• Please review our updated Terms and Rules here

What did I do to my PDP-8 today.

Some days ago my 8/e from Wüllenbweber Gymnasium failed to run OS/8 and the simple "Kill The Bit" game (that mainly does an XOR between AC and SR). I changed most cards and suspect the SR to be the problem. From my spares I found a SR that worked, so I will repair the panel next days.

Both panels have differnet type of switches, the one with the reed relais is my spare.

Which one was the former part in production?



1704788095344.jpeg

1704788119884.jpeg


Have fun,

Volker
 
gnupublic said:
Both panels have differnet type of switches, the one with the reed relais is my spare.

Which one was the former part in production?
Click to expand...
The one in my 8/e is the top one. I don't think I have seen one with the reed switch. Is there an etch revision on the boards? I don't want to dismantle my machine to look for it.

The rework wires in your bottom photo makes me think that would have been the earlier one but that is just a guess as they could have screwed something up making a change.

Do the switch angles match your front panel graphics with the reed switch unit? I remember some discussion about this when Panelman was making his replacements.
 
The one with reed switches is the earlier one. I think I have one of each. The front panel graphics do differ between the two to get the angles correct, as Doug brought up.
 
From the electro mechanical standpoint I would have preferred the reed switch unit. From an esthetic standpoint I would prefer the Reed switch unit. It just looks better. But I bet it cost quite a lot more as well.

It will be interesting to know what broke?
 
I like the reed switch panel other than the fact that the switch can turn continuously, 360° and more. Just a funny quirk of it.
 
...don't forget the switch feeling with the reed. So soft.

Did not know the front panel graphic differences. I will look what fits best. I think it's an early 8/e, because the CPU has the old EAE layout.
 
antiquekid3 said:
The one with reed switches is the earlier one. I think I have one of each. The front panel graphics do differ between the two to get the angles correct, as Doug brought up.
Click to expand...

The front panel doesn't always perfectly line up even with original front panel and front panel board.
 
With the reed switch it seems to be 36° per switch, so it is exact 180° like the printing on the front panel graphic. The normal switch is a bit less, I guess 33° per switch. Thats is the difference between

SR36degree.png

and

SR33degree.png

So the panel of my 8/e is for the normal switch.
From the production dates of the IC's there is no real difference. Mostly 73 for the reed and two SRs with normal switch 73 and 74.

Have fun,
Volker
 
I was able to locate the problem with the buggy SR of the 8/e. Here are some timing signals from the defect one and a working one. The one that is working shows less Signals on SR->DATA. From this comparison it was easy to track the Problem to E15 a DEC314A that was broken. This has to mask SR->DATA only when an LAS or OSR instruction is running.

The interesting part for me was to see, that if you only test the loop:
7604 LAS
5000 JMP 0
there was the correct SR in the AC. It depends on the sequence of instructions if the too often SR->DATA leads to a problem or not. It was possible to run OS/8 in this constellation. Maybe this was an old error in this panel. Not a new grown one.

pasted-from-clipboard-2.png
 
Volker,

Good job finding the problem!

I see that E15 decodes the OSR portion of the operate instruction. 7404 actually 1111XXXXX1X0 is the pattern it is looking for. Did you have a spare DEC384 to replace it with? This would best be described as a 7 input AND gate with inverted inputs from looking at the schematic but is equivalent to a 7 input NOR gate. There does not seem to be an equivalent TTL part. After a quick search I see that this has been mentioned a couple of times in the forum.
 
Last edited:
gnupublic said:
Dough, I still have 3 spare of Signetics 314A. I have ordered 3 Signetic 314 (no A) on ebay and hope they will fit.
Click to expand...
I appear to have confused myself. Somehow I thought E15 was a DEC384. It looks like the Signetics 314 is a 7 input NOR gate. The pinout looks correct so is probably the right part. What is the diff between the no suffix and the A suffix?

The forum software let me correct a mistake in my previous post. I changed 7 input OR to 7 input NOR. I really needed to wake up more before I posted that.
 
DougIngraham said:
What is the diff between the no suffix and the A suffix?
Click to expand...
It's usually an indication the part is selected for a particular attribute (ex: leakage current, fanout drive current, rise-time, etc..)
ex: DEC part # 19-09972 can be DEC6314 or DEC6314A. Apparently the difference wasn't considered significant enough to warrent a new part number.
But a 19-09704 is a DEC314 and a 19-10391 is a DEC5314.
The 5314 should have lower leakage than a 314 from what I recall. So a 5314 or 6314 should be able to substitute for a 314 but not the reverse (even though it often works just fine). Maybe there's intermittent failures with temperature extremes or long-bus (60/80-slot) configs?
 
Today I got the M4 brass spacer for mounting the SUNON 20W fans for the power supply of my pdp8/e desktop.
The spacer are 30mm but the inner distance of the fans is nearly 31mm. So I soldered a washer on top of the spacer and cut a tangent line:

IMG_7666.JPG

This mounted to the fans should be stable enough and should not turn when fixing the screws.

IMG_7668.jpg

I put it together with the cooler cage:

IMG_7669.jpg

and finally mounted it. This is very tight-fitting. You do not really see if there are all cables free ore squeezed. But in the end it was in and everything works as expected. The new SUNON fans have a fair sound, better than the PAPST 4606N (19W) with metal fan blades. I was unsure to take the 12W Papst or the 19W Papst, but te order of the 30mm spacer came first, so I could not resist to bring all together.

Next step is to check the board that came with this nice desktop, but on first sight it looks quite good. M8650 has a problem and M8310 also. But the rest maybe OK.

IMG_7692.jpg


Have fun,
Volker
 
Volker,

DEC originally used two washers to make the spacer fit. They used an assembly order trick when putting it all together to make it go smoothly. I spent way to much time when I last replaced the fans figuring it out. The spacing on currently made fans is enough different that I ended up doing it a different way. I am glad you found a solution that worked for you. That is a pretty nice looking E. I have not seen a cover like that one. I am certain you will resolve your remaining issues
 
Thanks Dough for the kind words. There was one fan inside without spacers. Just nuts, I had to pull the screws while turning, to stop the nut from turning itself. It was a pain to get it out. The other fan had 3 spacer but without beeing fixed. So I had to find a solution for mounting the new fans. Next time I will not use brass, I will cut some parts with the lathe.

And yes, it's a GE Terminet data printer 300. A friend saved it from a collection, and gave it to me. I got it some weeks ago. Inside it's very clean. Maybe never used. But it's not top of my list of todo's at the moment.
 
Today, one week of holiday comes to an end.
But it was great, today I finally got my PC04 running! I was able to load a RIM and a BIN papertape.
The biggest problem are these aweful cables. The core wire is broken at the Berg connecter, where it is crimped. That is awful to repair. I tested all contacts, and they seem to be OK. But after I got all back in the rack, it does not work anymore. I will have to make new cables and try again.
I noticed, that the photodiode amplifier needs the bias in a quite small range to work. But all cells where OK and I only hat to adjust the light very little.
I'm glad I had at least one working M840!


1712525989370.jpeg

Have fun,
Volker
 
Congrats. I wish I could get mine to work. That is an interesting front panel on your 8/e where is it from?
 
The 8/blue is from Wühlenweber Gymnasium of Bergneustadt, a school in Germany. They had two PDP-8/E, one normal and this one in blue. This front was build from a DEC technician. This is done very professional and the switches looks like PDP-15 switches. The frontplate is a 2mm steal plate with the colored printing, also the labeling. The cutout for the diodes are punched. The SR is from an 8/m with diodes. in the right corner this is a thermal sticker that lost it's function over the years.
The housing is an 8/e with two backplanes inside. I like this uniq machine very much. But if you toggle the switches, it is a bit confusing. I got both machines from Gerold Pauler who got them from the school 20 years ago, or so. (The Dectapes of the school are here: https://www.pdp8online.com/images/images/Wullenweber.shtml )
 
You must log in or register to reply here.
Back
Top