• Please review our updated Terms and Rules here

PDP-8/I Repair

antiquekid3

antiquekid3

Veteran Member
Joined
Nov 10, 2009
Messages
593
Location
Alabama
I figured I'd start a new thread for my journey into bringing back a PDP-8/I from the dead. I acquired this system several years ago from a relatively local collector, but life and other projects have gotten in the way of working on it. In fact, I still have other projects that are battling for my time, so I expect this will come and go in intermittent bursts.

When I got the system, the previous owner had bypassed the original 704A power supply and replaced it with a 716 and H721. I would like to get the 704A working to get the system back to originality. A while back, I removed the power distribution board and capacitor for the ferroresonant transformer, in hopes to get those fixed up and tested, respectively. However, based on a little testing of the capacitor (and a replacement I purchased), I'm beginning to think there may be another issue, and the cap may have been a red herring. In any event, I removed the rest of the 704A supply today from the rack to test on the bench. I already noticed some of the spade connectors were not looking too hot. In fact, it looks like some of them had gotten quite hot before and degraded the insulation.

Anyways, here's an album with some photos of the system. It'll be nice to have a functional machine with a TC01/TU55/TU56 and RK05 to dump some of the tapes and disk packs that I've got.

Kyle
 
I see a set of four that look like they got excessively warm at some point. It may be possible to chemically remove the corrosion and restore a good enough contact to prevent overheating. Otherwise you might have to put new connectors on. I've trieds to replace the wires before, and had that turn into a nightmare, as things are zip-tied in with each other in a seemingly random pattern, and if you're not careful you end up effectively rebuilding the whole harness.

Vince
 
Enjoyed looking at the photos of this system and am looking forward to reading about the journey to get it running again.
 
I removed the 3-board set from the DW8E when I first got the system, with the intent of testing them in my 8/M. However, I never got around to that. I still have a few RK05s that I need to work on, but perhaps the one with the system is the nicest candidate now.

What's the difference between the RK8-E and the -F? I am sure I knew at one point, but can't remember.

I think I've got a full 32k in the system. 8k in the 8/I backplane, 16k of MM8Is above the 8/I, and 8k above the DM01. Not too shabby!
 
antiquekid3 said:
What's the difference between the RK8-E and the -F? I am sure I knew at one point, but can't remember.
Click to expand...
The DW8E only supports one Data-Break device per backplane. For a cost reduction the DW8/F doesn't have the ability to remove some of the Data-Break signals from the Omnibus. Since it is the only Data-Break device on the bus, it doesn't matter.
 
I probably ought to restore them together, as you suggest, Roland. We will see how my patience is by the time I haul the good one home and up to the lab! No kidding, I developed a hernia and had to have it surgically repaired because I thought I could haul an RK05 up the stairs by myself. Well, I did, just not without consequences!

I'll be delighted to have just one running for now.

Thanks for the info on the DW8 and such. I'll certainly be curious to do more digging on that and understand the implications of the cost-cutting measures.

Are there other limitations of what devices can go in the box?
 
antiquekid3 said:
Thanks for the info on the DW8 and such. I'll certainly be curious to do more digging on that and understand the implications of the cost-cutting measures.

Are there other limitations of what devices can go in the box?
Click to expand...
I don't think that there are any limitations, other than just one Data-Break device per Omnibus in the DW8E. There are actually two Omnibus in the DW8E. You need two sets of interface cards in the DW8E to use both Omnibus.

There are some test programs for Warren's FlipChip tester for some of the DW8E boards.
 
antiquekid3 said:
Anyways, here's an album with some photos of the system. It'll be nice to have a functional machine with a TC01/TU55/TU56 and RK05 to dump some of the tapes and disk packs that I've got.

Kyle
Click to expand...
Thanks for the photos Kyle. It looks like a nice system. I am looking forward to your progress reports while restoring it. I wish you good luck and success!
 
m_thompson said:
I don't think that there are any limitations, other than just one Data-Break device per Omnibus in the DW8E. There are actually two Omnibus in the DW8E. You need two sets of interface cards in the DW8E to use both Omnibus.

There are some test programs for Warren's FlipChip tester for some of the DW8E boards.
Click to expand...
I've never realized the DW8E had two separate sections. They refer to it as the I/O bus converter and the I/O bus expander. Nifty!
 
The RK8 set that I pulled from the DW8E has an M7104B2, M7105B, and M7106C per the etch marks. The M7104 handle does not have YA on it. I see a single green wire added on the M7105 and many on the M7104. The M7106 appears unmodified. I can get photos soon. Any thoughts as to if these are the -F or a -E set?
 
The M7104 in the PDP-12 at the RICM says YA on the top of the handle. It will be interesting to see if your board set works.

We had lots of issues with the power supply in the DW8E. We have LTspice simulation files for it if you want to learn about how it works.
 
Added more photos to the album.

The power supply is identical to the rear-mounted 8/M or /F supply, correct? I understand there was also an /M and /F supply that was mounted on the side. My 8/M thankfully works fine—hopefully the DW8E will as well!
 
I believe that the same power supply was also used in some 11/03 chassis too. You should reform the large caps before you power it up.
 
And, added more photos to the album. I wouldn't use imgur if it weren't for the fact that the forum doesn't like uploading photos from my phone, as apparently they are too big. I figured the forum software could scale them down accordingly, but I digress.

I took the G813 5V regulator card out of the power supply. It's so funny to me that DEC opted to use the same Flip-Chip form factor for this PCB, complete with gold fingers and handles, despite it not going into a backplane. Only two connections are made to the edge connectors, declared as test points in the schematic. I thought maybe that was for automated testing, but I would've expected more connections than that if so.

Anyways, one of the 150μF 15V tantalum capacitors caught fire and burned part of the board, and seems to have roasted the SCR for the crowbar circuit. This could certainly cause some issues! I cut and filed away the burned portions of the PCB in hopes of getting rid of any potentially conductive carbon. This may be more concerning at higher voltages, but it's better safe than sorry. This also gives me hope that the transformer and resonant capacitor are still good. I'll keep my fingers crossed. I will double-check the rectifier for the 8V since it may have been taxed by the short that was the tantalum cap.

I went ahead and stuck two new radial caps (220μF 50V aluminum) in place of the old ones, but I've ordered a new SCR (C106BG) to replace the 2N4441 and axial capacitors to use instead of the radials.

Soon, I'll get the large caps (2x 160,000μF 20V, 1x 57,000μF 30V) hooked up to a power supply to start reforming them, as Michael has suggested.

Talking with Doug tonight, he reminds me that, while you can try to run this power supply without a load, it will likely trip the crowbar every time. I haven't studied the regulator or crowbar circuit much yet, but I certainly believe his experience with Vince's machine! I have some large power resistors which will work nicely for a load.
 
I added more photos to the album. I'll continue doing that as I take more photos, but I'll also inline some here for added interest.

Tonight was very eventful! I reformed the capacitors and found them to be in good health.

tD7GpUj.jpg


The current was limited to 10 mA, and I monitored it over several hours.

uOwAzVF.jpg


Eventually, the leakage dropped to an amount I deemed reasonable, and figured any more reforming could be done in circuit instead of on the bench.

mqdFlkN.jpg


While the capacitors were out, I did some low input voltage checks of the ferroresonant transformer and rectifiers. To my delight, everything seemed good.

I put the capacitors back in and clipped on test leads to monitor the voltage of the +8V, -15V, and -30V rails.

dpxU3UI.jpg


I found that, at 120 VAC input, I was seeing very reasonable output voltages. In fact, even at 60 VAC, things seemed adequate. This tells me that the ferroresonant transformer is indeed providing some regulation, presumably by saturating the core or something like that. Power electronics were never my specialty. Here are the output voltages at 120 VAC.

eNAAfcB.jpg


Since I was bringing it up slowly with a variac with a Kill-A-Watt on its input, I happened to notice the current clearly was at a local minimum at 114 VAC or so. (My line voltage runs about 123 VAC, so I figure the variac dial will read a little low.) This corresponded with a local maximum of the power factor: 0.81.

08bJ8WK.jpg


Lastly, I checked the ripple on the two negative rails, and it seems reasonable with a very light load of the bleeder resistors.

utYtK5U.jpg


All that is left is to check the pass transistors for the 5V regulator and await the arrival of the SCR. Oh, and I will also replace the power paddle with a new one I got from Jack. My flex cable is in very rough shape. I don't think it'll be much longer before I'm assessing the actual logic of the computer!

Kyle
 
Looks like a Victrola or equivalent in the background. Got one of those also. I've had problems with the Faston/quick disconnect having high resistance so having to clean/re-seat them.
 
I've got a small phonograph collection. A few Edison machines (the Diamond Disc machine is the one you eyed), a Columbia, and an off-brand that I have the original 1922 sales receipt for.

I'm expecting all sorts of issues with the spade connectors! Is there a source for new male spade connectors with the colored plastic shrouds?
 
You must log in or register to reply here.
Back
Top