Conference vmszoo::rc

    Frank,
    
    Sounds like the clue here may be "engine VIBRATION."  I'd look
    real hard at things like: is the receiver packed too tight?...vi-
    bration can be induced into it if you've "protected" it to the ex-
    tent that it's not free enough to "float" a little;  is the antenna
    routed properly?...if too close to the battery and/or servo's, you
    can inductively induce glitches into the system. A good rule of
    thumb is to consider the antenna wire to be 3-inches in diameter,
    which means keep the antenna a minimum of 1 1/2" away from other
    system components wherever possible and exit it from the fuselage
    as close as possible to the rcvr. location;  any metal-to-metal   
    connections anywhere, i.e. throttle arm-to-metal clevis, metal-hubbed
    wheels, etc?...I know modern radios are "supposed" to tolerate this
    kind of electrical "noise" but, WHY CHANCE IT?  Avoid metal-to-metal
    everywhere you can; it's not unusual for a system to chatter a bit
    when glo-plug is connected/disconnected...is your plug disconnected
    when the system chatters with transmitter on?...if so, try it with
    the plug disconnected. If the symptom disappears, you're probably
    OK.
     
    If none of this helps, try to borrow a transmitter on the same fre-
    quency (most any brand "should" work well enough for this test)
    and see if it behaves in the same manner.  If the symptom remains,
    something in your airborne installation is vibration sensitive or
    you may have a bad crystal or other component in your rcvr.  If
    the problem goes away, something's amiss with the x-mttr., maybe
    just tuning/alignment.  In either of these possibilities, send the
    entire system back for an R&R...it costs a lot less than replacing
    even one airplane!
                      
    Hope I've given you some ideas that help to identify the "gremlin."
    Let us hear how this situation comes out.
    
    Adios,	Al

    Al-
    
    The receiver is a very tight fit, I'll try to repack it a little
    looser.  Also, I'll check the antenna routing closely; it probably
    is within the "magic 3" of the servos.  Also, I have a music wire
    throttle control from the servo right to the engine; this may add
    it's own "music" as well!
    
    Sounds like I have several areas to try... I'll try correcting them
    in this order (happens to be most likely source of noise, easiest to 
    fix order).
    
    THANKS for the suggestions... I'll let you know how it goes.
    
    Regards-
    Frank.
                                 

    My late, lamented Legend exhibited a problem seen on a few of my
    other planes that also used two aileron and two flap servos.
    
    While holding the plane in my right hand, ready to launch, and the
    xmtr. in my left hand, the ailerons would go crazy. Actually, a
    very regular up/down beat as if they were emulating an ornithopter.
    
    If I kept the xmtr. antenna at right angles to the wiring in the
    wing it wouldn't happen. At any distance greater than two feet plane/
    xmtr. it didn't happen so it was mainly an annoyance rather than
    concern.
    
    I decided to experiment, so hooked up all six servos on the bench,
    and arranged them as in the plane. By holding the xmtr. antenna,
    fully extended, parallel to the aileron servo wiring they would
    both start jittering when the antenna got ~ 6-8 in. away. The flap
    servos were never affected.
    
    As a first fix, I installed one ferrite bead (the type with two
    holes through it) over the power and ground leads on both sides,
    plus put a 220pf cap across the p & g leads on one side only.
    
    Made no difference at all.
    
    As a second fix, I removed the cap, and put a second ferrite bead
    on both sides, between the signal and ground leads. All four ferrite
    beads were located ~20 in. inboard from the servos, only because
    there was a handy splice where I could break the line.
    
    Results: Eureka ! I could lay the antenna right on top of the leads
    and not a twitch.
    
    The wiring will go back into the new Legend, and no more embarrasing
    speeches to the assembled multitudes on the fundamentals of signal
    coupling. ("Jeez, mister. Why's your airplane doing that?").
    
    Terry

This interesting radio diagnosis just in from the use net.

Bye          --+--
Kay R. Fisher  |
---------------O---------------
################################################################################


I thought the net might be interested in this...

I lost a plane last weekend when the Rx apparently suffered a major failure.
I know it wasn't just interference, since the PCM failsafe would have activated
and the plane would have just flown away (Yes I was straight and level at the
time, flying advanced pattern).

After picking up the pieces (surprisingly few, just broke the nose off at the
canopy, and dislodged the wings and everything else inside) I found that
one of my airleron connectors (servors in the wing panels, connected via
a 'Y' at the receiver) was disconnected. No big deal I thought, it just came 
off in the crash. However, thinking about this later I remembered that
my planes demise began with an uncommanded right roll, so I started wondering.

Anyway, I took my complete radio setup into the lab and did some measurements.
I found that if one servo connection in a 'Y' setup is intermitant it can cause
both servos to act eratically, both stop, both jump around, one stops the other
jumps, etc. After some probing around with an oscilloscope, I discovered that
if the ground lead breaks contact, while the power and signal lines are still
connected, then both servos stop. Why is this I wondered. After measuring
the signal lead I found my answer. The signal no longer switched between
GND and Vbat (Vbat depends on your battery level) it was being biased and
now went between approx Vbat/2 and Vbat. This is why the other 'good connection'
servo stopped too, it could no longer find a good signal input.

This explained my loss of airleron control, but how about the other channels?
Were they affected by any of this? The answer is yes!

I now measured the signal going to the Elev servo, while causing the intermitant
on the airleron servo. I found that there is significant noise injected
into the pulse , but this alone would not cause a problem, it was mostly
due to the airleron sevro motors turning on and off and loading the power
supply. The major problem was that the Receiver lost it's timing, ie.
the frame rate of the signal to the elev servo was not constant, there were
many extra pulses and (I think this is the real problem) the width of the pulse
was seen to change. 

While I was  causing the intermitant by hand, I speculate that a real 
intermitant in the plane, where things are really vibrating, would be disaster
for the Rx.

After all these measurements, I suspect that my crash may have been caused
by a simple bad connector causing an intermitant which caused the Rx to lose it.

I use the Ernst connector clips on all my 'permantent' connections in the plane,
ie battery to switch, extension leads, and I think now I will also take the
time to use them on my airleron connections, even though it may be a pain to
secure them each time.

Total cost of this learning experience: Plane ~400, Radio ~450 (if I decide not
to reuse my old one) :( :( :(


++++++++++++++++++++++ The full NEWS header follows +++++++++++++++++++++++++++
News Article 16299
Newsgroups: rec.models.rc
Path: nntpd.lkg.dec.com!nntpd2.cxo.dec.com!pa.dec.com!decwrl!spool.mu.edu!howland.reston.ans.net!torn!nott!bnrgate!bmerh85!bnr.ca!tomt
From: [email protected] (Thomas Taylor)
Subject: Intermitant Connector measurements
Message-ID: <[email protected]>
Sender: [email protected] (Usenet News)
Reply-To: [email protected] (Thomas Taylor)
Organization: Bell-Northern Research Ltd.
Date: Sun, 25 Jul 93 14:00:26 GMT
Lines: 52

    Verrrry interestink!. Say "Good night" Lucy!.
    
    I noticed an interesting result on my little servo ma usage test rig.
    When I was plugging in servos the reading between the Nicad and the RX
    would jump at least 1/2 an amp.
    
    Sometimes the existing servos would jump as well. I wonder if the last 
    note is the reason why. Plugging in a servo is like a loose servo!.
    
    E.

    On the subject of measuring servo signals and the like. I had a idea
    about analysing all the servo outputs by connecting them to a
    laptop/notebook for graphical analysis of a flight. One could have a
    second rx on the same channel hooked up via a parallel interface to
    log the data. Lots of questions arise on if its possible but it would
    be useful. Problem with the PC parallel interface although generally
    spec'ed as I/O they are output only and hence a require a proper
    parallel card and my DECpc notebook aint got room for such.. 
    
    I interested in this project and hope to get somewhere with it if I can
    solve the interface problem. The serial interface would probably be too
    slow and would require additional logic to convert the channels to a
    serial stream. Although if the DSC was active while doing normal tx'ing
    the serial stream migh hack it. But then I don't now the data rate. Any
    info would help ie data rates, signal specs etc.
    
    Eric.

    
    Eric,
    
    Are you sure about the parallel port being output only? I've used
    software like LapLink that uses the parallel port I/O with no problems.
    
    
    Regards
    
    Nigel
    

Another possibility would be using a parallel to serial printer buffer 
and inputting the serial stream. These are generally used in the output 
direction but I can't see why you couldn't drive the parallel side and 
get a serial stream out as input. That would give you some buffering so 
timing wouldn't be quite as critical. You're going to need to decode 
the pulsewidth for each channel anyway so you've got a lot of flexibility 
with the interface.

    Are you planning on putting the Notebook PC in the plane?  You need a
    way to store the stream in the plane or to transmitt it back to mother
    earth.  Most parallel ports can be programmed as inputs - otherwise
    Laplink wouldn't work as stated above.  
    
    	The data rate is about 20 updates per second for PPM.
    
    	The pulse width is 1 - 2 Msec.
    
    A small board with a uP and some memory could store a flight's data for
    later upload to a PC.  I'd then use the serial port.  ALso, you would
    not need to store every frame received by the Rx.  One sample per
    second would be plenty for replaying a flight.  This would be only 600
    samples for a 10minute flight (per active channel)  
    
    Charlie
    

Earlier he stated that he'd use TWO Rxs, one to fly the plane and one 
on the ground to "listen in".

    Robbe now sells an onboard computer for R/C airplanes.  I do not
    know too many details as I read about it in the new products column
    of a foreign publication.
    
    I believe it has speed sensor transducers, altimeter, current
    sensors(this could probably measure servo loads) and a few other
    bells and whistles.
    
    Maybe some of our German contributors could fill in the details, or
    you might try calling Robbe International in New Jersey (908)359-2115.
    
    
    Static measurements taken on the bench can be very misleading
    particularly in propeller performance analysis, evaluating
    electric motor/battery/prop combinations etc.
    
    
                                                      Regards,
    
                                                      Jim
    
                                              

    Maybe the parallel port is bi-directional - have to do some checking
    first.
    
    The on-board uP with the sensors is the idea way to go especially if
    one can get information about the orientation of the plane and then do
    some match up with the servo movements. It should give a fairly good
    result. Since I have developed the CGM (computer Graphics Metafile) I
    could link in the code and produce an animated flight for displaying on
    a PC or VAX. I have written a standalone viewer for the  PC so all I
    have  to do is decode the rx, calibrate the data vs actual movements,
    create the graphics file and then display it. I can back fill the
    movements of the plane graphic with the background color and it should
    look good. Then one could close in on an aerobatic and find out what
    fingers went the wrong way (often happens all the time).
    
    I must call my local hobby shop for details of the Robbe product sounds
    like interesting stuff. 
    
    Once we have the altitude, speed, rotation etc sensors...
    
    Regards,
    
    Eric.

    Just checked the decpc_portables notes conference and one of the notes
    says that in bye gone days the parallel port was unidirectional, which
    is what I read in my older documentation. So I had a look at the setup
    and it allows the port to be I/O or O. So off I go... I was looking for
    a chance to get back into writing ISR's in ASM for the 80386. At 20
    frames a second and 1-2ms pulses there is plenty of time to work the
    port. Not forgetting that I have to be able to process all channels in
    that time. I will look at designing in X samples per second with N
    channels. The samples could be reduced when the servo movement is
    mininal and increase it when a large servo throw occurs thus I can have
    reolution when its needed. 
    
    Better still I can earn some bonus points back home, after all it is
    not RC but computer work... should go down 8-).
    
    Now I just have to find the time to do it all.
    
    Eric.

    Re: .11
    
    Jim,
    
    sorry, I'm not familiar with what robbe came up with this year. In
    fact, now that I think about it, I haven't even seen a robbe news
    catalog of this year. Might have to do with the local hobby shop and
    their relationship with the responsible robbe representative. I know
    that this relationship is bad. I'll keep my eyes open whenever I go
    somewhere else.
    
    There are some interesting changes going on here in Germany: Last year
    Graupner came up with a unit you can measure and store in-flight data
    with - in their catalogue. Not sure whether they ever delivered any. It
    was prepared to measure current (high current, I seem to recall up to
    40 A) and was prepared to connect to additional sensors for (air)speed,
    rpm, height, vertical velocity etc. They didn't offer any of the
    additional units in this year's catalog, so I doubt they actually
    started to carry it.
    
    Aeronaut this year has a system to store multiple sets of data in
    flight. I have their catalogue at home and can look it up if you're
    interested.
    
    Even more interesting: The Bundespost (German FCC) lifted the ban for
    on-board transmitters last year. That means that the small vertical
    speed indicators are now legal! They feature a low-output TX on board, 
    a small receiver in your pocket and an earphone for the pilot. Their
    sound is pretty much like in real sailplanes. They have been sold for
    quite some time, but were until then illegal for use in Germany. Their
    small output of 1 mW however made it hard to detect them. 
    
    Now that this concept is legal, it opens a complete new market. There
    was a big ad in the June FMT for a complete telemetric system that will
    transmit a lot of data from the plane to the ground (altitude, vertical
    speed indicator, airspeed, rpm, voltage, motor battery current,
    receiver current etc.). A full blown system is expensive, though...
    
    Another idea is what Simprop came up with last year: At least one of
    their receivers stores flight data. If you hook it up to your TX (with
    the nice LCD display), it will tell you after the flight how often it
    went into failsafe and why (interference or low battery). So you can
    detect danger even if you didn't see anything during the flight.
    
    There is a lot more room for technology in our hobby...
    
    Please ask detailed questions if you're interested in more, I'll try to
    dig it up for you.
    
    Best regards,
                  Hartmut

Re: .13

Eric, just record the values when there's a change. That should cut down 
on the amount of data to store (the LG only goes up at the beginning and 
down at the end)

                                                    
    
    >> -< Here's one European... >-
    
    I'm another !
    

    >>Re    Now that this concept is legal, it opens a complete new market. There
    >>was a big ad in the June FMT for a complete telemetric system that will
    >>transmit a lot of data from the plane to the ground (altitude, vertical
    >>speed indicator, airspeed, rpm, voltage, motor battery current,
    >>receiver current etc.). A full blown system is expensive, though...
    
    Could you forward a copy or if it is not in english put the details in
    here. 
    
    Eric.

Re: .15

> the LG only goes up at the beginning and down at the end
                                           ---------------

... and this not even every time... (duck)

    Re: .16
    
    Of course, the ad is in German... I'll try to remember to bring it in
    and put the details into a new note (I hope the telemetric system
    doesn't have anything to do with servo jitter).
    
    Best regards,
                  Hartmut

    Another question regarding servo jitter...
    
    Last weekend, I test flew my electric Fiesta (14 cells, Ke 40/10). I
    had put a new (used) dual conversion receiver into it and a new (used
    motor controller (Graupner Power Mos 45). I did a range check first
    which showed questionable results (only short range). The receiver was
    checked by the factory before I put it into the plane.
    
    I proceeded to fly anyway, but did get hit several times (was fairly
    good when the plane was close by), so I landed as quickly as possible. 
    I have to admit that the arrangement of the radio systme is not ideal,
    due to space and CG limitations. It goes like this:
    
    	Motor | Receiver   | rudder servo | spoiler servo |wing| motor
              | Controller | elev. servo  | RX pack       |spar| battery
    
    
    I did have problems occasionally before, but had expected them to go
    away with the dual conversion receiver, but it became worse. The
    antenna is routed through a pushrod housing in the fuselage.
    
    I did a couple of tests while at the field. Disconnecting the motor
    battery didn't help! Only when I took the battery out of the plane, it
    seemed to be better. Since I had spectators, I didn't want to spend the
    whole night on failure analysis. I tried to find the problem last night
    in the shop, but it seems that the problem is hard to reproduce inside
    the house. I intend to reroute the antenna and hope that will fix it.
    Any other ideas on what to look for would be greatly appreciated.
    
    Best regards,
                  Hartmut

Eric Cullen,

Since you have just a single bit of data to put into the laptop, you
could use one of the HANDSHAKING signals that are always an input on a
parallel port.  This would work on any PC parallel port, not just one of
the "newer" bidirectional ones.

BTW - I think that the Laplink cable uses this trick and actually
transfers only 4 bits at a time 'cause that's how many input handshaking
lines there are (plus a fifth one for handshaking of the Laplink data.)

                                - Dan

    Dan,
    
    Yes I guess I could use the HANDSHAKING signals (ACK,BUSY,PAPER OUT,
    SELECTED, etc), if I were to use on 'older' type parallel port. To
    drive the whole thing on inturrupts I would need to have additional
    logic setting a strobe or data ready signal so that I wouldn't have to
    read the port all the time to see if the data changed. But for
    simplicities sake I guess reading the port via a timer tick would be
    good enough and to go process the data if it is new.
    
    The other Eric.

    
    Here is one for ya, I have a Aristo-Craft Challenger 720 7channels/dual
    rates. The problem is when I move my rudder (which is aileron using it
    as a three channel) that is ok, when I use my elevator that is ok but
    when I use them together I get interferance badly. Can anyone help me
    with this????
    
    Thanks
    
    Bruce
    

    I suspect two things.
    
    1. A bad cell in the RX pack - Measure the votage under load
    
    2. A damaged RX. (Has it been crashed?) - Do a range check.
    
    
    RX's have a voltage below which they start to lose it!. :-) Operating two
    servos at once will pull the votlage down, perhaps below what the RX
    will tollerate. Usually this means you have a bad cell.
                                                  
    
    I have seen the behaviour that you describe in both of the above
    instances.
    
    Brain empty - moving on. :-)
    
    E.

    
    Thanks Evil,
    
      This weekend I will try a new battery pack. It hasn't crashed that I
    can remember, once on a lake with my Cub but that really wasn't hard.
    
    Bruce