• Please review our updated Terms and Rules here

Reviving the PDP-12 at the RICM

We planned to do some fine tuning in the PDP-12's paper tape reader/punch, but it would only run at about 10% of the normal speed. This was one of the original symptoms. We quickly confirmed the -15VDC power supply in the DW8E Omnibus expansion chassis had failed again. At least this time if failed hard so we could finally track down the intermittent fault. We powered the supply through a Variac so the -15VDC would not go above -18VDC, trip the crowbar, and blow the fuse. We replaced two transistors, Q24 & Q25, in the power supply regulation circuitry. The power supply will now regulate, but can't be adjusted low enough to make only -15VDC. We need to do some more research and finish this repair next week.
 
Today I replaced all of the transistors for the -15V regulator in the DW8E. It behaves differently, but it still does not regulate.
I am going to get a tutorial on how the regulator works, and how to debug it from a colleague, and try it again.
 
I looked at some pictures of an 11/05 that was just donated to the RICM. It has the same power supply as the DW8E. We have two 11/05 and an 11/10 in the warehouse. Hopefully one of the power supplies works and we can use it for comparison. The 11/05 manual has a nice chapter on how the power supply works, including scope images of the switching signals. The schematic looks like a linear, but it is actually a switcher. We have just been using a DVM for debugging, and now it looks like we should have been using a scope to see the switching operation.
 
Mike,

This is a really common power supply - it's used in the 8/M and F, the 11/05, some versions of the 11/35, a couple tape drives, etc. It's usually called out as the "54-09728 Regulator Board". There is an Illustrated Parts Breakout (DEC-5409728-IPB-1) available and it is the basis of the H740D Power Supply (Maintenance manual DEC-11-H740A-A-D). These manuals should be available on the web but might not be on bitsavers. Let me know if you have trouble finding them and I'll post them here.

Jack
 
It is indeed common. It is in both the slim 11/10 and the full height BA11-D based 11/05 I have. Fortunately both worked directly. One had a failed input smoothing capacitor though.

If it didn't switch at all you would either have the -39V at the output which mean that the crowbar would trip or you would have 0V. So it just have to do some kind of switching. Since you replaced all transistors then one obvious next step is to check the zener D4. R26 could be suspicious but I don't think the circuit will switch at all if it is open circuit.
 
The voltage drop across the zener is 6.2V, right on spec.
We disconnected the little electrolytics and made sure that they were OK.
The big electrolytics tested OK when we started working with this chassis.
We need to connect a 'scope and see if it is actually switching.
It could be as simple as a bad resistor.

At least I have lots of spare transistors if the power supply in the newly donated 11/05 doesn't work.
 
Sorry, late to the thread. So, please forgive me if some of the following suggestions have already been made.
ASSUMPTION: That this supply is the same as the one in the 8m with which I am familiar!

1) What do you mean by not regulate? Does the voltage go to zero, drift around, or?
2) You need a load or the circuit won't fuction.
3) Check the wiper on the trim pot used to set the voltage; they can go open circuit.
4) Unlikely, but confirm continuity through the inductor and connections to the PWB.
 
AteEm said:
ASSUMPTION: That this supply is the same as the one in the 8m with which I am familiar!

1) What do you mean by not regulate? Does the voltage go to zero, drift around, or?
Click to expand...

The initial problem was that the DW8E manuals for this chassis don't describe how the power supply works. From the schematic, we though that it was a normal linear power supply. Most of our initial debugging was done the wrong way. We found a detailed description of the power supply in the PDP-11/05 manual. Now that we know it is a switcher we have to redo our debugging.

AteEm said:
2) You need a load or the circuit won't fuction.
Click to expand...

We had a small load on the power supply. We will try a bigger one.
In the DW8E chassis the PC8E is the only board that uses -15V, so the normal load is very small

AteEm said:
3) Check the wiper on the trim pot used to set the voltage; they can go open circuit.
Click to expand...

We did, and it works OK.

AteEm said:
4) Unlikely, but confirm continuity through the inductor and connections to the PWB.
Click to expand...

The inductor and the connections to the PCB are OK. There are some notes in the Field Service Manual about a problem with loose fasteners causing problems.

We got a PDP-11/10 from the warehouse. It has the same regulator board and it works OK. We can compare the behavior of the two power supplies, and now that we know it is a switcher, we should be able to fix it.
 
m_thompson said:
snip...
We got a PDP-11/10 from the warehouse. It has the same regulator board and it works OK. We can compare the behavior of the two power supplies, and now that we know it is a switcher, we should be able to fix it.
Click to expand...

That is EXCELLENT!
Nothing like having a working circuit to compare, point by point, with the faulty one.
Sorry I can't be of more help. Sounds like you've done a thorough job of checking the 'easy' fixes.
But, with a 'scope and the working supply, you should be up and running in no time :)
 
The power supply in the 11/05 shows switcher behavior on the +5 and -15 without any load.

The power supply from the DW8E shows switcher behavior on the +5 with a load, but a steady 5.2V and no switching without load. It is possible that it really is switching, but at a once every two or three seconds rate. We will probably ignore this because it works OK with a small load.

We removed Q24 and were able to turn Q23/Q22 on and off, so those transistors work OK. Last week we noticed that the NTE288 replacement for the XA55 transistor Q26 had the Emitter and Collector leads swapped from the originals. Last week we replaced Q26 with a new part, but didn't see any difference. When I went to put a new Q24 back in Warren suggested that I check the Emitter and Collector lead locations. Warren refers to this process as adult supervision. Of course these NTE287 transistors also had the leads swapped. I put the new Q24 in and replaced Q26 and Q21 that were also in backwards. Now the power supply has a very low output voltage and is not switching. At least it does not trigger the crowbar and blow the fuse.

Warren and Alex will work in the power supply this week, and hopefully solve the mystery.
 
After replacing the smaller transistors again the power supply started working a little. Alex spent a lot of time reforming and exercising the caps, and now the power supply is working OK.

With the help of a power supply expert at work I modeled the -15V power supply using PSPICE. The switching function is very sensitive to the size of the output capacitor and very sensitive to the capacitor's ESR. I will make new models of the power supply in LTspice and put them on the RICM WWW site.

The paper tape reader will again run at full speed, but double steps when you try to read a single character. That behavior is adjustable on the PC8E controller card and will be the project for next week.
 
Alex and I are creating LTspice models of the H740 power supply so we can better understand how it works. The models are working OK, but still need some improvements. This power supply is a lot more complicated than I first thought. We will post the models on the RICM WWW page.
 
We have all three voltages in the spice model of the H740 power supply working. We need to add the voltage OK logic and clean up the DEC part library. We found a program that plots the standard gain traces for a transistor model. We will compare the transistor models to some of the real transistors to make sure that the gain in the model is reasonable for a 45 year old transistor. Once that is done we will post the LTspice model.

We started cleaning the RK05 disk drive. It is very clean and in good condition. It will need new seals between the blower and the card cage, and between the plenum and the disk pack. I think that 1/2" and 1/4" weatherstrip from Home Depot will work fine. The NiCad batteries for emergency head retract are toast. These look like standard 1.2V 2/3AA 400mAh cells. It looks like some cordless phones use the same batteries so I can buy an assembled 4.8V battery pack.
 
I ordered 2 replacement NiCd battery packs, one of the PDP-12 and one for my 8/e. I bought a selection of self adhesive weatherstrip at Home Depot (Bauhaus). You can see the weatherstrip installed on the RK05 on the 8/e here. I will do the same weatherstrip process on the PDP-12 next Saturday.

The K2 Emergency relay on the RK05 on the PDP-12 chatters after the drive has been running for about 10 minutes, May this is due to a lack of cooling, or maybe there is something wrong with the power supplies. Another project for next Saturday.

I ran the DHRKA-B diskless tests on the RK8F in the PDP-12. The RK05F has special timing modifications so it can work in the DW8E chassis. The Maindec fails at test #79. Test #77 & #78 do a data-break to address 0000, and that works OK. Test #79 & #80 do data-breaks to address 7777. These two tests fail. The possible failure modes are: one address bit from the Current Address is not being driven onto the Omnibus. I can modify the test to try addresses with just a single bit on and use field 1 so it doesn't modify the diag code. None of the CA bits are being driven onto the Omnibus. The CA on the Omnibus is not being transferred to the Posibus. There is a broken Posibus receiver in the PDP-12.

Next Saturday I can put the DHRKA-B diag in a scope-loop and look at the Omnibus signals. If they are OK, I will trace the signals back through the Posibus and into the PDP-12. This failure should not be too difficult to debug.

We discussed cleaning the RK05 disk pack. The pack cleaner in the warehouse is for an RL01/RL02, so that won't work. We are thinking of disassembling the pack and cleaning the surface with 97% alcohol. I am not sure if that will make the pack cleaner, or get more dust on it from the environment.
 
What’s the RL pack cleaner look like? On the old RL-01 first generation drives they had a stupid brush that came out and swept across the disk pack before head loading but my limited understanding was that if you came across it you were supposed to disable it and it disappeared by the time the RL-02 was in production. I have a bunch of RL packs along with working drives but would think that unless you have clean room environments disassembling a disc pack can do more harm than good. I have a process now of any time a drive is open to clean it out with a shop vacuum first before installing a pack just to be certain nothing is around to interfere with the disk operation. Think that may be one of the big improvements of the RL over the RK in that it’s a closed loop system and if kept clean will stay clean.
 
Hi All;

QBus, I don't know about RL-01 or RL-02's, but my CDC 'Hawk' Had the 'toothBrush' and it's main function was to move any Large particles of Dirt or Dust from the the Disk Pack..
None of the Drives of this Era had or needed Clean Room air, But they did have a Large Filter and this filtered Air was what Lifted the head or kept the head off of the Surface of the Platter.. Since I don't know for sure whether it actually lifted the Head. I do know that If the Air flow got below a certain amount that I would have a Head Crash.. Usually from a Dirty or clogged Filter..
Most of the Time I was able to catch it in time and Save the Head, but not the Platter.. And so I would take a Platter from a Disk Pack and Replace the Fixed Platter.. The Fixed Platter was the one that Usually went for some reason.. Also the CDC had one fixed Platter and one Removable Platter each held about 5mb, I don't remember if it was Megabit or MegaByte.. I know the difference, I just don't remember what it was..
I think Your Shop Vac is a good Idea..

THANK YOU Marty
 
m_thompson said:
We are thinking of disassembling the pack and cleaning the surface with 97% alcohol. I am not sure if that will make the pack cleaner, or get more dust on it from the environment.
Click to expand...

Environmental dust on 2.5 meg packs isn't a problem, anything will get blown off during powerup.
The heads fly pretty high on this class of disk, and the heads and platters can survive a lot.

You want to completely inspect the surfaces for any 'tar' spots, and get rid of them with isopropyl, and do the same for the heads.
I've cleaned over 100 packs this way.
 
m_thompson said:
We discussed cleaning the RK05 disk pack. The pack cleaner in the warehouse is for an RL01/RL02, so that won't work. We are thinking of disassembling the pack and cleaning the surface with 97% alcohol. I am not sure if that will make the pack cleaner, or get more dust on it from the environment.
Click to expand...

RetroHacker_ and I have had excellent results with hand-cleaning RL01 and RL02 packs, which should have tighter tolerances. We did it in RetroHacker_'s kitchen after a little prep. Make sure you wear low-lint clothing. You will need:

Gloves (we used nitrile, latex is probably fine -- your hands will be EXTREMELY dry if you don't use gloves, and will likely crack/bleed/burn from the alcohol)
99% isopropyl alcohol (higher is better, get at least technical grade, don't use drugstore 91%)
Kimwipes or other lint-free wipes
Low lint cotton swabs with long wooden stick
Bright flashlight (we used a HexBright which IIRC is 150 lumens, LED)
Tools for disassembly

Additionally, I brought a wash bottle for the alcohol which made it much easier to dispense with less worry of dilution from humidity. Get the vented type, the alcohol will push itself out otherwise. We also used Goo Gone to remove stubborn sticker residue from the outside of the packs.

All of the consumables are available from McMaster-Carr. You don't need to get anhydrous alcohol, but the 99% is significantly better than 91% or 97%. It leaves *no* residue and its solvent action on gunk for cleaning disk heads is noticeably better.

Our work surface was a Formica countertop. We sprayed it down with Windex and wiped it down with a regular paper towel. Then we wetted a Kimwipe with alcohol and wiped it down again.

Process:

Clean the outside with soap/water, then alcohol. Pop it open in a relatively dust free space (wipe your work surface and tools down first). Inspect for signs of head crash. Roll a Kimwipe around the wooden stick of one of the swabs, holding the end closest to the swab's head to keep it from unrolling. Soak it with alcohol from the wash bottle. Hold it across the surface of the platter, angled or straight doesn't matter as long as you can get the whole platter in one swipe. Rotate the platter under the wipe (this worked best with the RL01/RL02 packs vertical). After the alcohol has evaporated (seconds), take your bright light and keeping it at an acute angle with the platter, look for stuck on bits on the surface. If you find any persistent bits, wet the tip of a swab with alcohol and remove them, then rewipe the platter. Using the corner of a Kimwipe with a little alcohol, pick any stray bits of dust off the platter. Close it up and set it aside, it is now clean.

I need to actually write this process up at some point. We took pictures of the various steps.
 
Some years back I was involved with an Amateur Radio Group AMSAT and was involved in the production of amateur radio satellites. I was introduced into the concepts of “clean room” operations at that time and have tried to keep separate work arias clean and free from contaminations. No, don’t think you need a high quality filtered air clean room environment for working on media but can’t see where it would hurt!
It’s always better to think in terms of at least a clean work surface and clean tools when doing this stuff as opposed to just popping the pack open on the same work bench that you use to clean out power supplies, drill holes and file down burs.
With a small effort you can clean up a work space and maybe get down to the 5 to 10 Micron level of contamination for sensitive work but I have seen people just pop this stuff open before on the same work bench they use for everything else without giving it a thought.
I know the vacuum exhaust in the room but if you buy the cheap HEPPA filters they are good down to about 2 to 5 microns or so.
I don’t know how with every RL drive that I have up and running I find contaminations when looking at them up close. I cleaned them insides and out with towels and alcohol till I found no dirt but still from time to time small parts of the paper towels or other debris in the drives. I will admit that maybe I am a bit obsessive about keeping the packs and drives as clean as possible but it’s not like they are making any more of them and I want to try to keep the internal air filters, packs and heads in as good an environment as possible.
Have been thinking of bringing in a raised floor and maybe separating off a part of the shop just for the computers but have not reached that level yet.
 
You must log in or register to reply here.
Back
Top