Conference 7.286::digital

re: .0

Here's what the Dick Best book says about it.  Maybe some of these people 
are still around?


    G8018      DESC: CORE POWER SUPPLY REGULATOR FOR 8A400,8A500 (H9300)
         WHERE USED: 8/A

 STATUS DATE: 12/79
 STATUS CODE: 5  BUILD TO STOCK

 ENG MGR:  SNWS SAMUEL N WICKS @NIO
 DES ENGR:
 PROD MGR: SFK  SANDRA F KENDALL @DAS
 SVCS:
 MFG REP:  ALR  ANGEL  RIVERA @AGO

 PROD CAT: FA   - MODEL NUMBERED PRINTED CIRCUIT BOARD ASSEMBLY (MODULES)
 PART TYPE: MOD      VOLTAGE CODE: 0    FCC CODE:
 MAJ SUPPL: 017FS - FIELD SERVICE WAREHO

I've sent Guenther the responsible engineers names and a contact for the
documentation for G8018.

Judy Kilgore
Storage Tech Design & Documentation

    
    I prefer to think of the pdp-8 as the original RISC processor:-)

	I've got an 8/e sitting up in my attic.  I bought it about 9 years
	ago from UNH for $15.  I'm still looking for replacement bulbs
	for the console :-)

	Sadly I no longer have my 11/40, I was still living at home at the
	time and my folks told me to either get rid of it or they would
	get rid of me :-)  I should of kept the cabinet however (and those
	big lead bars!) ....

    I am an MCS Engineer located in Milwaukee. You are probably referring
    to a Kearney and Trecker CNC milling machine. I probably have some
    power supply prints at home if you're really interested. Also, these
    particular PDP8A's have two power supplies. There is also fan sensing
    circuitry either bulit into the backplane or on a separate plug-in
    flip chip card depending on the vintage of PDP8A 20 slot chassis. The
    fans themselves and the sensors were a real problem in these systems.
    These can cause power supply "symptoms" as well. Anyway, the G8018
    should still be available. I currently see one in Willmington, DE and 
    one in Baltimore, MD logistics. A customer should be able to purchase 
    one thru their nearest Digital Servicenter for $371.
    
    Dale
    (414)-792-2537
                  
    

    Wow!!! - critical customer outage on a 20-year old machine and you've got
    online VTX CEO option reports, e-mail addresses, nostalgia, material
    availability estimates and even a knowledgable Field Service 
    diagnostic and replacement quotation....
    
    	All via NOTES in 2 hours!!
    
    Gotta love it
    
    /Chris/
    
    		

    re.7: -< Go NOTES Go >-
    
    When my cousin called me this morning I thought "oh boy, these
    mechanical engineers have to change the way they think about life
    cycles...we'll never find anything about that piece". Now it seems like
    he can order it from local digital office or in the worst case he might
    get a copy of the print set (under non-disclosure I heard).
    
    NOTES is a perfect tool...but is only as good as the people using it!
    Cheers to all...
    
    Guenther

Guenther:

  And if that ain't enough, the odds are high you could buy an
  entire 8A620 CPU box. Alternatively, a PDP-8/E would almost
  certainly run the same codes unless they're using the 8/A FPU.

  If they're not using all the slots, the smaller 8/F or 8/M
  would work too.

  And I'll bet a bunch of us could chip-shoot the power supply. :-)

                                      Atlant

    I have been in field service for 22 years, so this machine was state
    of the art back then.  Besides the suggestion in .6, another senario
    that I ran into could be the culprit.
    
    On the backplane, or what is referred to as the "centerwall" in this 
    case, there is a circuit that feeds a 120V triac.  Over to the left of
    it is a 47ohm resister that often failed, and occasionally took out the
    triac itself.  The triac provides AC to the power supplies.  If the 
    resistor fails (and unfortunately it sits off to the side and not
    easily seen - it feeds the 5v to the gate of the triac) then the triac
    doesn't turn on - hence dead power supplies!  Also, both power supplies
    must be in the machine!  They are effective interlocked by etches on
    the power supplies; one being out does not complete a necessary
    circuit.  
    If a fan sensor is out, the symptoms are that the system will stay on
    for about 15 to 20 seconds and then shut down.  If this machine has the
    old fan sensors, you will see a flat etch board with a vane on it.  If
    it fails, just disconnect it.  On the "newer" models, there was a small
    magnet on the rotator blade of the fan, with a pickup that put out a
    pulse as the blade rotated. There is a green handled module located on
    the platform between the two power suppies that contains circuitry to
    check for these pulses at a rate greater than x(?).  If one of the fans
    slows down or stops...... power supply shuts down.  There are four
    fans, but only three are checked.  To replace a fan with this type
    sensor you need the proper fan, or
          take the white wire from the defective fan harness and connect it
          to the white wire on a fan that is working okay.  This will fake
          out the circuit, so you can replace it with a regular fan
          temporarily.  It is a TA450 fan, and I'll get you the part number
          if you need it.
    As .6 stated, there still out there.  We have at least 4 systems out
    there - some with paper tape, some with 8 inch floppies, some with RK05
    disk drives.  It was a darn good machine.
    
    

    
    	Only from New Jersey would you get a reply like .10 to the
    component level. Good memory on the stuff that bit you once.
    
    /Irv Luizza
    
    

On CNN they recently featured a Company that specializes in `geriatric' 
computers. They're called: GD California. They repair old stuff and even make
new parts for them.
The CNN.COM Sci-Tech pages still have the write up.

M

    
    Got one in Calgary, ALberta
    
    
    
    Image Database Review               Site detail display required (Y/N)
    ? Y
    Part number:    G8018-00       Description: H774 REGULATOR                
    Continue ?           Total on hand: 00001         VALUE: $         
    
    CCsite|OnHnd|DefOH|OTR  |Com  |Orderd|CCsite|OnHnd|DefOH|OTR  |Com  |OO    
    7WK 00|    1|     |     |    1|      |      |     |     |     |     |
    
    

    O.K.  How many of you can remember how to toggle in RIM - let alone how to
    do it upside-down on an 8/I? (Never liked those rocker switches).
    
    As to the 8/E owner - my experience had always been that if you
    replaced six burnt-out lamps on the control panel - by the time you
    buttoned it up you'd have seven other burnt-out lamps:-)
    (Better the LEDS on the 8/M)
    
    BTW, it's been only two or three years since my last OS/8 support call!
    

> (Never liked those [8/I] rocker switches)

  But at least you could quickly reset them all to zero by playing
  a "glissando" kind of move on them. The sharper edges of the lever
  switches of the 8/L and 8/E/F/M would hurt your fingers if you
  tried to do this. There, you needed a broad-hand kind of thwacking
  motion instead. Toggles reached their zenith with the 11/40/45/50/
  55 toggles: great feel and nice looking, too. But the '10 folks
  wil probably say the KI pushbuttons were faster, and they may well
  be right. :-)

 -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

  I've told this story before, but what the heck...

> As to the 8/E owner - my experience had always been that if you
> replaced six burnt-out lamps on the control panel - by the time you
> buttoned it up you'd have seven other burnt-out lamps:-)
> (Better the LEDS on the 8/M)
    
  The 8/E was no problem -- it had socketed lamps. But on the 8/I
  and 8/L, the impact of the spring-loaded vacuum solder sucker
  would definitely deliver enough shock to break marginal fila-
  ments. So what we used to do was to dismount the console PCB,
  whack it on the table a few times, and then temporarily re-
  mount and mark all the dead lamps.

  Then, having broken all (or certainly most) of the weak fila-
  ments, you could go ahead and replace the lamps, confident that:

    o Few, if any, more lamps would break as you unsoldered
      the duds, and

    o Few lamps would burn out between now and the next PM
      as you'd dispatched all or most of the weak ones.


  Now, do you remember the sequence of illegal key presses to get
  *ALL* the lamps on the five (?) rows of the 8/I/L data displays
  illuminated? (I'm not sure without the real thing in front of me.)

                                   Atlant

Re .15:

>But the '10 folks
>  wil probably say the KI pushbuttons were faster, and they may well
>  be right.

Nah, the KA rockers were the way to go, for the reason you stated.
				/AHM

	Another 8 story (probably worked for any system w/core memory).
	There used to be a set of programs that when run would play
	music if you placed an AM radio nearby.  Some of the tunes
	used to also beep the console as an additional instrument.
	I believe alot of them were Xmas tunes and one year someone
	ran a mike into the computer room and the tunes were broadcast
	over the PA system.

    When I was in college, a friend worked out the tunes-over-AM-radio
    program for an IBM 1401 that we had ... complete with ooom-pah bass
    lines coming out on the 1403 (old-style impact) printer.

> 	Another 8 story (probably worked for any system w/core memory).
> 	There used to be a set of programs that when run would play
> 	music if you placed an AM radio nearby.

  That would be MUSIC.SV.  It could read text files that contained
  up to four part harmony (but not full orchestrations! :-) ).
  Someone did a variant of it that used one the AC (Accumulator)
  LEDs on the 8/M. Optically coupled, you could get better fidelity.

                                   Atlant

    NOTES! - 
    
    Where else can ya find a community that rallys together to solve
    a problem in a few replies then spends the next 20 reminiscing to the
    point of "optically coupled oompah devices"
    
    Sheesh - you guys are really - well -     W E I R D  !
    
    /Chris/

Chris:

> Sheesh - you guys are really - well -     W E I R D  !
    
  Don't look now, but you're here too! :-)

  Anyway, thanks for the compliment!

                                   Atlant

    
    Would that W E I R D
    
    or         W I R E D ...
    
    Notes is a good thing,
    Dave Y.

Re .21: why single out Notes as *the* way of doing things. AltaVista Forum does 
pretty much the same thing, and is getting more public viewing than Notes ever 
did.

Notes has enormous inertia and tradition in Digital, but this conference could 
easily be a forum. (As soon as someone does a Notes/Forum gateway...)

It's not the tool that does the work, it's the people using it.

(And you'll probably find that quite a bit of this kind of thing goes on in 
newsgroups as well. comp.arch springs to mind, and there's probably some kind 
of comp.folklore as well.)

PJDM

> (And you'll probably find that quite a bit of this kind of thing goes on in 
> newsgroups as well. comp.arch springs to mind, and there's probably some kind 
> of comp.folklore as well.)

	the problem with newsgroups however is that someone looking for
	the same information a month later will have to ask all over again,
	with the answers repeated .... because most news servers don't
	archive the articles posted to the newsgroups for that long a
	period of time.  Hence the creation of the FAQ lists which
	makes USENET vs. DECnotes like dynamic memory (which needs to
	be continuous refreshed of the info will be lost) vs. static
	memory (no refresh).

	How long do Web forums (AltaVista or otherwise) maintain
	articles/notes?  (my guess is forever until manually deleted
	or explicit experiations placed on articles, ie. more like
	notes than USENET?)

	The main difference between AVF and Notes, in my opinion, is
	the method of navigation. It takes me about 3-5 times longer to
	"mouse" thru Forum than it does for me to go thru the same
	amount of data in a Notesfile. This is because Notes has the
	Enter key on the keypad mapped to an intelligent function of
	navigation. 

	It seems many engineers have forgotten the useful keypad. It's
	nothing more than a calculator keypad in the Windows/Internet
	world.

	That's my main gripe with AVF.
							mike

    And I have had it with remembering yet another password.  I don't care how
    you implement the next-unseen functionality, I don't want to enter a
    password to navigate.
    
    - Vikas

    Sooo, did the customer get satisfied??
    
    Not sure from the varied, yet highly interesting, and highly-Digital
    responses herein.
    
    Doug
    
    p.s. Someone should send this story to Digital Today.  We're not just
    new stuff, we're experienced engineers supporting legacy appications
    that saved the customer plenty on upgrades!

    Seems like there are 2 power supplies imediately available from the US
    but my cousin lives in Germany. Mail to colleagues in Germany have not
    been answered yet...they work only a 35 hour work week ;-).
    
    If someone knows how/who to contact in Germany, Mannheim area to get
    this part let me know.
    
    Thanks
    Guenther
    
    P.S. It sure ain't easy to deal with a corporation like digital when
    you have to leave the NOTES safety net.