| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 1504.1 | You must really hate coupling capacitors.... | DSSDEV::HALLGRIMSSON | Eirikur, CDA Product Manager | Tue Jul 05 1988 15:21 | 13 |
| An active case ground? The mind croggles. Are you sure that this
won't be ground-loop city? What happens when something with a three
wire plug ties your case to the electrical service ground? It's
really quite clever but it doesn't give me a good feeling. It should
be safe in that it is on the far side of a wall bug, but the case
could well be trying to be 1/2 the supply voltage offset from "real"
ground.
A head pokes up over the wall and sees the diagram.
"What's that?" Takes a closer look. "Serious perversion!"
--Eirikur
|
| 1504.2 | TLO-72 JFET Low Noise Op Amp. | SUBSYS::GLORIOSO | | Tue Jul 05 1988 17:59 | 5 |
| If you want better signal to noise, try the TLO-72 instead of the
TLO-82. It's the same pinout and is still inexpensive. However,
you can't get 'em at Rad Shack. Try Active Electronics in Westboro,
or mail order from Digi-Key or the like.
Scott.
|
| 1504.3 | Method in the madness... | MIDEVL::YERAZUNIS | by an unnamed spokesman | Tue Jul 05 1988 18:24 | 25 |
| Yes, active case ground. All the wall-bugs I tested are at least
110VAC isolated from case to either output lead (which you'd expect,
as a flipped-over insertion on a non-polarized bug could then give
you 110VAC on the outside of the barrel connector. UL would have
kittens...)
And no, there is no ground-loop problem whatsoever. It works very
very well indeed. Most ground-loops I see are relatively high
impedance (like 100K or so). The active ground here is more like
10 ohms.
Here's another way to view it: the active ground op-amp is not driving
ground, it's alternately tugging the wall-bug output lines up and
down around whatever the 1/4" cord shields decide "ground" is. This
makes a very nice split supply out of a cheap wall bug (and keeps
me from having to put a hummmmmy transformer inside the mixer case).
-----
I'll make another try at measuring the residual noise tonight, now
that I found my DVM (it was in the bottom of my gig bag). If it's
worse than 90 dB, I'll go get some TL072's right away.
(I only used the 082's because they were on the wall, a mere 6"
from my fingers. Sorry to use the R-word in your file. :-) )
|
| 1504.4 | Real EE's don't need interstage DC blocking... | MIDEVL::YERAZUNIS | by an unnamed spokesman | Tue Jul 05 1988 18:36 | 11 |
|
Added info: I have a couple pieces that _do_ tie the case to service
ground (the ESQ and the Xpander both). It's not a problem, it's
a feature. And it doesn't cause ground loops.
The only "problem" I had in this project was discovering that
one stomp box puts hash on the input. I thought _that_ was a
ground-loop problem, but it persisted when I shifted the stomp box
over to battery power! That's where the resistor in the EFF OUT
line came from...
|
| 1504.5 | Just a tad more info pleeeze | SALEM::AMARTIN | MY AHH..DEEDAHZZ | Wed Jul 13 1988 00:22 | 6 |
| I really do not know what all this stuff is but, I printed it off
and am having my uncle (electronic type) build it. Is it possible
to get (his words) part numbers or some sort of number for the parts???
The will aminly make it easier for him to locate them. Thanks in
advance. AL
|
| 1504.6 | Ultra-Simple 2x1 input-merger??? | FGVAXZ::LAING | Jim*261-2194*DEC MemorabiliaCollector | Wed Jul 20 1988 18:02 | 11 |
| I'd like to make a much simpler, 2- or 4-in, 1-out, mono mixer.
It would be used simply to "add channel(s)" to my current 8-ch
mixer in an "emergency". I wouldn't need stereo, FX loops, etc.
As I'm not any good at electronics, could someone describe to me
how I'd make such a thing? Maybe I'd think if it as "a Y-cord
for summing inputs, that WORKS", i.e. SIMPLE way of merging a
couple of inputs. Would it have to be active (i.e. powered?)
Any simple plans for such a gadget would help me out alot ...
-Jim
|
| 1504.7 | may be cheaper to buy ... | MIZZOU::SHERMAN | incompetence knows no bounds | Thu Jul 21 1988 10:11 | 5 |
| How about trotting down to Rat Shack and shelling out $20 for their
switch box/mixer? It can mix up to 3 or 4 stereo channels, as I
recall. The internals are passive (resistors).
Steve_who finally_settled_with_a_Rat_Shack_$120_7-channel_mixer
|
| 1504.8 | Radio Shack mixers | HPSTEK::RHODES | | Thu Jul 21 1988 10:25 | 6 |
| >Steve_who finally_settled_with_a_Rat_Shack_$120_7-channel_mixer
Stereo? What features.
Todd.
|
| 1504.9 | Rat Shack mixer modifiable! | CTHULU::YERAZUNIS | Smurf _Terminator_ | Thu Jul 21 1988 11:10 | 19 |
| The Rat-shack 4-channel has a few misfeatures:
1) it's very noisy (doesn't use op-amp, just two transistors);
2) it's very easy to overdrive;
3) it sounds _awful_ when overdriven (not a nice "overdrive", more
like an ugly clip.)
4) The stereo mode is pretty much useless.
However, it does have reasonable pots and jacks in it. Buy one, rip
out their circuit (which is a 2-transistor amp), and throw in the
op-amp active ground/mix bus/buffer in .0 Route the EFF SEND wire to
OUTPUT. Bag the .0 EFF RET amp, jacks, etc.
That ought to quiet it down and make it much more insensitive to
overdrive.
|
| 1504.10 | | MIZZOU::SHERMAN | incompetence knows no bounds | Thu Jul 21 1988 13:12 | 6 |
| Well, the $20 'mixer' I was thinking of is supposed to be used to
hook a couple of tape decks to a stereo. It only has resistors
and switches inside. The $120 mixer is their top-o'-the-line unit.
It has good frequency response, but otherwise is minimal on features.
Steve
|
| 1504.11 | Passive mixers? | FGVAXZ::LAING | Jim*261-2194*DEC MemorabiliaCollector | Thu Jul 21 1988 13:51 | 7 |
| Will passive internals (resisters, note .7) handle impedence
mis-matches and level mismatches? Will it add lots-o-noise?
If it's passive, "only" resistors, maybe I could make one? What
would a schematic be for an ultra-simple, 2x1 mono passive
"mixer"? Sounds like for 4 or 5 bux I could make one up?
-Jim
|
| 1504.12 | Passive will work, within limits. | CTHULU::YERAZUNIS | Smurf _Terminator_ | Thu Jul 21 1988 14:02 | 28 |
| I built a 2x1 mono passive mixer, it worked OK but not great (levels
on the output were very low, and because of the high output impedance
it was very susceptible to interference, hum, etc. Under clean
conditions it was quite quiet, but it tended to pick up hum if you
routed the output cable near a power line. It didn't handle
impedance mismatch very well; it just got quiet when it ran out
of oomph.
Why are you afraid to make an active mixer? You probably already
have a wall bug around, and the BiFet chips are not particularly
expensive or static-sensitive or delicate. The amount of work is
trivially greater (believe me, the biggest work item in the whole
mixer is drilling all them holes for pots and jacks!). At least
try it with a 9V alkaline battery first, that may whet your appetite
and you can certainly go gigging with it, as long as you bring a
spare battery.
Spend $8 on a good wirestripper (like the Snapit "beak") and it'll
go together so smoothly you won't believe it.
-----
If absolute cheap is the order of the day, sure, go ahead and build a
passive 2x1. Get a big enough box so you can make it 4x1 or x2 when
you get psyched. I really recommend putting it all in a metallic
box because of the hum pickup problem associated with unshielded
high-impedance wires.
if you don't you're going to
|
| 1504.13 | how to build a passive | CTHULU::YERAZUNIS | Smurf _Terminator_ | Thu Jul 21 1988 14:04 | 4 |
| How to build the passive mixer: Just build only the input fader
circuit in .0 (2 copies) and connect the "A" mix bus to the output
jack.
|
| 1504.14 | More info please??? | SALEM::AMARTIN | My AHDEDAHZZ REmix, by uLtRaVeRsE | Fri Jul 22 1988 00:06 | 2 |
| Yes but how bout telling me what the part numbers are for these
parts???? I have been waiting....please???
|
| 1504.15 | The part numbers are unimportant. Really! | CTHULU::YERAZUNIS | The light that burns twice as bright burns half as long | Fri Jul 22 1988 11:39 | 43 |
| The only important part number is
TL082 !
Every other part on the list is sold by specification, not by an
arbitrary part number. Seriously! Go into a Radio Shack and look
on the pegboard! They often put price stickers over the part number,
but they never obscure the specification.
For example
1K resistor
_ANY_ 1K resistor (1000 ohms, K means 1000) will work. And even
the worst of the weenie salesmen at Radio Shack know what part that
is.
I suggest you just print out the plans (just type PRINT .0 ) and
take it with you to RS. They should be able to help you out.
Again I repeat myself: Radio Shack stock numbers are UNIMPORTANT
(and change!) The specification is the ONLY important part!
(by the way, TL082 isn't a RS stock number, it's a generic type
specification...)
The TL082 chip has what each pin is connected to printed right on
the back of the bubble-pak card. Pin 1 has a little dot on top
of the package, and they count around from there (see the drawing
on the back of the card. It's pretty clear).
If you're still totally confused, perhaps you ought to first buy
one of the "Introduction to Electronic Projects" type books, like
the Radio Shack "Engineers Notebook I" (which aint, but it's about
the right thing for someone starting out).
Buy the book first (about $4), read it, and see if things become
clearer. Or visit the local library, they often have introductory
electronics books that show you the things you need to know.
-Bill
|
| 1504.16 | Thanks. | SALEM::AMARTIN | My AHDEDAHZZ REmix, by uLtRaVeRsE | Sat Jul 23 1988 00:01 | 2 |
| THANK YOU!!! Really app. it. I wasnt sure if it was or wasnt
important. Sorry for pestering you. :-) Thanks again. AL
|
| 1504.17 | A little more stereo | CTCADM::NICKERSON | | Mon Jul 25 1988 08:34 | 8 |
| Bill,
I must be missing something. Could you elaborate on making the mixer
stereo? How would you handle an ESQ and a TR-505 which are stereo and 2
guitars which are mono? How would you balance the guitars across
channels.
Dana I'm_going_to_build_this_mixer
|
| 1504.18 | Maybe this helps.. | CTHULU::YERAZUNIS | The light that burns twice as bright burns half as long | Mon Jul 25 1988 11:27 | 121 |
| It would depend on how I wanted to handle each of the instruments.
I would probably handle the ESQ (at least) by hard-panning one output
hard left and one hard right. Probably likewise on the TR-505.
Now the guitars are a tad different. If you don't mind having GUITAR
LEFT and GUITAR RIGHT as separate gain controls, build an 6x2 mixer.
If you really *must* have GUITAR GAIN and GUITAR PAN, it gets a bit
hairier.
-----
Now for the nitty-gritty changes.
First off, to give better stereo separation, I would put buffer
amplifiers on each of the inputs. (that's the last circuit block
in .0) . These buffers are connected as follows: cut the wire from
the center contact of each input jack, wire it to the input lead
of the buffer amp. Wire from the output lead of the buffer amp
to the top end of the channel's gain potentiometer.
Second, (here's where you have to decide whether you want separate
left and right gains, or a single gain and a pan control: If you
go for separate gain controls, build a second input attenuator circut
(the circuit where the tail ends all connect to point A (mix bus))
for each input channel. This second set of circuits should have
their "tails" connected together but separate from the "A" mix bus.
Call this second mix bus "B". Now duplicate the effect buffer and
final buffer circuits for this B bus. A and B are now your LEFT
and RIGHT outputs. If you want a third bus (monitors, say) you
can do this yet a third time (but my guess is you'll really NEED
those input buffer amps if you want a third channel.)
If you really want CHAN GAIN and CHAN PAN, then it gets a tad
hairier. The following changes will work but it won't be "quite pro" as
pro pan pots have curves that look like this:
LLLLLRRRR
R L
R L
R L
and this one looks like
L R
L R
L R
L R
B
L R
L R
L R
L R
L R
There's really no way to get the "pro" style without getting special
pan-pots, but that will probably be OK for your purposes anyway.
Build an input circuit that looks like this:
from input jack (or input buffer amplifier)
<------------
|
| -----------------------------------
| | |
| | ---| |---
\ | | \ to A bus / 5K
chan / <-----o | /<---\/\/--- \<---\/\/---
gain \ | | \ 5K / to B bus
(50K) / | | / \
| -----(--| PAN(50K x 2) |
| | |
| | |
GND GND GND
Now the tricky part- the second and third pots are really THE SAME
pot. You need to get a stereo (sometimes called 2-gang) linear taper pot
and wire it as shown (one side gets more gain as the other side
gets less). You may need to go to someplace other than Radio Shack
for this part. Sometimes they don't come this way, but they are
made to order by "stacking" single-gang linear pots with a tinkertoy-
like connector between them. That's fine. Ask the electronics
dealer deskperson for help on this one, as what you use will vary
with what they have in stock (and what lines they carry). You can
use anything from 50K up to 1 meg for the panpot, it really doesn't
make a big difference as long as all channels use the same value
of panpot, and the panpot value is equal to or greater than the
gain pot value.
(note- an audio-taper stereo pot will NOT work here! Sorry!)
Don't forget the resistors (5K's) between the output of the pan
pot and the mix busses. They are necessary!
I strongly recommend you use input buffer amplifiers for this circuit.
Now the last part- the guitars! Guitars put out a much weaker signal
than keyboards do. You may get away with just plugging in your
guitars and cranking up the guitar signal output; but that still
may not give you enough OUT. You can fix that by adding/substituting
a different input buffer amplifier.
The easy way to do this is to use the same circuit that you already
have to build for the EFF SEND buffer, and use that on the channels
that will be high-gain (for guitar). (the EFF SEND buffer is the op-amp
circuit that had a diddle-pot in it.) Build two (or four) extras so
you can put them into the guitar inputs, instead of the standard
input buffer inputs.
-----
Does this help ? (and now do you see why they charge so much for
mixers? With so many potentiometers, there's a lotta bucks tied
up in things that don't ICify well. )
-Bill
|
| 1504.19 | moved from 1549.* | GIBSON::DICKENS | booting system -> toe pain | Mon Jul 25 1988 18:16 | 155 |
|
This really belonged here as a reply, so I moved it.
Bill, if you'll delete 1549.1 I'll delete 1549.0. ... -Jeff
<<< NOVA::$111$DUA0:[NOTES$LIBRARY]COMMUSIC.NOTE;1 >>>
-< ** Computer Music ** >-
================================================================================
Note 1549.0 would-be mixer builder fantasizes 1 reply
GIBSON::DICKENS "Surfing with my Buick" 36 lines 19-JUL-1988 16:39
--------------------------------------------------------------------------------
I'm building the 4x1 mixer, without the effect loop, which I don't
need, at least not in the output path. I'll let you know how it
comes out.
Some questions:
How many input channels can I have within reason ? Would 8 be a
problem ?
If I can make it stereo by adding another mix-bus and mixer amp,
can I add a third and fourth bus the same way ?
Can I put the effects loop in the input channel instead of the output
by just putting two buffer amps before the fader, with the insert
point between them ?
How would I put a master-volume in the mixer amp ?
Eventually, I would like to build an 8x4 mixer, with two busses for
stereo out, and two for effects sends. Two of the inputs would be
dedicated to the effects returns, giving me a 6x2 mixer with two aux
busses.
But before I go off the deep end here I'm going to build a smaller
model first. Also, I'd want to re-arrange the controls a little, so
there would be a fader, pan, and send 1&2 instead of 4 sends for each
channel. And by now I'm already thinking about pricing some miniature
pots, etc to maybe make the thing a little less unwieldy. My gut feel
is that yes, (if the audio quality is sufficient), this would be worth
building. Recently I've heard of a lot of people building their
own EQ-less mix matrices for small studio work. Is this what they're
doing ?
================================================================================
Note 1549.1 would-be mixer builder fantasizes 1 of 1
CTHULU::YERAZUNIS "Since when do electrons carry c" 100 lines 19-JUL-1988 20:00
-< The Mad Mixer! >-
--------------------------------------------------------------------------------
> Some questions:
>How many input channels can I have within reason ? Would 8 be a
>problem ?
Shouldn't be; if you're worried double the value of the resistors
that all connect to the mix bus. You might want to test this by
modifying your 4x1 prototype by adding 4 resistors going from the
mix bus to chassis ground. If you get adequate levels, then
the 8-input mod will be fine for your application.
>If I can make it stereo by adding another mix-bus and mixer amp,
>can I add a third and fourth bus the same way ?
Theoretically, yes. Practically, it depends on just
how low an output impedance your instruments put out.
It *might* be OK. If I was going to do build anything past
a 2-bus (stereo) mixer, I would put one of those unity-gain
buffers between each instrument input jack and the level
pot. That would assure you that your instrument wouldn't
run out of oomph, and also prevent any possible crosstalk.
>Can I put the effects loop in the input channel instead of the output
>by just putting two buffer amps before the fader, with the insert
>point between them ?
You still need the final output buffer amp, but you don't need
to put buffers around the input effect insert point. To test
this, plug the instrument directly into the effect and the output
of the effect into your 4x1 prototype input. If you're satisfied, then
go ahead.
However, if you want to input both the dry instrument and the
effect return on different pots, you probably should stick a
unity-gain non-inverting buffer in front of the effect send jack
so the digital hash coming back up the line from the effects
box can't clobber your dry signal.
>How would I put a master-volume in the mixer amp ?
Get a 10K audio-taper pot. Connect one end to chassis ground,
the other end to the EFF RET switching jack, and the wiper
then goes to the final buffer amp input.
>Eventually, I would like to build an 8x4 mixer, with two busses for
>stereo out, and two for effects sends. Two of the inputs would be
>dedicated to the effects returns, giving me a 6x2 mixer with two aux
>busses.
If you want EFF RET to show up in EFF SEND, you best go through
and count inverting buffers. Make darn sure that you have
an even number of them, else you can get into a very
nasty feedback situation. Also be careful of the max
gain; you can "lock up" the op-amps if you drive them
too hard even if there is no feedback oscillation.
Probably best to build it (electrically) as a 6x2 mixer, which
feeds (through buffer amps) a pair of EFF SEND jacks. Then
the outputs of the 6x2's are buffered _separately_ internally,
and go with the two EFF RET's into an independent 4x2 mixer.
That way, your effect boxes can't rake you over the coals.
You can build it all into one box; just keep the two pieces
electrically isolated except for case ground.
>But before I go off the deep end here I'm going to build a smaller
>model first. Also, I'd want to re-arrange the controls a little, so
>there would be a fader, pan, and send 1&2 instead of 4 sends for each
>channel. And by now I'm already thinking about pricing some miniature
>pots, etc to maybe make the thing a little less unwieldy. My gut feel
>is that yes, (if the audio quality is sufficient), this would be worth
>building. Recently I've heard of a lot of people building their
>own EQ-less mix matrices for small studio work. Is this what they're
>doing ?
I don't know what other people are doing, but this is what
_I'm_ doing. It's fun, and I get exactly what I design
(not necessarily what I want, but that's another story
altogether :-) )
Build the 4x1 first, and see if you can hear any noise
anywhere. (if you don't buffer the inputs, there is a
possibility of crosstalk on 4x2's, but it's benign and
variable depending on your instruments output impedance.
Buffering the inputs with the unity gain buffers will
fix that cleanly.)
--------
Active Electronics (Westboro) has TL072 op-amps for the low,
low price of $6.75 for eight. That's considerably cheaper
than Rat Shack, and for a supposedly better op-amp too.
Today I went looking for A/D converters.... is the world ready
for homemade digital mixing boards?
:-)
Bill
|
| 1504.20 | no smoke, but... | GIBSON::DICKENS | booting system -> toe pain | Mon Jul 25 1988 18:40 | 24 |
| Well, I built the 4x1. After getting over the fear of blowing my
ESQ and my Porta-1 to El Salvador, I test. I get loud hum, no signal.
But no smoke, either.
Found an open ground, test again, still don' work. Tiny bit of
signal, but no hum.
Grrrrrrr. Next time I build on breadboard first. Grrrrr again.
At this point I thankfully had to go away to the country for the
weekend.
Now I have to breadboard a test circuit so I can figure out whether
the TL082s I have are burnt from my misadventures. Any hints on
this ?
I've already checked the circuit with my ohm-meter (with the IC's
out) and every thing seems logically correct. No solder bridges
or other ugliness apparent. Any hints on how I might proceed ?
Thanks.
-Jeff
|
| 1504.21 | counting those inverting amps | GIBSON::DICKENS | booting system -> toe pain | Mon Jul 25 1988 18:53 | 31 |
| re "making sure you have an even number of inverting amps" from
.19, aka 1549.*:
Isn't this an inverting amp that you're using as the mix amp ?
V++ from wall bug-------------------
|
|
|\ |
| \|
Case GROUND---------------------|+ \ 1K res.
| \___._______.____/\/\/\----To EFF SEND
| / | | jack tip
------- |- / | |
| | /| | |
| |/ | | |
V | | ----To shorting contact on
Joined "A" wires-----/\/\/\/-------)----- EFF INSERT jack (the one
10K trim | "closed-circuit" jack)
|
|
|
- from wall bug---------------------
I'm probably exposing the shallowness of my knowlege of analog
electronics here, but please tell me why it's not an inverting amp.
Thanks,
-Jeff
|
| 1504.22 | | GIBSON::DICKENS | booting system -> toe pain | Mon Jul 25 1988 19:02 | 11 |
| Another question:
Now that we're building guitar pre-amps, has any given any thought
to using some kind of overdrive or other means of getting a "chewier"
sound from the guitar ? Anyone built an ic "voicing preamp" ?
Of course what we really need is a good ol' 12ax7 (?) in there for
natural chewiness or (what the 'ell) crunchiness !
-Jeff 8-; (half-serious)
|
| 1504.23 | Debug Script | CTHULU::YERAZUNIS | Gordian Knot Lock Co. | Tue Jul 26 1988 11:46 | 122 |
| Yep, the circuit in .21 is an inverting amp (you are right).
The hitch is that there is only one of them in the signal path IF
you loop EFF SEND back in on a regular input (not on EFF RET).
That means that a rather nasty oscillation can happen, unless you
insert another inverting amplifier somewhere in the chain.
Or you could do a somewhat more complicated gain-controlled op-amp
circuit and make the .21 circuit not invert.
----
Re: mixer not working (didn't expect to have to write a debug script
this early)...
0) It's unlikely you've blown the op-amp chips; they're pretty tough
little buggers. You might want to pick up a spare if you've really
been abusive, but it's unlikely you'll need them.
remember that a ")" in any circuit diagram means THESE WIRES CROSS
BUT DO NOT CONNECT!
1) First things first. Check to see if the power stabilizer (active
grounding) is working correctly. Use a meter, and check that the V++
and V-- busses are at about +6 and -6 volts (compared to the case). If
that isn't so, then there's something wrong with the active ground
circuit.
1a) Check to see that you have 12V between V++ and V-- AND
that it's the right way (I recall that pin 4 of the TL082
wants to be - and pin 8 is +. Check that first off. If
it's backwards, the chips probably survived but it won't
be a very clean mixer :-).
1b) Check the junction of the two 5K resistors at the head of
the active ground circuit. THAT junction should be
at the +6/-6 volt point. If it isn't, then the rest of
the system will never see the correct ground reference and
nothing will work. The only possible failures here are
that the 5K resistors are bad.
1c) Check the output pin of the TL082 (either pin 2 or pin 7,
I recall). It should be at the +6/-6 volt point. If it
is, and the case aint, then you have a bad connection to
the case, resolder it.
2) Power is OK, but nothing seems to be coming out. Time to check if
the signal is making it to the mix bus. Connect a couple of jumper
wires from a 1/4" plug (tip and sleeve). Connect the sleeve to case
ground, tip to the "A" bus. Connect the other end of the 1/4" plug
into a sound reinforcement amp. REMOVE the IC's.
2a) Do you hear anything? Hum? If so, you've got hum coming in
from somewhere else, because there's no chips in the circuit
at this point.
2b) Plug a keyboard into an input, pot up that input, play something.
You should hear a signal coming out; it'll be weak but definitely
there. The gain pots should be able to control it. If this
doesn't work, you've got a mistake in the first circuit (input jack
potentiometer circuit, the one with the 100K pots.) Check all 4
channels, while you're at it.
3) If you got to here, you know you have good power and ground and
that the mix bus itself is OK. Time to check the EFF SEND amplifier.
Remove your 1/4" cable, put the chips back in.
Plug a 1/4" cable into your sound reinforcement amp, and touch the
other end with your finger. Hear hum? Yes is the right answer.
If you don't hear hum, turn up the sound reinforcement amp till you
do. Now turn it back down until the hum is just about gone.
Position the 10K trim pot at about mid-point.
Plug the 1/4" cable into EFF SEND. Play something. You should
hear about the same level of signal as you heard with the 1/4"
plug jumpered to the mix bus. You can adjust the gain with the
trimpot.
3a) If you hear something faint, but the trimpot doesn't bring
the gain up, then either the trimpot is bad, the chip
is bad, or the wiring is bad. Check the trimpot with your
ohmmeter. Swap chips with the power stabilizer and see
if the chip works there. If so, then check your wiring!
3b) If you don't hear anything, then remove the 1/4" plug and
jumper it to the output pin on the chip directly (might
help to have the sequencer looping a short tune for you).
If you hear it (and the gain trimpot works), then your wiring
from amp to jack panel is bad, or the EFF SEND jack is bad
(and the chip and trimpot are OK.)
3c) Just on a hunch- insert a 1/4" STEREO cable into EFF INSERT
but leave the other end of that cable disconnected. If things
start working now, you've wired the EFF INSERT jack backwards!
3c) If you get a good signal (and can vary it up to very LOUD)
out of EFF SEND with the trimpot, then the EFF SEND amp
is OK.
4) Only one place left now: the final buffer amp. Pull your cable
out of EFF INSERT. Plug cable from sound reinforcement amp into
MIX OUT. Play something. You should hear it, gain variable with
the 10K trim pot on the mix bus amp.
4a) You don't hear anything (or it's very weak). Pull the
sound reinforcement amp cable, jumper sleeve to ground and
tip to the output pin of the IC. Play something (sequencer
handy here). If you get good output, then either the
wiring from chip to MIX OUT is bad or the MIX OUT jack itself
is bad. If you don't get anything on the IC pin, then
the chip is suspect.
4b) You now have good signal all the way through. Why are you
troubleshooting? If it ain't broke, don't fix it! :-)
Hope this helps...
-Bill
|
| 1504.24 | it works ! | GIBSON::DICKENS | booting system -> toe pain | Tue Jul 26 1988 15:02 | 47 |
| Wow. Great job in .23.
It turns out I got it working last night anyway.
When everything was wired right, I plugged it in to my Porta-1,
which I was using as a headphone amp, and hooked up the ESQ to
an input. Then I punched up the ever-popular GROWL patch. (You
said loud)
I noticed that even with no signal coming from the synth, the vu
meter was going up and down in a slow oscillation. Strange, I thought,
since I couldn't hear anything. When I played at this point, I
got a very weak and distorted signal.
Then, while I was twiddling things trying to get it to work, I noticed
that if I turned down the input pad on the mixer (from the mic.
setting towards the line setting) I would get more signal, but still
not enough.
So I tried a 270 ohm resistor in the output line, and bingo, it
works. However, in calibrating it I found that it wouldn't distort
even with the trim-pot turned all the way up. Does this mean 270
is too much ? Also, now it works best with the input pad on the
mixer about at halfway. Would it be better if the mixer output
was at line level, or at mic. level ?
Also, do you have any idea what was happening before I put the output
resistor in ? It sure looked weird.
In general, though, it sounds Great ! I was pleasantly surprised.
It seems to have more than adequate frequency and dynamic range.
Also, I could not detect any hum or hiss at all at any setting.
And I haven't even closed the case yet. Of course that might have
something to do with the fact that I used mini-coax to connect up
all the jacks and pots to the board, and grounded the shields.
Another weirdness: The volume response of the pots is not linear at
all. Even when turned all the way down I still get some signal, and
then you can turn it 3/4 of the way up with little change. Then the
last 1/4 is useful. These are the cheapo rat-shack 100k audio taper
pots, or so the package said.
Thanks again for all your help, Bill.
-Jeff
|
| 1504.25 | | GIBSON::DICKENS | booting system -> toe pain | Tue Jul 26 1988 15:07 | 10 |
| > The hitch is that there is only one of them in the signal path IF you
> loop EFF SEND back in on a regular input (not on EFF RET).
IF ? I thought that the final buffer amp was non-inverting too.
I still don't get it. If the output of the mix amp is inverted,
when does it get un-inverted ?
-Jeff
|
| 1504.26 | Doesn't matter. | IOENG::JWILLIAMS | Zeitgeist Zoology | Tue Jul 26 1988 16:43 | 6 |
| For most audio purposes, it doesn't matter if it's inverted or non
inverted, with one exception: Stereo outputs must be the same polarity.
Otherwise, your ear can't tell for beans.
John.
|
| 1504.27 | Good! | MIDEVL::YERAZUNIS | Just a puppet who can see the strings. | Wed Jul 27 1988 12:39 | 21 |
| re .24
MAJOR WEIRDNESS!!! You mean you need a 270 ohm resistor
between your MIX OUT and the Porta-1?
Possibilities: the Porta-1 has a lo-Z input all the time and
depends on the previous hardware to (hopefully) have an inline resistor.
That's just barely possible. It makes sort-of sense if we remember
that the TL082's shut themselves down on short circuit. That slow
cycling may have been the TL082's taking themselves on and off line!
I hope the cycling is gone now. You might want to up the 270 ohm
resistor up to 1K ohm.
Now do you believe just how incredibly quiet these active-ground
systems can be? So they're perverted? Who cares?
-Bill
|
| 1504.28 | Thoughts on pots | MIDEVL::YERAZUNIS | Just a puppet who can see the strings. | Wed Jul 27 1988 12:43 | 9 |
| re the pots doing all their work in the last 1/4 turn:
Do they turn left-low to right-high (standard) or backwards from
that? If they're backwards, that would explain it; you're on the
wrong side of the audio taper! If not, then I gotta go think
on whys and wherefores...
Good idea using coax, that certainly will make it quiet. Gee, you
might have gotten away with a plastic box then.
|
| 1504.29 | midi-processor project | GIBSON::DICKENS | booting system -> toe pain | Fri Jul 29 1988 17:38 | 38 |
|
Next project: Rather than chase after a digital mixer, I would
rather build a generic midi processor.
Cheap Z-80 SBC or other cheap micro + small memory + midi interface
(oh, and then there's that soft stuff)
The biggest problem I could see is how to develop the s/w.
We'd need a development system with a cross-compiler and a prom
blaster. I know, I know, an 80's weenie who just isn't up to entering
code with a set of switches...
My first application would be:
A "gate machine". It would store a set of 16 binary gates and then
recall them via a midi patch select command. The gates could then be
built into your cheap mixer, providing mute-automation, effect on/off
automation, or anything else (binary) your imagination could come up
with. Once this is done it should be straightforward to add more
gates.
If I was going to build one I would build it into the mixer cab to
reduce the spaghetti factor. Since I just ran head-on into several
hundred mil-spec 1/4" jacks, I think I'll put a patch bay in there
too.
Once we've developed the basic midi transciever s/w and h/w, we should
be able to build a crude mapper or any other midi-gadget that could be
dreamed up.
Another thing that comes to mind is a buffered MIDI to RS232 converter
that would allow you to dump sysex data upline to a vax, etc.
With some good VCA chips we could even build our own automated mixer !
What thinks you ? Could it be done ? Might it not be a lot of fun ?
-Jeff
|
| 1504.30 | It could be interesting. | PANGLS::BAILEY | | Mon Aug 01 1988 16:17 | 16 |
| > cross compiler...
Probably better to use assembler, especially since a Z-80 cross
assembler, written in C, is availible in the DECUS-C distribution.
You might want to look at a micro which has some stuff already built
into it. The 6801 springs to my mind, since it has some RAM, a
serial port (I think it might be fast enough for MIDI) and some
parallel I/O. There is also a 6800 cross assembler in the DECUS-C
distribution, I think.
I have often thought of building such an item, but I always spend
my energy writing code for my current MIDI system (an Atari-ST).
It seems like it would be fun, though.
Steph
|
| 1504.31 | Ah, a bite ! :-{|} | GIBSON::DICKENS | booting system -> toe pain | Mon Aug 01 1988 17:38 | 15 |
| I wouldn't mind if it could only only respond to patch change and
control the 16 gates. That shouldn't take much speed... All I'd
need for backing store is a NVR. Is that difficult to do/get ?
I realize to do any real time work we're probably talking 68000
instead of 6800. I'm a *rank* amateur in this h/w biz, but I'm
interested.
I'll need something to keep me busy this winter !
Yes, cross-assembler was what I meant. We're talking "small and
tight" here.
-Jeff
|
| 1504.32 | Cheap micros | MCIS2::ROACH | | Mon Aug 01 1988 18:44 | 2 |
| I can't verify this, but I think someone told me that the processor in
Yamaha's FB-01 is the Z-80.
|
| 1504.33 | Don't over-design. | PANGLS::BAILEY | | Mon Aug 01 1988 19:14 | 22 |
| Most 8-bit micros are more than enough for the task, especially
if you don't want it to give you a mega-pixel, bit-mapped user
interface. A 68000 (or almost any other 16-bitter) would require
enough support components to be annoying.
I think the easiest way to do NVRAM is with a CMOS low-power-standby
RAM and a lithium (non-charging) battery. More fancy schemes (like
EEPROM) may require funny voltages (like +12--a working scheme could
probably use only +5).
The best way to do a technology-driven design like this is to get
spec sheets for some parts (like the 6801) and estimate what you
can do with it, and in how many parts. Write some test code for
doing what you want, and calculate the time that it will take to
execute it.
Designing hardware from old, commodity parts is as easy as falling
off a log. Or at least as easy as reading from a cookbook.
I have also heard that the FB01 uses a Z-80.
Steph
|
| 1504.34 | I <heart> NVRAMs | MIZZOU::SHERMAN | socialism doesn't work ... | Tue Aug 02 1988 10:17 | 15 |
|
> I think the easiest way to do NVRAM is with a CMOS low-power-standby
> RAM and a lithium (non-charging) battery. More fancy schemes (like
> EEPROM) may require funny voltages (like +12--a working scheme could
> probably use only +5).
Not so! The really *good* NVRAMs have on-board charge-pumping and
don't require the extra voltage sources. In addition, they have
automatic save features that detect power outage and save the contents
within about a microsecond.
Steve_who_designed_such_a_beast_while_working_for_Mostek_and_who_knows
_*all*_about_Fowler-Nordheim_tunneling_elements_;^)
|
| 1504.35 | | GIBSON::DICKENS | booting system -> toe pain | Tue Aug 02 1988 16:36 | 1 |
| Keep 'em coming, the idea is developing...
|
| 1504.36 | | GIBSON::DICKENS | booting system -> toe pain | Mon Aug 08 1988 16:37 | 12 |
| More parts prices:
100K mini audio-taper pots are $1.31 each in quantities of >=50
at U-doit. Roughly 1/2 off.
Anyone interested in a cooperative buy ? Anyone have a better price
?
Also looking for rack-mount cabinetry cheap..
-Jeff
|
| 1504.37 | Me, too | AKOV88::EATOND | Moving to NRO! | Mon Aug 08 1988 16:51 | 9 |
| RE < Note 1504.36 by GIBSON::DICKENS "booting system -> toe pain" >
> Also looking for rack-mount cabinetry cheap..
If you find any, I'm interested, too. I've been looking through
catalogs for a long time and have found no inexpensive sources.
Dan
|
| 1504.38 | Some interest here | MIDEVL::YERAZUNIS | The light that burns twice as bright burns half as long | Mon Aug 08 1988 17:51 | 4 |
| Are these rotary pots or sliders for $1.31 ea? (I'm thinking I
want to go to an 8x4 board myself, and with sliders for
channel gains. )
|
| 1504.39 | | GIBSON::DICKENS | booting system -> toe pain | Tue Aug 09 1988 12:46 | 7 |
| Those are rotary pots. I'll ask about sliders next time I'm there.
They wanted in the $20-30 range for just a blank relay rack panel.
I thought that was a little pricey. I'm looking into getting some
sheetmetal custom bent at a sheetmetal shop. A friend of a friend,
etc...
|
| 1504.40 | DODO TIME........ | SALEM::AMARTIN | My AHDEDAHZZ REmix, by uLtRaVeRsE | Wed Aug 10 1988 06:09 | 10 |
| DUMB QUESTION TIME AGAIN!
Is it better/worse/safe etal to use a heavier box instead of sheet
metal or what ever the dir's call for???
What I mean is.....
Would it make it a quieter mixer to user a heavier metal??
Good, bad, indifferent Ideas????
|
| 1504.41 | It'd Be More Durable Tho.... | AQUA::ROST | Life is serious, but art is fun | Wed Aug 10 1988 09:37 | 7 |
|
Re: .40
Heavier metal won't improve the shielding.
|
| 1504.42 | Use stranded wire for anything that might move. | CTHULU::YERAZUNIS | I don't know about apathy or ignorance, and I don't care! | Wed Aug 10 1988 11:01 | 15 |
| ...but harder to drill the holes. Do you plan to do heavy gigging?
(another hint: use stranded wire EVERYWHERE except for connections
between points both on the IC perfboard (use solid wire there).
Wire fatigue can become a problem with solid wire.)
-----
I just used a Rat Shack metal box- the one with the black top and the
little rubber feets.
A reattempt at measuring residual noise failed miserably.... it's
DARN QUIET. That's all I can really say.
-Bill
|
| 1504.43 | Rack mount enclosure source ... | DECWET::BISMUTH | Wait and think. | Wed Aug 24 1988 03:35 | 28 |
|
Meant to post this a while ago - too busy to keep up ...
For those looking for rack mount enclosures, I discovered that:
ANGO Electronics Corp.,
16 N Third Street,
Harrison, NJ 07029
Tel. 201-484-1115 (Fax 201-484-0530)
makes a nice series of enclosures. They go from 1 to 5 su's in height,
standard ecma mounting dimensions. Variety of guages of aluminum
are available, either natural finish, anodized black or beige.
Prices start at around $30 (one of).
Their units break down into essentially flat sections, which makes
layout, drilling, etc. easier.
They also have some (not a lot) of hardware for making up rack mounts
cabs.
Payment terms are the usual credit cards, cash, check or COD.
Robert
|
| 1504.44 | Sounds like fun | TOWNS::MUSUMECI | | Mon Dec 12 1988 18:49 | 79 |
|
Bill,
First off let me thank you for including your "Cheap, Quiet Mixer Plans"
in the COMMUSIC notes file this past summer. It was the inspiration that
got me to try and build some simple audio circuits. I have a good bit of
audio expeirence but none in electronics. I've been fooling with different
OP amp circuits for the past 2 months and have built some simple circuits.
My goal is to make a 8x2 mixer with 6 line level inputs and 2 low level
inputs. I have 2 prototype circuits that I'm now playing with. The first
is a 2x2 line level stereo circuit using the stereo input circuit that you
showed in the notes file(the one with the dual linear ganged pan pot). I
used a 100k pot for gain and 100k for the dual pan pot. I have 2 of these
and one goes to the A mix out amp ( op amp with 10k trim) and the other to
the B mix out amp. the output of these go to inputs of my Tascam 244 porta-
studio which is input to a power amp. I also have a 2x2 low level stereo
circuit that is the same as the line level execpt the inputs have a 10k
trim op amp before the 100k gain pot. I build these circuits on a solderless
breadboard. I use the active case ground ( tho I have know ideal how it works)
and 2 9 volt batteries. The line level circuit sounds nice and clean but I
have to turn up my 244 to about 6-7 to get a good listening level. I use the
output from my cassett deck for input to this circuit.
The low level circuit when used with a high imped. mike is very low. I have both
10k trims up all the way and the 244 up to 8-9 and the sound is about the same
as plugging the mike directly to the 244 with the gain on 2. Now I don't pretend
to know how to trouble shoot circuits but I believe I have them wired as shown
in the notes file( I got caught wiring the dual ganged pot wrong but figured
it out). while we're on that subject, I came across a stereo pan pot circuit
that I built that works real nice , uses a single ganged linear pot and seems
to give a "pro" style of pan. I built is as shown in the book and would like
to use it when I build my 8x2 mixer for the input pans. Can I use it as it
is with your input circuits or do we have to change the value of the resisters?
24.3k 10k 58.3k
In----*----/\/\/\---*----/\/\/\---*-----/\/\/\-------*-----Left
| | | | out
| | | | \ |
| | pan right ----|_ \ |
| 10k < ^ | \__________|
| -----> | ----|+ /
| | < v | | /
| --- | pan left --- | /
| - | -
| |
| 24.3k | 10k 58.3k
*----/\/\/\---*----/\/\/\---*-----/\/\/\-------*-----Right
| | out
| | \ |
-----|_ \ |
| \_________|
-----|+ /
| | /
--- | /
-
If you have the time I have some questions.
1. How can I get a little more gain on my line level circuit ?
2. How can I get alot more gain on my low level circuit ?
3. Can I intregrate the pan pot above into the your circuits ?
4. Will I get more gain by going to 12V power ?
5. Can I use a 741 op amp for the power stabilizer without degregation ?
Thanks for your time and the effort you put into the note in COMMUSIC
Chris Musumeci
|
| 1504.45 | | TOWNS::MUSUMECI | | Mon Dec 12 1988 18:51 | 144 |
|
From: 38821::YERAZUNIS "Running from the Turing Police" 8-DEC-1988 12:26:33.45
To: TOWNS::MUSUMECI
CC: YERAZUNIS
Subj: Re: Hello and thanks
Answering some of your questions:
5) Can I use a 741 for power stab?
Don't know. Haven't tried it. Might work, might not (power dissipation
capability, among other things). Try it, they're cheap.
4) Will I get more gain by going to +-12 volts?
No! The gain of an op-amp circuit is entirely determined by the resistors
it uses, and so long as the output doesn't try to go outside the power
supply range, it will never notice. (YOU will notice if the op-amp tries to
go outside the supply range- they will clip and distort badly!)
3) Integrate pan-pot circuit above?
Sure. Connect the real (pre-pan) input to an input pot (to control gain)
and then the output of that goes into your circuit above. Your "left-out"
and "right-out" then should get 5K resistors and the other end of the
resistor goes to the mix busses.
2) Getting a lot more gain on your low-level circuit?
Build a non-inverting buffer amp with gain, put the buffer amp
between the input jack and the "in" connector of your pan circuit,
and crank in as much gain as you want. See "Basic Op-Amps" below.
You want an inverting buffer with a 10K pot to control gain.
1) Get a little more gain on your line-level circuit?
Just crank the output signal amplifiers a little hotter. If you don't
have enough range, replace the 10K pot with a 100K pot.
---------------------------------------------------------------------
Basic Op-Amps
There are two general cases for op-amp circuits that amplify, one
doesn't invert and one does. The final buffer amp in the CQM is
non-inverting, while the the signal amp (the one connected to the "A"
mix bus) is inverting.
A: "Non-inverting Buffer" - this is where the input signal
goes into the op-amp "+" input, and the output signal is looped back
into the "-" signal. Sometimes there's a resistor or trimpot between
the opamp output, the "-" input, and ground, and sometimes one of the
resistors is missing (infinite ohms, as in no connection) The general
circuit looks like this:
in |\ out
-----|+\--------
--|-/ \
| |/ /R1
--------|
\
/R2
|
| to gnd
For this circuit, Vout = Vin [ (R1 + R2) / R2 ]. The bigger you make
R1 compared to R2, the more gain you get. The CQM final buffer amp is
a degenerate case of this- there R1 = 0 (direct connect) and R2 = open
(many megohms). This makes the term (R1+R2)/R2 approach very close to 1.
So, to get more gain in any amp that looks like the final buffer amp
(input on "+", output connects to "-" input):
1) CUT the wire between the "-" input and the output.
2) put a pot (10K should be fine) with one end on OUTPUT, the
other end on ground, and the moving contact to "-" input.
3) vary the gain as you wish with the pot.
Be careful- you can get huge amounts of gain with this circuit, but as you
approach max gain (R1 goes to infinity), the circuit will get very noisy and
may well go unstable.
If you're going to have an outside knob to adjust the gain, you can put
another _fixed_ resistor (say, 100 ohms) between the lower end of the
pot and ground. That way, the user won't be able to increase the
"effective" (R1+R2)/R2 to infinity and the circuit will stay stable.
-----------------------------------------------------------------------
B: The second kind of op-amp circuit is the inverting, summing amp. Here, the
case ground goes to +, and a bunch of resistors "sum up" the input signals
(maybe from more than one source) and THAT is connected to the "-" input.
The idea in the summing amp is that the sum of the currents going into the
summing node (such as the "A" mix bus curcuit) all adds to zero.
gnd |\
-----|+\____._____________________out
--|-/ |
| |/ \
| / Ro
---------|
|
V1--\/\/\---| R1
|
V2--\/\/\---| R2
|
V3--\/\/\---| R3
.
.
.
The Vn's above are input connections. The R's (except for Ro, the output
resistor) could be mix pots or could be fixed resistors. The CQM used
the 5K resistors on the input fader circuit for this purpose.
The output signal for any one of the inputs is
Vout = Vin (Rout/Rin)
So for the 3-input system displayed above:
Vout = V1(Ro/R1) + V2(Ro/R2) + V3(Ro/R3)
but this can be extended to any number of inputs. As insurance against
crosstalk, you should use a buffer amplifier to drive each of the Vi's
(see above for inverting buffer amplifiers). Theory says that in a perfect
world the inverting buffers are unneeded, but reality says they
prevent crosstalk. Use 'em if you need 'em.
-------------------------
Hope this helps; feel free to post your question and this response in
NOVA::COMMUSIC. It might help someone else someday too.
-Bill
|
| 1504.46 | Wow! What response! | CTHULU::YERAZUNIS | Turpentine, acetone, benzine... | Mon Feb 20 1989 17:56 | 11 |
| I'm _still_ getting occasional mail about how the CQM mixer plans
detailed here (thisnote.0 and onward)...
Just a question: How many of you out there have built one?
How big (4x1? 4x2? 8x2?) did you build it?
Did it work out OK?
-Bill
|
| 1504.47 | I GOT ONE AND USE IT TOO! | SALEM::AMARTIN | Pop any uppers lately Kitty? | Tue Feb 21 1989 04:35 | 8 |
| Well bill, the one that I (not really, someone else did it) built
is great. Aside from the probs that I had a while ago (memba??)
things are a ok...
OH, the 4x2 was the one for me..
Ill tell you something though, it is a bit noisy sometimes.
Not sure why that is but it is what I need for now.
|
| 1504.48 | I made one... | SMURF::NEWHOUSE | | Tue Feb 21 1989 09:05 | 8 |
| I made one, not quite from these plans, but almost identical.
8x2 with no effect sends no nadda. I used your wall bug power supply
jobber and it works beautiful. It was eating batteries very fast
and that became distressing during recording sessions. I was thinking
of posting the plans for the one I built since it is close to yours,
but is a little different. Line level only, pretty quiet.
-Tim
|
| 1504.49 | ECO #1 | CTHULU::YERAZUNIS | it's.. it's DIP ! | Tue Feb 21 1989 10:11 | 28 |
| Just to keep this note up to date, MUNCSS::BURKE asked if there
were any outstanding ECOs to the CQM plans. There is just one:
Add a resistor (value 1K to 10K, it's not critical) between the
tip contact (not the shorting contact!) and GROUND on the EFF
RETURN jack.
This resistor really cuts down the switching transient "KACHUNKKK"
when you "hot-jack" an effect into the EFF loop. It slightly decreases
output levels, but it's minimal.
------------
I'm not surprised that an 8x2 CQM would eat batteries ! The active
ground power stabilizer system tends to draw a fair amount of current
even when the final output signal is null (it does active nulling,
remember?)
If you wanted to run on batteries, you should disconnect the active
ground circuit and use TWO batteries (one to supply V++, one to
supply V--) and reference them both to case ground.
-------------
Go ahead and post the plans to the 8x2. Someone may want to build
one!
-Bill
|
| 1504.50 | Yes, please do. | NORGE::CHAD | Ich glaube Ich t�te Ich h�tte | Tue Feb 21 1989 11:05 | 9 |
|
Yes, please do post the 8x2. I need two 8x2 submixers for my sampler and soon
to be had R8. Thanks
Chad
(I'd like to build the CQM 8x2 variant as it is cheaper than buying and after
I buy the R8 the $$$ will be gone (not that they're there now :-) but that is
in Tom's topic :-) )
|
| 1504.51 | PLEASE POST 8X2 plans! | MRSVAX::MISKINIS | | Tue Feb 21 1989 11:12 | 3 |
| Yes, PLEASE POST THE 8 X 2 PLANS. I want to build one also!
_John_ (Man with NO mixer, except a little 10in 2out box)
|
| 1504.52 | ambitious? | HJUXB::LEGA | Bug Busters Incorporated | Tue Feb 21 1989 12:47 | 21 |
|
Has anyone gone beyond 8x2?
After reading this note awhile ago, and seeing the prices on new
8/16 x 2 commercial mixers, I am really into building a
16 x 2 with at least 3 effect sends, no eq, and panning on each
channel, mounted in a 2U rack space ala KM802.
I figure I could mass produce each channel on an individual card
with edge or right angle dip, and use a backplane of dips to buss
everything together.
Does this sound insane?
I even could enhance it with a bar graph LED for each channel, using
the circuit in EM a few issues back.
I'd use rotary pots to avoid making the front panel effort a metalshop
project. I'd also get the best (MIL-SPEC maybe) op amps, for the
best audio performance.
Waddya think.
?
Pete
|
| 1504.53 | Design decisions | CTHULU::YERAZUNIS | it's.. it's DIP ! | Tue Feb 21 1989 12:59 | 27 |
| You could do it. DIP sockets aren't the best (or even necessarily
the cheapest) backplane-building method, but they should work OK.
One thing I regret on my mixer is the non-slide-pot faders. I really
like being able to set gains with a rotary pot, and then "pencil" the
faders up and down together.
You also have to decide:
1) Do the effect sends read signal before or after the faders?
(in the CQM, they go after- but not everybody wants this)
2) Should the effect returns have their own faders or should it
be at a fixed level in the mix. (CQM is fixed- it's an
"insert loop", not a mix-back-in.)
3) Are the effect sends stereo? How about the returns?
-----
Does anyone out there have a listing for 16-bit parallel A/D and D/A
converters? We may want to cost out "going digital" versus a pure
analog board (my hunch is that the digital board will be about equal
in price, because we can get rid of so _many_ potentiometers. )
-Bill
|
| 1504.54 | digital board-->digital faders? | AQUA::KANOUN | | Tue Feb 21 1989 13:34 | 15 |
| Well, this topic caught my attention, and I had to jump in. I'm
curious how you plan on controlling signal levels in a digital board
without any controls.. sure, you would get rid of the analog pots,
but you need *something* for controls. They do make digital faders,
but they're even more expensive than analog. They look similar to
an analog slide fader, but they have digital in/out.
I think the main reason the good boards cost a lot is because of
the controls. Some of the better faders go for $50-$100 *each*!
Maybe you could control the whole thing by computer, with a simulated
board on the screen? Use the mouse to set things. Sounds slow.
Regards,
-Keith
|
| 1504.55 | 4:1 | MIZZOU::SHERMAN | quality first cause quality lasts | Tue Feb 21 1989 13:51 | 5 |
| As far as marketing goes, I would imagine the 4 to 1 ratio applies.
In other words, if you plan to retail your mixer for $1000, better
make sure it only costs you $250 to build (including labor).
Steve
|
| 1504.56 | Just kicking around ideas... | CTHULU::YERAZUNIS | it's.. it's DIP ! | Tue Feb 21 1989 15:00 | 6 |
| I was considering either one pot (the infamous data entry slider!)
or no pots and just a lot of raise/lower/select buttons.
This digital board _is_ just a lot of brainstorming, you know.
-Bill
|
| 1504.57 | That's a possibility... | AQUA::KANOUN | | Tue Feb 21 1989 16:43 | 14 |
| re: .56
OK, now I see what you were thinking of. Those are certainly
possibilities. 'Hope I didn't sound too negative.
I was reading some time ago in Mix magazine about various ideas
for the "mixing board of the future". One far out idea was to
completely replace the user interface part of the board (i.e. faders,
switches, etc) with a giant flat touch screen video display. I guess
a lot of people *like* to have controls they can grab onto though.
Interesting idea, but I don't think it would fit into the "Cheap"
part of this topic... 8^)
-Keith
|
| 1504.58 | 8x2 line mixer plans | SMURF::NEWHOUSE | | Wed Feb 22 1989 11:54 | 129 |
| Description:
Here are plans for an 8x2 line level mixer. There are no effect sends,
recieves or aux busses. Each input channel has gain and pan. I have
found it sounds good and costs somewhere between $30 and $40. It'll get me
by until I can justify buying one with all the functionality I'd like. Even
then I can still use it to do submixes. This mixer is very much the same as
the previous plans posted here.
Disclaimer:
I know very little about this. It was designed (a few times) by
a friend and I just put it together. Any thoughts, comments or
corrections are welcome - post 'em. I think you should use explicit comments
and/or updated plans - that way folks like me can follow the modifications.
You EE types will be bored by the following...
Plans:
10k
+--/\/\----+
CH 1 100k | | LEFT
IN +--/\/\----+ | |\ | OUT
+ | | 22k 10k | | \- |
o----+ | | +--/\/\--+------/\/\--+--+---|- \---+------o +
| 10k | |\ | | | | | / o -
- / <---/\/\---+--|-\- | | 10k / CH2-+ gnd-|+/+ |
o \ | \----+--+ pot \<-gnd ... | |/ |
| / 10k +--|+ / | (pan)/ CH8-+ gnd
| | POT | | /+ | | +--/\/\----+
gnd | (gain) | |/ +--/\/\--+ | 10k |
gnd gnd 22k | | |\ | RIGHT
| 10k | | \- | OUT
+------/\/\--+--+---|- \---+------o +
| | / o -
CH2-+ gnd-|+/+ |
... | |/ |
CH8-+ gnd
Battery Supply Wallbug Supply
+
+ gnd - | 5 k
| | | o----+-/\/\--+ |\
o o--+--o o + | | \+
+ - + - wallbug +---|+ \-------+---gnd
9v 9v - 5 k | | / |
o----+-/\/\--+ +-|-/- |
| | |/ |
- | |
+------------+
General:
The top diagram shows the circuit for 1 complete input channel and the
2 output channels. For each input channel duplicate channel 1 and join
them where it says CH2 ... CH8.
I used 10k pots for the pan. There is no center 'lockin' on the ones
I got from RS. Drawback - but I figured if I can't hear the difference
then there is no problem.
Use either the battery or wallbug power supply.
The wallbug power supply is the same one posted in another reply - I just put
it here to be complete. It works great. The battery supply does not last
very long and became *very* frustrating. Thanks for the plans for this!
I use a wallbug from an old appliance. It is 9vDC 500mA and seems to work
just fine.
Layout that I did was junk. First of all I didn't buy a metal box
my friend happened to have an old wooden wine box. It works great.
I won't tell you the one liner about the wine mixer we made...
I put all input channels on one little RS circuit board. I put the
power supply and output channels on another. The only shielded wire
I used goes to and from the ins and outs to the boards. All other wires
to pots and stuff was just plain old wire. Even with all this hacking
it is pretty quiet. It doesn't seem to drop out much either - at least
I couldn't hear the difference when switching my CD player
through it.
The op amps I used were DUAL BIFET OP AMP LF353N/TL082 from RS. There are
2 per chip. So for an 8 channel mixer you need:
1 - for output channels
4 - for input channels
1 - for wallbug power supply (using only 1 of the op amps on the chip)
----
6 - total chips
I used sockets for the chips. These particular chips are 8 pin and I got
16 pin sockets and doubled up - just to minimize the space and distance.
I had used other op amps before this and they sounded like junk. I think
you'll find alternatives for these chips - but I found them convenient.
I used all RCA ins and outs. RS had them in mountable groups of 1,2,4,8.
Real convenient - I used the 2 and the 8.
For all you non-Heathkit folks like myself...
You probably want to do this with a friend who is a heathkit type of person.
You'll want to sit down and make a parts list before heading out to the store.
It took me about 4 to 6 hours to make.
Part list for each input channel:
2 10 k pots
3 10 k res
1 100k res
2 22k res
1 op amp (1/2 LF353N/TL082)
1 rca female (in)
Part list for both output channels:
2 10 k res
2 op amps (1 LF353N/TL082)
2 rca female (outs)
Part list for Battery Power Supply:
2 9v batts
2 9v batt connectors
Part list for Wallbug Power Supply:
1 wall bug (I used one from an old answering machine)
2 5 K res (or there abouts)
1 op amp
1 female phone plug (power in)
General Parts:
box - remember you'll need room for 16 pots. maybe more if you ever add in
left and right gain, etc.
knobs - most expensive part of mixer
pc board(s) - I used 2 RS boards part # (276-170) (at the time anyway).
chip sockets - if you're gonna use 'em.
hammer - for the 2am test run and the left side of channel 4 is dead.
-Tim
|
| 1504.59 | A brave new world... | MUNCSS::BURKE | Jim Burke, @UFC | Wed Aug 23 1989 21:23 | 81 |
|
Well, I finally did it. I really went and done it. I bought all the
components for two 4x1 mixers, and went ahead. This is brave you see,
since I'm a pure software engineer, and am not too clever with a soldering
iron and other wiry thingies.
*** Result ***
* I have Elastoplasts on:-
left forefinger (hacksawing through potentiometer shafts),
left thumb (wire-stripping with unsuitable instrument)
right palm (soldering iron)
right elbow (...ditto...)
* Stretchy surgical bandage on:-
left wrist (not sure about this)
* Assorted punctures and lesions to both hands.
* Three whole days of self destruction. Bugger this - I'm sticking
to software.
Well, I don't give up that easy....
------------------------------------------------------------------------
On to the serious technical bit... The rest of this is meant for
Bill, but suggestions are welcome from any quarter.
Both mixers actually work; however the accompanying noise is
unacceptable.
Modifications from CQM Rev A
----------------------------
1. I'm using screened cable for most of it (see 1504.??). I connect
one end of a screen to ground, and the other I don't connect.
Q: does it matter which end you ground ?
2. For some inputs, I don't want the Effect. So, I've added two
switches to bypass the Effects on two of the four channels. This
is a single-core, two-way microswitch. Do I need anything else,
or is the seemingly-obvious two-way switch sufficient ?
3. I'm using TL-072 instead of TL-082.
________________________________________________________________________
Random questions:-
1. Is "CASE GROUND" = "BOX GROUND" = "GROUND" ? ie. is the box
casing grounded to the circuit (perfboard) 'ground' ?
The reason I ask, is that when I ground the perfboard to the
case, it makes godawful noise and sounds like something's about
to blow. What should I connect the 'perfboard ground' to ?
2. Is "EFF RETURN [shorting contact]" the TIP part (shorting) of the
socket, or the shield bit ? I assume the former.
________________________________________________________________________
Debug steps
-----------
[refer to Note 1504.23 (Debug script) for the numbering]
1) The Op-amps are powered OK. The meter shows the power supply is
OK, as in your suggested checkpoints.
2) There is no hum without power, but when I apply power, some hum
appears. I can't get rid of this hum. THIS IS THE PROBLEM.
3) There is a good signal, controlled OK with the pots. However,
the 10K trimpot does not seem to have much (if any) effect on
the gain (viz. volume: I'm cycling a TR505 pattern).
The trimpot only alters the surface noise a bit.
________________________________________________________________________
One last thing I noticed: I crank all pots down. I crank Chan#1
up to max (nothing is plugged in to Chans 2,3,& 4). The volume
increases as it should.
However; when I crank up the other chans (2,3,4), Chan#1 also
increases a little bit. Is this a clue ?
To summarise; why does this 10K trimpot not do anything ? I've
checked it with the meter, and it seems OK.
It certainly doesn't alter the volume of the input signal.
Anybody who can help me, going only on the above info, deserves a
medal.
help.
Jim (needing-anything-fixed?) Burke
|
| 1504.60 | | MIZZOU::SHERMAN | ECADSR::SHERMAN 235-8176, 223-3326 | Thu Aug 24 1989 01:00 | 10 |
| You didn't fry any of the chips, did you? Be sure to double-check
your wiring and look for breaks and cold solder joints. Cold solder
joints are grainy whereas good joints are nice and shiney. Also,
look for little bits of solder or wire between terminals.
By the way, all you need now (besides working boards) is a pair
of polyester pants with self-inflicted solder wounds and you'll
be a technician ... ;^)
Steve
|
| 1504.61 | Self-damage quotient goes down w/time. Avoid Polyester | KALLON::EIRIKUR | Hallgr�msson, ACA and CDA Prod. Mgr. | Thu Aug 24 1989 11:39 | 5 |
| Ouch, Steve! I used to own a few pair of polyester pants with solder
wounds!
Eirikur
|
| 1504.62 | Some more things to try | GUESS::YERAZUNIS | Like a shadow from the tomb... | Thu Aug 24 1989 12:32 | 89 |
| Ahemm....
PC ground = box ground = 1/4" jack ground. This is why you MUST
MUST MUST use a wall bug power adapter with isolation and a wall
bug socket that is isolated on both sides. If you connect either
side of bug power to the box ground you will fry the active ground
circuit in a few seconds (unless the chip has auto overcurrent
protection, in which case it'll just shut down on you and you won't
have any active ground.)
Or- try a 9 volt alkaline battery. It'll only last a few minutes
but it will let you verify that the problem is in the CQM power
supply section rather than somewhere else.
Do both CQM's misbehave the same way? If so, you may have an
interpretation error in the schematics, rather than a bad part.
Try having a friend who's good at electronics look over your wiring
and the schematics.
Other possibilities:
1) You fried a chip... you poor soul. I hope you used sockets...
Try swapping chips around (wear an antistatic kit).
2) You miswired something, possibly left something unconnected.
Not using a metallic case (socket shield ground is via the metallic
case) will cause this "miswire" effect.
3) Your wall bug is very badly filtered. You need an oscilloscope to
check it for sure, but you can "slum it" with the following circuit
(though a 9V alkaline battery is cheaper if all you need to do is
test). All the diodes in this circuit are 1 amp, 50 volt. (the
resistor and third diode are there to provide a "soft-start" on
power-up)
wall bug +
----------|>|----.---------.-to active ground +
| |
/ |
25 ohm \ --|>|--|
/ |
|---
+ | 20,000 microfarad (or bigger)
======== 15 volt
| electrolytic
|
wall bug - | to active ground -
---------|<|-----.----------
4) You have a ground loop from the R8. Unplug all of the 1/4" plugs
(sig in/out, eff send/ret, etc.) Leave the power plug in.
Plug in a headphone into the sig out. Do you get a hum? If so, it's
an internal problem; see #1 or #2 or #3. If the hum is gone,
then your problem is because something you plug in (or a pair of
things you plug in) has a voltage leak.
The CQM, as designed, ties all the signal grounds together right
at the CQM. If two things in your setup disagree as to what "ground"
really is, you'll get a hum.
This isn't the CQM's fault, since the hum voltage can be many volts
(up to 180 volts peak-to-peak) and there really is no good way to
get rid of it.
-----
The 10K internal diddle-pot really does do something! It sets the
internal gain of the first (pre-effect) stage. On my CQM, if I
turn that pot all the way one way, I get almost no signal, and as
I turn it the other way, the gain goes WAY up and I start getting
hiss.
If in your system the pot doesn't do anything, then the first stage op-amp
is probably fried. The CQM's design is such that if that chip goes,
you'll still get some signal through, but not much, and there will be a
lot of crud mixed in.
-----
If this doesn't work, mail me a round-trip ticket and I'll fly out
there to Munich and personally fix that CQM. :-)
-Bill Yerazunis
|
| 1504.63 | It works !!! (sortof !) | MUNCSS::BURKE | Jim Burke, @UFC | Fri Aug 25 1989 09:11 | 31 |
| well....
1. MIX OUT connected to headphones:-
Both mixers work fine - no noise, and the trimpot varies the gain
perfectly !! I had the case ground connected to the perfboard
ground, as per your schematic.
It works *exactly* as you said - it is dead quiet, and there's
plenty of volume in the cans.
2. MIX OUT connected to amp (an Acoustic 2x12 combo guitar amp):-
I cannot connect the case ground to perfboard ground - the noise
ramps, as though something's about to blow. So, I have to leave
this disconnected.
The trimpot has no effect.
The signal is weak.
I also tried my Hifi amp - same problems [both amps are powered
by a 3-pin (grounded) plug. It must be something to do with the
power supply and/or grounding.
I tried a 9v alkaline battery. The hum disappeared. However, the
signal was very faint.
Incidentally, I'm building this at home here in Edinburgh - the UK
power supply is 220/240v, 50 Hz.
Any further ideas ?
Many thanks for the help so far,
Jim Burke
|
| 1504.64 | | MIZZOU::SHERMAN | ECADSR::SHERMAN 235-8176, 223-3326 | Fri Aug 25 1989 11:44 | 4 |
| Hmmm. Sounds like you may be dealing with some impedance mismatch
between your mixer output and your amps?
Steve
|
| 1504.65 | Next step... | GUESS::YERAZUNIS | Try to never run out of altitude, airspeed, and luck all at the | Fri Aug 25 1989 13:48 | 51 |
| Well, it's encouraging to hear at least that the CQMs are not fried.
(to be honest, I thought they'd be radioactive slag after what you've
said. :-) )
It's not an impedance mismatch problem; the CQM output drive buffer
is not sensitive to impedance mismatch (being a unity-gain voltage
follower has it's advantages).
You MUST connect jack-shield ground to perfboard ground; else the
audio signal will just float and the diddle-pot won't do diddly.
Just solder the jack grounds to the perfboard ground to the case
ground and leave it that way; your problem lies elsewhere than
those connections.
-----
You definitely have a ground-loop problem. Your Acoustic 2x12 is
the one "bit" in the system that changed, and then all hell breaks
loose. If I had to point a finger at this point, it would
point right at the A/2x12.
Try the following:
1) Check your CQM-->A/2x12 cable. Use a good ohmmeter; the resistance
for center-to-center should be less than an ohm, likewise for the
shield. A broken or intermittent shield could be causing this
problem.... or maybe a broken/intermittent MIX OUT jack??? Swap
this cable with another cable just for luck.
2) Disconnect everything except power to the CQM and the A/2x12.
Reconnect the perfboard ground to the case ground. Is the hum still there?
If so, then you have a double problem: your wall bug is not
isolated and your A/2x12 has a ground leak. Try the alkaline
battery routine. Does it _still_ hum (CQM on battery, A/2x12
on AC, no other connections?
3) If not, then stay on battery power, and jack in other inputs one at a
time (that is, plug "A" in. Then unplug "A" and plug in "B". It
doesn't matter which of the 1-4 input jacks you use). Whenevery an
instrument causes the hum to come up, then that instrument as well
as the A/2x12 are both ground-looping. Get them fixed.
-----
Jim Miskatis (sorry about the spelling) and I were kicking around ideas
for an optoisolator box at the last LERDS-BIM. Maybe we should press
on with that idea?
-Bill
|
| 1504.66 | Opto-box: yes | KALLON::EIRIKUR | Hallgr�msson, ACA and CDA Prod. Mgr. | Fri Aug 25 1989 14:16 | 14 |
| Opto-isolator box: Sounds like a really great thing to have around (at least
the plans).
Can you get good (low) distortion results with something like that?
A ~$50.00 (US) optoisolator box that ran forever off of a 9V battery, that
wasn't any bigger than the battery plus the connectors, would be a heck of a
product in our field. I'd invest.... Call it, uh, Pocket-Isolator....
Eirikur
|
| 1504.67 | Miscellaneous Mispellations | MRSVAX::MISKINIS | | Fri Aug 25 1989 17:20 | 1 |
| That's spelled John Miskinis!
|
| 1504.68 | <BLUSH ON> | MUNCSS::BURKE | Jim Burke, @UFC | Mon Aug 28 1989 09:36 | 35 |
| ahem..... (do I feel silly ?)..... I feel that I must now record:-
"Of sound mind and body do I, Jim Burke (S/W engineer), hereby
acknowledge that there is more to hardware than manipulating
bits of wire and [un]screwing things."
[eg. unwrapping a Band-aid with one hand]
--------------------------------------------------------------------
Now then. After I discovered that the CQMs worked fine with
headphones and also my Portastudio, I just went ahead and put the
lids on, and finished it off. Installed everything (SGM's), wired
them all up, checked it all out - spot on.
As I write this (from home), I have just connected a CQM to my 2x12
in order to experience the terrible noise/hum again. Result: dead
quiet, works fine (!??@&). I'm not saying no more.
I discovered something that may have been confusing things.
As I stated earlier, I introduced a mod to the CQM. I decided that I
didn't want effects on all four channels, and so I wired two channels
with switches which bypass the effects. I did this with a single-pole
two-way microswitch, which switches the channel input from the 'A
line' to the EFF RET line. Works fine.
HOWEVER, the little trimpot doesn't control the 'un-EFF' signal. A
bit obvious (even for me).
Many thanks Bill - you have been invaluable. The longetivity alone
of this note stands testament to your efforts.
Jim Burke
<BLUSH OFF>
------------------------------
I like the idea of the opto-isolator box. Anyone want one built for
them
|
| 1504.69 | Those switches sound weird... | GUESS::YERAZUNIS | Where there's life, there's threat. | Tue Aug 29 1989 19:45 | 38 |
|
Sounds like you've had your first experience with a mix-board gremlin.
They're rare little critters with sharp metallic teeth that can
shred your fingertips, glowing red tongues that burn your hands,
and forked tails that plug right into the house power line and feed
AC hum right into the speaker leads.
The only way to appease them is one of the following:
1) Drop-kick the possessed mix board across the stage in front
of a sellout crowd, then beat on the board with a ball peen hammer
while screaming "I hate you I hate you I hate you"!
2) Drip a stream of fresh human blood into the FX return.
3) Spend at least $800 at a music store on a spare mix board,
and then carry it around, still in the pink plastic wrap,
forever.
-----------------------------------------------------------------------
Good to hear that the CQMs work fine for you, too. I'm not sure
that the FX bypass the way you have it wired will do what you want
it to do. It certainly wouldn't do what I would want it to do.
When you throw the mix bypass switch, you're "mixing" the FX return
signal with incoming signal from the SGUs without any mix resistors.
Try adding a 10K resistor from FX return to the junction of the
chan1/chan2/buffer line, and another 10K resistor in the line from
the FX defeat switch to that junction.
When you throw that switch, the mix levels are gonna get screwey,
be prepared.
Good luck,
-Bill
|
| 1504.70 | Bandaid Rev 2... | MUNCSS::BURKE | Jim Burke, @UFC | Tue Aug 29 1989 20:28 | 11 |
| 1x10k+.....
...sounds like I've got to de-wire the b***s now and upgrade to-----
CQM; Version 'Bandaid'; Rev 2....
await further announcement...
PS Bill: If this causes it NOT to work, then not only will I revoke
my declaration, but I will also fund my own trip to your locality,
and...
|
| 1504.71 | Let me rephrase that | GUESS::YERAZUNIS | There's no force like brute force! | Thu Aug 31 1989 15:58 | 28 |
|
I didn't say it would make it work, I only said that it would
probably make it work better.
Then again, I could be worng. It's been a while since I dug into
a CQM (they never break) so who knows what will happen when you
put those wiry thingies in there. :-)
Of course, if you are completely and unabashedly happy with the
way it works right now, then leave it be and don't open it up.
You realize, of course, that you are likely to break it by taking
it apart now that it's working.
Crah's Laws of Things:
1) Things can be in two states: working and broke;
2) Taking things apart can cause them to change state.
3) There are more ways to get into the broke state
than there are ways to get into the working state.
-Bill
|
| 1504.72 | I build one 6x2 | SNEEZY::DICKENS | What are you pretending not to know ? | Wed Jan 17 1990 12:46 | 38 |
| Just to let you know, I've finally finished a CQM.
I built it as a 6x2 with no level controls. 2 inputs are mono, mixed to center,
2 are panned hard left, and 2 to the right. So I use it for 2 stereo sources,
and 2 mono sources.
It's actually built as a 12x2, but I need to find a bigger box to mount all
the jacks in.
I found that the most important thing I had to learn was how to debug the thing.
My debug procedure went something like this:
1. Before the chips are plugged in, check the resistance from the case ground
to every point on the board that should be grounded. An Ohmmeter with one of
those "continuity beepers" makes it easy. This will find any open ground.
2. Then with the continuity checker (or ohmmeter) check the resistance between
every adjacent pair of pins on the board that you've used. This will find all
those microscopic solder bridges. I found that a lighted magnifying glass
helps, after you've found two pins that shouldn't be connected, but are.
3. Then I put the chips in, apply power, and check the voltage at the + and -
supply pin of every chip.
Do this before you try to put sounds through it, and you'll be a lot happier.
Also, I found a book that will be of great interest to most CQM hackers:
Sams' Understanding IC Operational Amplifiiers
I found it at U-doit, but Active has it too. Any place that carries the Sams
books can order it for you.
Thanks again to Bill Y. and the rest of you,
-Jeff
|
| 1504.73 | Other than that I think I understand | LOOKUP::ADSUPPORT | | Wed Jan 17 1990 15:51 | 14 |
| One question (from not-yet-but-soon-to-be-EE person):
What is |\ | from the base note and .58?
| \|
---|+ \ I assume it's the TLO82, but what's the
| \ pinout, etc.? I don't want to wait until
| / Cicuit Theory 420 to build this mixer.
---|- /
| /|
|/ |
Thank you,
--mikie--
|
| 1504.74 | Yet more Q's | LOOKUP::ADSUPPORT | | Thu Jan 18 1990 08:24 | 15 |
| I re-read all the replies and came across something interesting; in
one of them someone mentioned that all the specs included here are line
level. And if you needed to use mic-level, you had to add "unity gain"
to each of the mic level inputs. Does anyone have a schematic for this?
Also, does anyone know the schematic for a simple one-pot tone
control?
I'm planning on building that 8x2 and adding switchable level
inputs and adding tone control per channel. Maybe even a hard on/off
effect loop.
I'd appreciate any feedback.
--mikie--
|
| 1504.75 | | DOPEY::DICKENS | What are you pretending not to know ? | Thu Jan 18 1990 12:36 | 17 |
| re .73 A while back I got crazy and bought the TI Linear data book, which tells
you all about every op-amp chip you've ever heard of. I believe that Rat-shack
also sells a somewhat boiled-down semiconductor data book. It might have what
you need, too.
re .74 I think what you want is a unity-gain input buffer for line inputs,
and a higher-gain input buffer for mic. (lower level) inputs.
The book I mentioned in .72, Sams' Understanding IC Op-amp circuits, has a
whole chapter on designing op-amp active filters. It even gives you an example
BASIC program to calculate the resistor and capacitor values. You might be
able to boil this down to a useful 1-pot tone control; I haven't played with
filters yet. My next project is to try the pan circuit from .58, and maybe
add some low-level inputs for guitars, mics and the like.
-Jeff
|
| 1504.76 | More CQM answers | GUESS::YERAZUNIS | Knowledge is a deadly friend when no one sets the rules. | Wed Jan 24 1990 18:24 | 23 |
|
Re: tone control: Check the "Active Filter Cookbook" for circuits.
Re: mic inputs: Microphones can put out two kinds of output- lo-Z
and Hi-Z. Use a rat-shack plug adapter to get to Hi-Z, then
put an op-amp stage with a gain of about 10 (may need more) in the line.
The reason to put this gain-10 stage in is because microphones
and guitars put out only about .01 to .1 volts - not much. Synths
put out around 1 volt. The gain of 10 brings them up to par.
The reason for the unity-gain buffer is that if you're building a
something X 2 or more CQM, you can get crosstalk effects between
different inputs. The buffer prevents that from happening (almost
completely- super-duper ultra-expensive boards use two buffers-
one at the input, and one more for each bus being mixed to.)
-----
It's good to hear people actually using building CQMs. I hadn't
expected anyone to build them 8x2 with extra channels, though. :-)
-Bill
|
| 1504.77 | Watch for impedance mismatch | TOWNS::MUSUMECI | | Wed Jan 24 1990 22:03 | 35 |
|
If you are going to provide inputs for mikes and guitars then you should use
at least 100k resisters on the - inputs of the op amp for those inputs. Else
you run a good chance of loading the sources( mike or guitar). Using .58
as an example...
Plans:
10k
+--/\/\----+
CH 1 1M | | LEFT
IN +--/\/\----+ | |\ | OUT
+ | | 22k 10k | | \- |
o----+ | | +--/\/\--+------/\/\--+--+---|- \---+------o +
| 100k | |\ | | | | | / o -
- / <---/\/\---+--|-\- | | 10k / CH2-+ gnd-|+/+ |
o \ | \----+--+ pot \<-gnd ... | |/ |
| / 100k +--|+ / | (pan)/ CH8-+ gnd
| | POT | | /+ | | +--/\/\----+
gnd | (gain) | |/ +--/\/\--+ | 10k |
gnd gnd 22k | | |\ | RIGHT
| 10k | | \- | OUT
^ +------/\/\--+--+---|- \---+------o +
| | | / o -
| CH2-+ gnd-|+/+ |
| ... | |/ |
| CH8-+ gnd
A noninverting amp may be better here.
Chris
|
| 1504.78 | filter fun | DOPEY::DICKENS | What are you pretending not to know ? | Thu Jan 25 1990 13:27 | 61 |
| I don't know how many of you read "Electronic Musician", but the CQM sans
input buffers is duplicated almost exactly in an article on homebrew line
mixers in this month's issue. Also covered is a passive mixer.
----
Another article in the same issue describes a homebrew 12ax7 tube preamp
(for the hacker who understands the dangers of +300V).
----
I've been thinking about building a box that would provide some boost at
a given frequency, say 2000 hz, to "color" a guitar input.
I picked the following circuit for a multiloop feedback bandpass filter from
the Understanding IC Op Amps book:
+---------------+
| |
+--|(-+-/\/\/-+ |
| c1 r3 | |\ |
| | | \ |
o---/\/\/-+--|(---------+-|- \__+__________o
r1 | c2 +--------|+ /
+-/\/\/-+ | | /
| r2 | |/
gnd gnd
It looks a lot clearer in the book, but it's hard to make vertical resisters
here.
Anyway I don't have the associated math on hand for this, but I'll try to
bring it in tomorrow. But basically I have this formula where you give
the center frequency, gain and Q, and the value of the C1 and C2 you are using.
(C1=C2). Then it gives you the values for R1, R2 and R3.
It occurred to me that I could make R1-3 ten-turn trimmers that sit in a
dip socket. Then I could program the filter by doing the math, setting
the trimmers and plugging them in. Cool, eh ?
Well, I have some questions.
What unit is Q in ? Is Q a slope ? How sharp is a Q of 10 versus 1 or .1 ?
Also, since this filter circuit uses the inverting input, will the output be
the inverted, as well as filtered ?
This is relevant since I thought I'd like to boost up one band, and then use
a summing amp to mix the filtered, boosted signal back in with the unfiltered
signal. I would expect this to yield a more useful equalizer. Any problem
forseen with this ?
If the filter did produce an inverted output, and I did mix it with the
unfiltered input, wouldn't phase cancellation cause the boosted frequencies to
be actually be cut rather than boosted in the final output ? Would this not
make a fairly decent band-reject filter ?
Thanks,
-Jeff
|
| 1504.79 | | SMURF::NEWHOUSE | | Fri Jan 26 1990 17:17 | 5 |
| re: .73 (what is op-amp pin out...)
Sorry so long, you must have figured it by now... but in case
Just go take a look at the diagram that tells them pins on the
chip - my Rshack package had it on the back. It matches the
picture.
|
| 1504.80 | | SNEEZY::DICKENS | What are you pretending not to know ? | Tue Jan 30 1990 13:21 | 41 |
| Last night's installment of CQM hacking was a learning experience.
I was playing around with different preamp gains to try to get optimum
performance from:
1) A Shure SM58 with transformer-in-cord,
2) A Gibson electric guitar with "zebra" high-output pickups
3) A Seagull acoustic guitar with "The Baggs pickup", which is a piezo unit, I
think.
Several maddening hours were spent trying to discover the source of the loud
hummmmmm I was getting with any gain more than 1. The simple final cure was
to cut off the 5 or 6 inches of wire that was connecting the phone jack I was
using as a test input to the solderless breadboard. I soldered about an inch
of tinned bus wire to the two contacts on the phone jack and stuck it right
down on the breadboard, so that there was only about 1/2 inch between the jack
and the heavy aluminum ground plane of the breadboard. It also helped to ground
the breadboard ground plane (it wasn't).
The preliminary results are as follows:
1) SM58 = gain of 5 to 10 depending on spl present was ok.
2) Gibson = 1 to 2 1/2. Any more would pin the meters. Note that even with
gain near 1, real loud playing would cause distortion. It seemed that some
kind of compression was in order, unless you exercise extreme restraint
in playing.
3) Seagull/Baggs = a gain of 10 sounded great. There was noticably more
high-end present in the signal than I've heard using any other preamp, including
my dbx 166 compressor's high-Z input. I might try even more gain than 10 !
Also, a minimum input resistance of 100K seemed to be a must as a previous
reply indicated.
Anyone have any thoughts on my filter ideas, the metrics of Q, etc ?
-Jeff
|
| 1504.81 | PLAY AROUND A LITTLE | TOWNS::MUSUMECI | | Tue Jan 30 1990 21:45 | 78 |
|
RE: .80
>>> Several maddening hours were spent trying to discover the source of the
loud hummmmmm I was getting with any gain more than 1. The simple final cure
was to cut off the 5 or 6 inches of wire that was connecting the phone jack I
was using as a test input to the solderless breadboard. I soldered about an
inch of tinned bus wire to the two contacts on the phone jack and stuck it right
down on the breadboard, so that there was only about 1/2 inch between the jack
and the heavy aluminum ground plane of the breadboard. It also helped to ground
the breadboard ground plane (it wasn't). <<< ------------------------
----------------------------------------
I think the lack of grounding the ground plane was the main reason for the loud
hum. You should keep all leads less then 3 inches as these tend to act like
antennas and pick up all kinds RF signals. Usually you don't get this Antenna
problem with low gains ( < 10).
>> The preliminary results are as follows:
>>
>> 1) SM58 = gain of 5 to 10 depending on spl present was ok.
>>
This is great for a mic output. Most mikes need a gain between 10 and 100 to
get up to a decent level for a power amp. Are you sure there isn't gain
somewhere else in the circuit?
>> 2) Gibson = 1 to 2 1/2. Any more would pin the meters. Note that even with
>> gain near 1, real loud playing would cause distortion. It seemed that some
>> kind of compression was in order, unless you exercise extreme restraint
>> in playing.
Again the gain sounds low to me. But the less gain you need the less noise.
>> Anyone have any thoughts on my filter ideas, the metrics of Q, etc ?
I'll give it a shot, but I don't have the expierence you need to really
understand active filters. You really ought to get a active filter cook
book and expierement.
A bandpass filter has a bandwidth associated with it. This is the difference
in HZ between the upper and lower points where the bandpass filter response
falls 3 DB below it's peak value. The center frequency of the filter is the
goemetric mean of the upper and lower 3 db cutoff frequencies. It is NOT as
some will tell you, half the difference between the upper and lower 3 db cutoff.
Q is the ratio of the bandwidth to it's center frequency. A bandpass filter
whose center frequency is 1000hz and whose bandwidth is 10hz has a Q of 100.
If it's bandwidth is 100hz then it has a Q of 10.
The inverse of Q is called damping. Low damping is equivalent to high Q.
Bandpass filters using a single op amp are usually limited to Q's of 10 or
less. Multiple op amps can be used to construct bandpass filters with Q's
up to 400.
Alot of budget mixers have what is called shelving equalizers. Most
Hi Fi tone controls are high and low shelving equalizers. These equalizers
got their name because of the shape of their frequency response.
+20 ------------\ /---------------
\ /
\__________/
0 DB / \
/ \
-20 ____________/ \_______________
Low frequency shelving High frequency shelving
I have some of these circuits that I played with that turned out pretty good.
I'll try to find these and post them if there is interest.
Chris
|
| 1504.82 | | DOPEY::DICKENS | What are you pretending not to know ? | Thu Feb 01 1990 12:46 | 32 |
| re hum: Even after I grounded the ground plane, I found that by moving the
lead to the input jack away from the breadboard, the hum would come roaring
in. Just folding it back over the breadboard killed the hum completely.
It was an eduacation.
re gain: Yes, as I figured out after I wrote -.2, the mixer summing amp has gain
too. I don't know how much offhand, but probably between 1 and 5. So does gain
add linearly ?
I was most impressed by how great "the Baggs pickup" sounded with the homebrew
preamp. I may package that up in the smallest metal box I can find, and keep
it in my guitar case. Hmm, maybe if I used TL062s I could get away with a
9v battery power supply..
Thanks for the info on filters. I think my next project is a four op-amp
state variable filter that has independantly adjustable center frequency and
selectivity.
I'm also thinking about re-hacking my old guitar pedal box (which currently
contains an MXR distortion+, MXR dyna-comp, Boss chorus and Boss analog delay,
and adding some buffer amps before, between, and after the pedal effects.
Here's a new question: Has anyone got a schematic for a solid state audio
switch ? This would have one input and one output and would connect or
disconnect the output when a momentary (debounced) switch was pressed.
Just like the Boss pedal effects have. I want to be able to switch my eq
in and out reliably with a stomp-box type switch. It would also be good
to build a noiseless A/B box, or even a multi effects-loop controller.
(uh-oh, look out Mr. Bradshaw)
-Jeff
|
| 1504.83 | More answers | GUESS::YERAZUNIS | Dr. Frankenstein, I presume? | Thu Feb 01 1990 17:50 | 17 |
|
Two items (they didn't matter for my application; they might for
yours):
1) Tie the hot side of each input to ground with a 100K resistor.
This can cut down on the noise picked up significantlyf.
2) You _can_ make an audio switch out of one of those "CMOS quad
bilateral switch" chips from Radio Shack. Actually, you make
four of them. When turned off, the on-chip switches have
a few megohms resistance- when on, it drops to about 80 ohms.
This is an excellent way to build electronically controlled
switches (you can use a capacitor and resistor pair to make
the turn-on slow )
-Bill
|
| 1504.84 | | DOPEY::DICKENS | What are you pretending not to know ? | Fri Feb 02 1990 12:02 | 3 |
| I found that the 100K resistor tieing the input to ground was necessary
to keep unused inputs quiet. Also, it seems to kill the pop you'd otherwise
hear when you plug in or unplug an input.
|
| 1504.85 | :-) | GUESS::YERAZUNIS | Just a puppet who can see the strings. | Fri Feb 02 1990 12:45 | 9 |
| > I found that the 100K resistor tieing the input to ground was necessary
> to keep unused inputs quiet.
^^^^^^
UNUSED inputs? A MIDIholic with UNUSED inputs? Sir, I am frankly
shocked at your dereliction of _duty_. :-)
-Bill
|
| 1504.86 | | SNEEZY::DICKENS | What are you pretending not to know ? | Mon Feb 05 1990 12:56 | 4 |
| Yes, unused inputs. It's a shame, isn't it ?
Actually I'm configuring the mixer to handle visits to a friend's studio
that contains more SGUs than I have, but not enough quiet mixer channels..
|
| 1504.87 | 9 non-inverting amps would work??? | MAMTS2::RUYOUNG | | Fri Jul 20 1990 13:16 | 18 |
| RE: .58 8x2 mixer
Is there a problem with making it a 9x2 mixer? I saw a reply that
said, make sure you have an even number of inverting amps. Those are
the input amps, right? Couldn't I just sum 9 leads in stead of eight,
to make it 9x2? Or, if the inverting amps are a problem, could I make
them all non-inverting?
I really don't have a clue what I'm saying in terms of electronics;
I just figured since in doesn't cost that much to make 1 more channel,
why not?
I did think about 10x2 but that puts it into another price bracket
(everything is sold by $3.50/10 or whatever, and the output needs one.
so 9x2 is all I can afford.)
Can anyone help me out?
Mike
|
| 1504.88 | This Mixer Sounds Odd! | DRUMS::FEHSKENS | len, EMA, LKG2-2/W10, DTN 226-7556 | Fri Jul 20 1990 14:52 | 8 |
| I think the comment about an even number of inverting amps was with
respect to a chain of amps, so that the output has the same phase as the
input. I believe it's OK for there to be an odd number of inputs,
but I'se just a software architecture guy, so you'd best get a real
confirmation from a hardware weenie.
len.
|
| 1504.89 | Right | OTOO01::ELLACOTT | non_teenage_mutant_ninja_bassist | Fri Jul 20 1990 15:23 | 7 |
| RE.-1
That's right. Mixers should always have an even number of invertion
stages(amps) to keep the output in phase with the inputs, this should
also include the FX sends and returns. The actual number of inputs
is irrelavent except when designing the summing amp circuit.
FJE
|
| 1504.90 | 9x4. How silly can I make it? | GRANPA::RUYOUNG | | Fri Jul 20 1990 15:43 | 7 |
| Ok. I pretty much figured that, but as I said, I have no clue really.
It probably would be pretty easy to make a switchable "main/aux" bus
then, too, for multiple outs. Hmmm...
len, howse LKG these daze? Seems, like only two years ago...
Mike
|
| 1504.91 | have portasol, will travel | CTHULU::YERAZUNIS | My program can beat up your program. | Mon Jul 30 1990 16:21 | 14 |
|
Sure, you can have as many inputs as you want (assuming you
put the input buffer circuit in... otherwise you'll load down the
inputs too much).
The only "rule" you don't want to violate is that if you trace a signal
path through the system block diagram from any input to any output, you
should go through an even number of inverting amplifiers (non-inverting
buffer amplifiers don't count).
So, no inverting amplifiers is OK, two is OK, four is OK... But one is
not good, and neither is three, and neither is five...
-Bill (and you should see my soldering chops! :-) )
|
| 1504.92 | I've started building one! | KOAL::LAURENT | Hal Laurent, Loc: FOR, DTN: 378-6742 | Mon Aug 13 1990 12:46 | 61 |
| Well, I've finally done it! I've actually soldered my first connection
in anger! Since I'm very inexperienced in this electronic stuff, I first
built and tested the input buffer that I plan to use (see below). I know
it looks a bit strange, but I plan to support both balanced-line and
unbalanced inputs. Since I don't yet have any balanced-line input devices,
I built the degenerate case first.
I fried my first IC within seconds of applying power. Apparently the TLO74
doesn't take kindly to having the power connected backwards (sheepish grin).
I also managed somehow to cut my finger with the wire strippers. Still, not
as bad as Jim Burke's first experience :-). And, amazingly enough, it worked!
At least it did after one modification.
I first built the circuit below, except without the 500K resistor from input
to ground. Worked fine with a high-impedance mike as input, but when I
connected the Barcus-Berry pickup from my acoustic guitar I just got a lot
of hum and no signal. After trolling through some circuits in some books,
I added the 500K resistor from input to ground. Voila, both input devices
work great! However, since I'm rather lacking in basic electronics knowledge,
I don't understand why the extra resistor helped. To further my electronics
knowledge, could someone explain to me why the 500K resistor helped, and why
I only needed it for some input sources?
|
\
/
500K \<-------------
/ |
\ |
/ |\ |
| | \ |
10K | | \ |
--/\/\/\/\-*---| - \ |
| | \ |
----- | \ |
--- | \---*------o Out
- | /
| /
| /
In o-------*-------| + /
| | /
\ | /
/ |/
500K \
/
\
/
|
-----
---
-
For anyone that's curious, my goal is to build an 8-in 2-out stereo mixer.
I definitely want some sort of EQ on each input channel, but I'm not yet
sure what kind would be best. I plan on breadboarding some experiments with
parametric equalization as well as with high and low shelving equalizers with
a resonant equalizer for mid-range.
Hal Laurent
|
| 1504.93 | | DECWIN::FISHER | Locutus: Fact or Fraud? | Mon Aug 13 1990 13:28 | 9 |
| Just a guess:
If the op-amp is VERY high impedance, it takes virtually no current to produce
a significant voltage on the input. Thus, you get lots of hum and stuff just
from induction of power-line stuff into the wiring. If you put on the 500k
resistor, that is not enough to load down any kind of significant voltage
source, but it is enough to load down tiny sources like induction.
Burns
|
| 1504.94 | Supposed to work that way | CTHULU::YERAZUNIS | I only designed your eyes. You must talk to Tyrell. He designed | Mon Aug 13 1990 14:33 | 15 |
|
.93 is close; the problem is that HiZ devices like crystal pickups
and BiFET op-amp inputs have input resistances approaching 10E9 ohms;
i.e. close to the resistance of a pointed wire aimed at the ground.
So, any static/induced charge on the input wire isn't bled off, but
just sits there, forcing the op-amp into saturation. By putting the
500K resistor (often called a "bleeder" resistor because it bleeds off
the excess charge on a capacitor somewhere) you get rid of the saturation.
[and I'm glad to hear you're having good luck! Who says
microprocessors are the only way to do something musically
interesting? ]
-Bill
|
| 1504.95 | TLO-84 Quad Op Amp good S/N? | VFOVAX::BELL | | Tue Aug 14 1990 10:14 | 6 |
| Hey, whoever suggested using the TLO-72 for better S/N, is the TLO-84
the quad op amp of the same specs? It would make things a lot cheaper,
as in my projects I have to use tons of them.
Thanks,
Mike
|
| 1504.96 | Almost but not quite | OTOA01::ELLACOTT | non_teenage_mutant_ninja_bassist | Tue Aug 14 1990 10:41 | 9 |
| No, the TL074 is the quad version of that op amp. The TL07x series
is the same op amp as the TL08x, but has been screened for low
noise characteristics. This is why there is such a difference in
the price as few of the chips pass the tests. The places where it
is important to use the low noise version are where there is low
signal levels or high gain. In buffering applications with line
level signals there is little difference, in fact regular (as opposed
to bifet) low noise op amps will do quite nicely.
FJE
|
| 1504.97 | Miser needs mixer | STAR::ROBINSON | | Tue Nov 20 1990 14:53 | 27 |
| I suppose this is the place to put real cheap mixer questions. ;-)
I have had my eye on the cheapest 4 into 2 stereo mixer sold by
Radio Shack ($34). The other day someone offered me what is
probably an older version of that mixer called a
"Transistorized Stereo Mike Mixer". It comes in a 2"x3"x7" or
so metal box with four phone inputs and two RCA outs.
I was excited about saving $34 ;^) but was dissapointed with
the distorted sound it produced when I put the line outs
from my VFX synth into it.
Is the problem the mike inputs? The new Radio shack
mixer provides mike and line iputs I noticed.
If that is the problem, would there be a real cheap/worth it
way to change the mike jacks to line jacks?
If none of the above would this box be suitable for converting
to the mixer described in the previous 9x replies? I am a complete
novice but I do have a soldering iron and a radio shack meter.
I have even soldered a few things but I never knew what I was doing.
Or, does anyone have a four or six into two stereo version of the
build-it-yourself mixer they would sell for about $34?
Thanks in advance,
|
| 1504.98 | Go For It | AQUA::ROST | Drink beer: Live 6 times longer | Tue Nov 20 1990 15:37 | 12 |
|
Re: .97
The problem is that the mixer expects microphone level signals (about
100 mv max) and you're feeding it line level (maybe a volt or two).
The "line" input on the newer RS mixer is probably just uses a resistor
to lower the level.
Yes, you could gut it and built the super-cheap mixer using the case,
pots and jacks, as described.
Brian
|
| 1504.99 | ... | FULCRM::PICKETT | David - Brahms Berman Requiem? | Wed Nov 21 1990 09:17 | 4 |
| I liked the suggestion, a while back, to get an old RS mixer, rip out
the guts, and save the box, sliders, and connectors!
dp
|
| 1504.100 | What you mean resistor Kimo Sabi? | STAR::ROBINSON | | Wed Nov 21 1990 09:39 | 16 |
| >> The "line" input on the newer RS mixer probably just uses a
>>resistor to lower the level.
Just in case I am very ignorant about these things ;-) , what value
resistor would lower the level of the input from expected 100mv to one
or two volts line level. I noticed that there are resistors (unknown
value) connected between the phone plug tip leads and the pots. Is
this where I could add magic new resistors? Actually, I am
assuming the brown things with colored lines on them are resistors.
How am I doing? %-)
I will probably gut the thing eventually, but I have a few notes to
digest before I get there.
Thanks again,
Dave
|
| 1504.101 | A real engineer should check me here.... | RANGER::EIRIKUR | Eir�kur Hallgr�msson | Wed Nov 21 1990 11:55 | 22 |
| Another alternative, is to use one of the attenuating adaptors, or attenuating
cables that Radio Shack sells.
Or hack the mixer as follows:
This is a straightforward voltage divider. The ratio between the input
resistor (usually large) and the resistor to ground is the attenuation
(plus or minus some fudge). So, for a 10K input and a 1K ground resistor,
you get on the order of a factor of 10 attenuation.
input -/\/\/\/\----O-----> mixer circuitry
|
/
\
/
\
/
|
-----
---
-
|
| 1504.102 | Correct... jack outside = ground | CTHULU::YERAZUNIS | Parallel Decomposition A House Specialty | Sun Nov 25 1990 21:47 | 5 |
| Yes, Eirkiur has it right. 10K ohms between input jack and mixer
circuit, and 1K from the junction to ground (or outside of the jack)
-Bill
|
| 1504.103 | Cheap mixers continued... | TLE::ASHFORTH | | Mon Nov 26 1990 12:04 | 13 |
| I, too, sing and play on a shoestring (yes, I someimtes lose my balance...).
I have what is probably the "next-up" Radio Shack mixer, with inputs for phono
(ceramic/magnetic) times 2, tape, aux, and three mikes (with stereo
positioning). Currently I have my Amiga and a Kawai K1-II as line-level inputs,
and a Yamaha DD-5 as an (attentuated) input into one of the mike input jacks.
I'm using one of the Radio Shack attentuators, and it's working fine. The cheap
price is worth it to me, since I'm in about the same class as the original noter
as regards soldering; I'd rather do a whole project than chance messing up some
commercial gear with a "simple" mod!
I've been pretty happy with the mixer, bye the bye. For cheaps ($69, I think),
it's manageable until the tunes pay for their creation...
|
| 1504.104 | Potentiometer terminology | KOAL::LAURENT | Hal Laurent, Loc: FOR, DTN: 378-6742 | Thu Feb 21 1991 17:45 | 11 |
| Well, once again I'm threatening to actually build one of these things (I
know, I've said that before, but this time I really mean it :-). I've been
looking in catalogs for pot prices (sigh, I suppose I have to expect the usual
jokes here), and I'm a little confused by some of the terminology. Perhaps
some of you hardware types could enlighten me as to what it meant by:
o Locking bushings (as opposed to standard bushings)
o Plain bushings (I'm not sure what this is opposed to!)
-Hal
|
| 1504.105 | Locking bushings | CTHULU::YERAZUNIS | Beware of programmers with screwdrivers | Fri Mar 01 1991 14:46 | 17 |
| Locking bushings means that there's a little hex nut on the attachment
sleeve that actually threads onto a tapered thread. This little hex
nut is a separate nut from the nut that holds the pot onto the
mounting. Tightening this little hex nut causes the shaft housing to
be squeezed together and keep the shaft from being turned by wandering
fingers, vibration, etc.
To adjust the potentiometer position, you loosen the little hex nut a
half-turn with a wrench, then turn the shaft.
No, you probably don't want this kind of potentiometer for the CQM.
Just go to Radio Shack and buy whatever they've got. :-)
[and no, I have no idea what the difference between plain and standard
bushings are]
-Bill
|
| 1504.106 | Digging it up, the old corpse... | COMET::BELLMJ | | Mon Jan 06 1992 19:31 | 34 |
| To Bill Yerazunis or anyone else out there:
I want to make an "all or nothing" effects bus (no level
control--maybe) on the 8 (well, 12) x 2 in 1504.58. What I'm thinking
is that I could have a DPDT switch throw a left and right pair from an
input onto either the main out or an aux out. Then, make four output
stages to make it a 8 x 4. Right? That works with no problems I
think.
Well, what would prevent one from then taking the aux outs, running
them to an effects processor, and then back into the mixer on the main
bus? Would that work?
What's wrong (someone told me it is) with simply running the inverted
input line to the effects, and then back again before the signal is
re-inverted? Couldn't you then put a potentiometer between the effect
return, signal and ground to function as an effect level?
The problem is that you need to buffer ins and outs, right? Well, why
not put two inverting unity gains on either side of the effect loop, or
for that matter two non-inverting unity gains?
I see no problem with the way I made an aux bus, then re-insert them
onto the main bus. Except that it goes through twice as many op-amps,
which implies more noise (even if they are very quiet).
Any comments? I'm using this for drums, so I'd really like to put
noise gates/reverb/whatever on all the channels.
Mike
P.S. What are some part numbers for the low noise JFET op amps? I
mean, the LM classification. I have a DigiKey book, but they
don't list TLO's. I'd like to find single and double chips, BTW.
|
| 1504.107 | This is the important question | COMET::BELLMJ | | Mon Jan 06 1992 19:43 | 16 |
|
MORE TO THE POINT (it's flash of brilliance time):
Can you put two (2) output stages off of one (1) mix???
Ex: ---- left out 1
- left stuff ---| (with amps)
---------etc. op amp----| ---- left out 2
- right stuff
etc?
If I can put two output amps off of one summed channel, then I can mix
the amount of effect in with the dry signal.
Mike
|
| 1504.108 | Never ending questions | COMET::BELLMJ | | Mon Jan 06 1992 20:04 | 17 |
|
One more question. Rat Shack books always show a resistor between the
+ of an op amp and ground, in an inverting set up. None of the amps on
.58 have a resistor between + and ground. Is this on purpose? What is
the reason behind a resistor there, and would it make a difference
putting one in?
In my new fangled idea, I need some inverters (unity gain) and using
Rat Shack's calculating method for gain control, I need one between +
and ground, and I'm wondering if this will "screw up" the existing
trend of non-resistance.
Or, does someone know an easy way to make a unity gain inverter without
the resistor?
Mike
|
| 1504.109 | | AUSSIE::SULLIVAN | sw still runs on hw | Tue Jan 07 1992 04:49 | 25 |
| The resistor from the +ve input to ground is to reduce the d.c error
that is caused by the input bias current drawn by the op amp. The value
should be equal to the other two (feedback & input) resistors in
parallel.
i.e Rin*Rfeedback
R+ = --------------
Rin + Rfeedback
Theoretically, the output d.c error voltage then becomes:
Verror = Ioffset * Rfeedback.
Since the offset current is much smaller than the bias current,
the error is reduced.
I have never designed a mixer - I'm just quoting a textbook.
I would have thought that this resistor would only be necessary
in a dc coupled system.
Also, I don't know how well all this applies to modern op amps
that have very high input impedances.
Greg.
|
| 1504.110 | Pots are expensive! | COMET::BELLMJ | | Tue Jan 07 1992 13:08 | 22 |
| Ok, great...so using modern JFET op amps should require no resistor to
ground, then. The Rat Shack books are talking about 741's, which I
don't even think are BiFET.
Therefore, I can just match the pre-resistor and the loop resistor to
make a unity gain inverter.
Just one more question, though, that hasn't been answered (I know it's
only been 12 hours! Thanks for the swift response on .108, tho!), and
that's whether I can tag two output stages off of the summed input
stages with no problems. (All of them are buffered, so why not?)
As soon as I find out, I'm on to the ambitious project of:
A 16 x 2, (10 mic, 6 line) each input with gain, balance, and effect
select switch, three (countem!) effects lines (1&2 for mic, 2&3 for
line) with separate mix-in gain on the main bus.
I've got the whole thing designed, + or - a few points, and I'm just
waiting for the go ahead on that assumption I made...
Mike
|
| 1504.111 | correction | AUSSIE::SULLIVAN | sw still runs on hw | Tue Jan 07 1992 16:37 | 14 |
| > Also, I don't know how well all this applies to modern op amps
> that have very high input impedances.
The problem has nothing to do with input impedance, it is due to the
finite (and relatively constant) current drawn by the bipolar input
stage of the op amp - so, yes, the problem is MUCH less severe when
using JFET op amps, but not for the reason I gave. Sorry to go on about
this, but I don't want to give you wrong information.
I personally would not bother with this resistor, even if using a 741.
(in this application).
|
| 1504.112 | no problems with proposed mod. | AUSSIE::SULLIVAN | sw still runs on hw | Fri Jan 10 1992 06:08 | 12 |
| Well, Mike didn't let me get away without answering his original
question (we chatted about it off line). When I eventually understood
what the proposed circuit modification was (driving two inverting
amplifiers with the output from a single inverting amp), I could see no
reason that it wouldn't work, other than the extra load due to the
second amp overloading the output of the previous stage. I think that
the chances of this happening are extremely remote (without having a
detailed schematic, though).
Corrections welcome!
Greg.
|
| 1504.113 | Sounds good to me | CTHULU::YERAZUNIS | Welcome to Addendia, Mr. Lee. | Wed Jan 15 1992 13:32 | 20 |
| [sorry to be offline for so long. -bill]
That sounds about right. Give it a try (you should prototype it first
before you buy lots of chips and sockets and wires and pots, though).
The nice thing about unity-gain buffers is that they have really nice
output characteristics (with respect to driving other inputs, that is.)
Just make sure you have at least 10Kohms between the output of the
buffer and the input of the next op-amp or potentiometer in the chain,
and you ought to do just fine. And of course, make SURE that you will
not have an inversion in your effects loop (i.e. if there's an
inverting amp in front of the output jack, make sure there's another
inverting amp in the path before you get back into the main signal path.)
{jeez, I build this little box, just a 4x1 with effects loop, and now
people are building 12x4's out of the same circuits. Maybe I'm in the
wrong line of work???}
-Bill
|