• Please review our updated Terms and Rules here

PDP-11/44 Restoration

The loop is a fun little program and apparently is good for all of the PDP-11 family, Unibus and Qbus. As long as they all use the same ODT and default addresses for the CON (777560) and the output to the terminal (777566). You may want to look real close at the cable that connects from the backplane to the power supply, not the heavy straps but the little multi conductor cable and its plug along with all the connectors in the power supply itself. Been a long time since that things been working and you may have some issues with corrosion and dirt due to age and that may be causing some of the flakiness with the regulation. It’s important to remember that the power supply samples the voltage on the backplane and there are at least a couple points where that sample goes thru a plug.
Do you have that thing in a rack? Or are you hefting it around on a table top? My 11/34 is in a rack with the slide rails so you can slide the system in or out of the rack and when extended out of the rack it’s on pivots so you can flip it up to get to the undersides without killing your back.
 
This system is in a rack. It pivots upward, but sadly it does not slide out. Luckily I can access the card cage with out having to move it, but accessing the other side of the backplanes is a bit difficult. I will definitely try reseating some cables.

I actually think that the 11/44 uses a different ODT. I tried the L and G commands and they did not work. I had to use D and S to load the program. The 11/44 manual describes it's operation (http://bitsavers.trailing-edge.com/pdf/dec/pdp11/1144/1144_UsersGuide.pdf chapter 2). I do not know the output address for the console and the terminal.
 
L, E, D and S are fairly standard. Think it was my mistake or maybe on the old 11/03 and 23 G was for start and on the newer stuff S is for start. I often get confused with code and software because I am basically a hardware guy.
L should load an address, so L 1000 will load address 1000 and to see the contents of that address you use E for examine, D will deposit whatever you type into that address (in Octal!) and S is for start, not G for go.
Fairly certain that when your keying in code that you should only have to load the first address like 1000 and then every time you deposit a new instruction the address register will increment to the next address so if you type L 1000 (ret) D 005000 (ret) D 100137 (ret) D 1002 you will have entered something like this:

1000 005000
1002 100137
1004 1002

Saves key strokes, E works the same way. L for the starting addresses and every time you enter E the consul will display the next line of the program.
Imagine the software people will come on here and let me know how wrong I am but think that may help when playing around with ODT.
I was so inspired by all this that I built a little web page about PDP-11 loops at:


http://staff.salisbury.edu/~rafantini/PDP11Octaltricks.htm
 
Thanks for the link to your webpage. I am hoping to try examining the console I/O addresses soon.

I tried to turn the PDP on today, but the power light started to flash again. I reseated all of the cards I have in the system and the one ribbon cable connected to the first backplane (I have nothing connected to the second backplane right now). When I tried to run the system again the power light flashed for a few seconds then it started to become steady. After a few minutes the light started to flash again. The console did become unresponsive before the power light started flashing. After waiting a few minutes and trying the system again the console responded, but the light was flashing again. I did take a voltage measurement again today. In the few seconds that the power light was flashing the multimeter read 5.6 volts. After the power light stopped flashing it read 5.5 volts. Should I go back to testing the system without the cards installed again?

EDIt: I have been thinking and I think that the console not responding might be a problem with the terminal. The boot toggle switch made the run light illuminate after it was released. I have also had the terminal not respond when I was using it with a different computer. I guess right now it is a matter of more troubleshooting.
 
Last edited:
I thought I would give another update with the power problem. I can definitely attest to the problem being intermittent. I turned on the PDP-11 a few days ago and the power light stopped blinking a few seconds after I turned the system on. After a few minutes (10-20) of checking to see what boot roms I have and the console I/O addresses the power light started to flash again.

I highly suspect that there is a part that is going in and out of tolerance in the power supply. What makes me suspect this is that one time the power light started flashing and it continued to increase in frequency until it remained solid. I think that whatever voltage is out of tolerance was changing until the part warmed up, causing the power bad signal to start and stop rapidly.

I still need to check a couple of other physical connections to the PSU besides the obvious ones. I also think that the voltage that is out of tolerance is not the 5v, but another one because I checked the 5v and it was 5.4v so it was less than the previous tests. I am still trying to find the other voltages on the backplane.
 
Another update. I was testing the computer again and when I turned it on I received the message ?22 CP HUNG. Then I turned it off for a few seconds and then I turned it back on. The computer did not respond at all, and the LEDs on the memory cards were not on. I immediately turned it off to prevent any damage that may occur.

I removed all cards except for the console interface module, and re-attached the dummy load. At the backplane I was receiving 5.1v again. I find that odd because I am sure that the cards would draw more current than the dummy load would. The whole time the power light was flashing.

Based on these observations I think that I can conclude that the Bias and Interface module and the Logic voltage boards in the power supply are functioning okay. So until I solve the power supply problem I am not going to replace any of the cards to keep from damaging them (save the console interface module so I can see the power light).

On a side note does anybody know the connections for the TU58? I can find the right type of connector, but I need to know which pins of the serial port I should connect it to.
 
I am still studying the technical manual to gain insight into what could be wrong with the power, but I think that the memory voltage is probably fine. the jumper is not set on the CIM to monitor the 12v for the memory. I am beginning to suspect the 15v logic voltage is bad. My first attemt to remedy this will be to reseat the cards in the power supply.
 
Just an odd thought... have you monitored your A/C line voltage during this testing? (preferably at a point inside the chassis)
 
+5 should be 4.9 to 5.1 or 5.2, 15 can be 14.8 to 15.2 just as a rule of thumb. That’s what I would aim for. If the regulated reference voltage, not cetin if its +20 or what that’s on the bias supply is off or changes all the other voltages will be offset but if only one voltage is wrong and the others good it will be that one regulator. Maybe install a couple wires to carry the reference or bias voltage out to where you can see it and also look at it on a scope to be cetin it’s clean. Also remember if your voltages look good that there can always be an issue in the power sensing circuits themselves.
You have been working on this thing forever but think there is a light at the end of the tunnel. Got any kind of drives on that system? And the good news is that after about six months now I finally got my RX drives up and running.
 
When I have the cards in the system without the dummy load I get ~5.5v, which I think is a bit odd given that they should draw the minimum load. With the dummy load I get 5.1v exactly. I am still trying to find the +15v on the backplane again, but when I found it before I remember reading around 15.01v.

There are two bias voltages, -15v and +12v. According to the technical manual the +12v bias powers the relay that bypasses the inrush current limiting resistor. I do not know if it is within specification or not, but I do know that it is present because I can hear that relay activate. I will try to take a look at these voltages.

As for the AC voltage at the PSU, I can definitely safely check it right where it enters the power supply. I am unsure whether I can check it inside the chassis without disassembling it, and I don't want to have it energized with it disassembled because of the high currents it can provide.

The drives I have on my system are 2 RK07 drives and an RC25. I am unsure whether I have the proper controller for the RC25 though. There is also the TU58 drive, but I believe that the capstans are bad.

Congratulations on getting the RX drives working!

EDIT: I have not checked the AC going into the system yet, nor the bias voltage, but I did check the voltages at the backplane. Using the dummy load I drew about 5.4A (0.6A less than the minimum load). The voltages I read were 5.10V, 14.85v, and -15.01v. I have not attempted to check the +5v B and the +12v and -12v. As a side note these voltages were measured where the console interface module is placed. I decided to check it there since I could not find those voltages on the backplane (there are so many pins to check I probably skipped a few accidentally). Maybe the voltage monitoring circuit is bad.
 
Last edited:
I have an update. I tested the voltages again today, and they were all the same as yesterday. I also checked the voltage where it enters the PSU, and it was around 125VAC. I still have not gotten to check the bias voltages though. I checked the +15v and +5v with the console interface module in its slot, due to where it was in the case I could not test the -15v.

I did find another voltage on the backplane today, it was around 3.75v. I have a high suspicion that is suposed to be the +5B voltage. I still have not been able to find the memory voltages.
 
I have checked the voltages at the backplane using table 3-4 in the system technical manual (page 3-4 http://bitsavers.trailing-edge.com/pdf/dec/pdp11/1144/1144_SystemTechMan.pdf). I was able to confirm +5, +15, -15, and the +12v. I still need to check for the -12v.

The voltages that I have not been able to check are the voltages marked with VB. The pin designation does not fit the legend that the manual gives for the other voltages. Any ideas? Like I said earlier, I found a voltage around 2V on the backplane. I have a strong suspicion that this is the +5VB, but I want to be definitely sure that I am not seeing some other signal that the CIM might be sending to the backplane, or another signal from the PSU. Any ideas?
 
Just off the top of my head so I may be wrong but think that’s the battery backup voltage for the memory for when the systems off and you don’t need it. Still have the flashing DC light?
 
I still have the flashing DC light. After some testing for a few minutes it turned solid, but it typically always flashes when I turn the PSU on.

I will look into the battery backup voltage. Is this voltage still generated even though I do not have the battery backup installed in my system?
 
I only had a nodding acquaintance with an 11/44 some years ago and would hardly qualify as a knowledgeable resource, but I admire your persistence in fixing this thing and have a few comments anyway.

When powered up, I believe that the “B” voltages you refer to should always be present since these would be from that portion of the supply that powers memory. These are the same voltages that would be supplied by the battery backup (if you had one) to retain the memory in the event of a power failure. Unlike the other voltages, they should also be present if the keyswitch is in the STD BY position.

It would seem that the blinking DC light is trying to tell you that some voltage is out of tolerance and I don’t think that would have anything to do with whether there was a battery backup option installed or not. It sounds like you already may have done a fairly thorough check of all the basic power supply voltages and you might want to consider that it’s remotely possible that the voltage monitor circuit is setting off false alarms.

Finally, page 3-4 in the manual documents voltages on backplane connector P1 and on the backplane itself. Which place are you attempting to measure from and which ones can’t you find?
 
Thanks for the help. I am trying to find all of the B voltages. I cannot seem to find any of them because the table lists their location differently than the rest of the pins' voltages. Since you mention that it is on connector P1, then I think that is where it is located then. Would you happen to know where I could find those voltages on the backplane?

As for the other voltages, +5, +15, -15, and I think +12 are all within specification. I just cannot find the B voltages on the backplane.
 
Thanks for the help. I am trying to find all of the B voltages. I cannot seem to find any of them because the table lists their location differently than the rest of the pins' voltages. Since you mention that it is on connector P1, then I think that is where it is located then. Would you happen to know where I could find those voltages on the backplane?

As for the other voltages, +5, +15, -15, and I think +12 are all within specification. I just cannot find the B voltages on the backplane.

Ah, looking at that page closely I can see your dilemma. Well, fortunately in order to be UNIBUS compatible, all MS11 memory would have to conform to the same pinouts (check out http://bitsavers.trailing-edge.com/pdf/dec/pdp11/1144/EK-MS11M-UG-001_UsersMan_May79.pdf). In addition, looking at the 11/44 Technical Manual, it appears that slots 9-12 of the processor backplane are intended for memory. So I think that combining those facts would lead to the following assignments:

+5 VB ..... B09D1 - B12D1
+12 VB .... A09R1 - A12R1
-12 VB .... A09S1 - A12S1
 
Thanks for those locations. I never thought to look in the memory manual.

Today I powered on the PSU and it decided to output the proper voltages today. I went ahead and checked the voltages just to be sure, and all of them were within specification.

I did realize that I had the computer connected to power for a while. I am wondering if that reconditioned any of the capacitors that are in use when the PSU is not outputting the current.

I will try testing again later to see if the DC light flashes again.
 
Back
Top