• Please review our updated Terms and Rules here

Restoring a DEC PDP-11/05

wh3016

Experienced Member
Joined
Jul 20, 2022
Messages
50
Location
Massachusetts
I have attached a screen capture above. It is the same signals as the two 74175. Inputs to the ROM is the output of the MPC. The outputs of the ROMs are the inputs of the MPC.
Is that CONF? I’ll take a look today. I don’t think the screenshot posted.

Edit: Going off of drawing CONF, for E101 pin 12 (MPC06) is at 0.33 volts. Pins 5,4,13 (MPC 07, 05, 04) read 3.30 volts

For E111, the MPC 0,1,2,3 inputs read 3.30 volts

Using the KM, holding down the M Clk switch produces a clock signal on pin 9 for E101 and E111. When the switch is released, there is a steady 3.60 volt signal present. I noticed that the M Clk pulse switch only works when the M Clk Enable switch is flipped opposite the arrow on the KM-1 overlay.
 
Last edited:

MattisLind

Veteran Member
Joined
Sep 30, 2013
Messages
1,006
Location
Stockholm, Sweden
That basically confirms that the KM11 shows the right info.

But it is not the correct address as I mentioned earlier.

So now you have to figure out why.

The problem is that the MPC inputs are wire-ored with at least one more ROM that is used for microbranching.

Check the CPU inputs of the other PROM and try to figure out if that one is enabled somehow.
 

wh3016

Experienced Member
Joined
Jul 20, 2022
Messages
50
Location
Massachusetts
That basically confirms that the KM11 shows the right info.

But it is not the correct address as I mentioned earlier.

So now you have to figure out why.

The problem is that the MPC inputs are wire-ored with at least one more ROM that is used for microbranching.

Check the CPU inputs of the other PROM and try to figure out if that one is enabled somehow.
By “other PROM(s)” are you referring to E102 and E112- the next instruction high and low bits? And if so, what inputs/outputs should I be looking for that indicate the PROM is enabled?
 

MattisLind

Veteran Member
Joined
Sep 30, 2013
Messages
1,006
Location
Stockholm, Sweden
No. Check sheet CONE. E108 and E109 PROM or in to the MPC bus. There are also a bunch of simple gates that wire OR with the MPC bus. These chips has to be checked to see if these make havoc with the MPC bus.
 

wh3016

Experienced Member
Joined
Jul 20, 2022
Messages
50
Location
Massachusetts
No. Check sheet CONE. E108 and E109 PROM or in to the MPC bus. There are also a bunch of simple gates that wire OR with the MPC bus. These chips has to be checked to see if these make havoc with the MPC bus.
Here are the read outs I took today with the DMM:
(from sheet CONE)

E108:
1- 3.30 2- 3.30 3- 3.30 4- 3.30
5- 3.30 6- 3.45 7- 3.45

9- 3.30 10- 3.84 11- 3.54 12- 0.26 13- 0.24 14- 0.24 15- 3.90

E109 isn’t referenced on CONE. Did you mean E107? The way they wrote the seven looked like a nine to me at a glance.

If so, E107;

1- 3.45 2- 3.86 3- 3.86 4- 0.26
5- 3.87 6- 3.87 7- 3.84

9- 3.30 10- 3.30 11- 3.30 12- 3.30 13- 0.25 14- 3.88 15- 3.85
 

MattisLind

Veteran Member
Joined
Sep 30, 2013
Messages
1,006
Location
Stockholm, Sweden
You are right. It is a E107. Not E109. But in any case what is of interest is the enable signals. Not the rest.

E107 has pin 13 and 14 which both has to be low to enable the outputs. They are not both low. I.e. not enabled.
E108 has pin 15 as active low output enable signal. It is high so E108 is not enabled either.


Wire-ORing is hard to debug. It can be done using a current tracer probe. In this way it can be detected which output sinks a line.

In this case the address should be 241 or 10100001 which since it is active low corresponds to L H L H H H H L. But we are seeing H L H H H H H H.

I am not sure if you checked that the enable signals on E102 and E112 is actually low all of them? Pins 13 and 14.

Since MPC7, MPC5 and MPC0 are high and if those PROMS are in fact enabled my guess is that the PROM itself is bad. But it hard to tell. And it isn't easier since some ROMs take the active low outputs from the MPC and some ROMs take the active high outputs of the MPC. Probably because of fanout.

Perhaps desolder the PROM and try to verify the contents in a PROM programmer or other setup (Arduino).
 

wh3016

Experienced Member
Joined
Jul 20, 2022
Messages
50
Location
Massachusetts
You are right. It is a E107. Not E109. But in any case what is of interest is the enable signals. Not the rest.

E107 has pin 13 and 14 which both has to be low to enable the outputs. They are not both low. I.e. not enabled.
E108 has pin 15 as active low output enable signal. It is high so E108 is not enabled either.


Wire-ORing is hard to debug. It can be done using a current tracer probe. In this way it can be detected which output sinks a line.

In this case the address should be 241 or 10100001 which since it is active low corresponds to L H L H H H H L. But we are seeing H L H H H H H H.

I am not sure if you checked that the enable signals on E102 and E112 is actually low all of them? Pins 13 and 14.

Since MPC7, MPC5 and MPC0 are high and if those PROMS are in fact enabled my guess is that the PROM itself is bad. But it hard to tell. And it isn't easier since some ROMs take the active low outputs from the MPC and some ROMs take the active high outputs of the MPC. Probably because of fanout.

Perhaps desolder the PROM and try to verify the contents in a PROM programmer or other setup (Arduino).
Would failed PROMs also be the reason I can’t seem to single step anything with the M CLK switch?
 

MattisLind

Veteran Member
Joined
Sep 30, 2013
Messages
1,006
Location
Stockholm, Sweden
Yes. A failed Next address ROM makes the sequence of microstep completely undefined.

You need to check the inputs and outputs of the ROM again and try to understand if it makes sense, comparing with the micro instruction flow in the schematic.

The solder one of the NEXT address PROM at a time and read out the contents. Check what value is on the MPC bus after removing them. It shoud be all high since nothing else should be gated onto the MPC bus at this point.
 

MattisLind

Veteran Member
Joined
Sep 30, 2013
Messages
1,006
Location
Stockholm, Sweden
When starting up the machine in manual mode what is the input address of the various PROMs? I.e. What is the output of the MPC register after RESET. It should all low on the active high outputs and all low on the active liw outputs. It could be a good idea to check this before desoldering anything.
 

wh3016

Experienced Member
Joined
Jul 20, 2022
Messages
50
Location
Massachusetts
When starting up the machine in manual mode what is the input address of the various PROMs? I.e. What is the output of the MPC register after RESET. It should all low on the active high outputs and all low on the active liw outputs. It could be a good idea to check this before desoldering anything.
I took a look at the PROM inputs on 102 and 112. High inputs all read high and low inputs all read low.

Back to 102 and 112; Enable pin 13 is low and Enable pin 14 is high on both.

I’ll take a look at the other PROMs tomorrow, as I’d like to double check everything before I break out the soldering iron.
 

wh3016

Experienced Member
Joined
Jul 20, 2022
Messages
50
Location
Massachusetts
The plot thickens(?):

Through the generosity of another user, I was able to procure another M7260 and M7261.

Here’s the peculiar bit-

After swapping out the boards, I powered the system up and was greeted with the same blank front panel. I’ll admit, I had hopes of seeing at least something. I haven’t taken a look with my scope yet, so that remains to be seen. The KM11 is still showing the invalid MPC as the last time.

However, this got me thinking- and hopefully something who’s worked with more than one 11/05 might be able to chime in,

As @MattisLind mentioned, the microprogram PROMs and associated circuitry are wire ‘or’ed’ with each other, so there are multiple possibilities that could screw up the microprogram steps. Assuming my KM11 is working correctly, what is the likelihood two different 7261’s can give the same invalid address?
 

MattisLind

Veteran Member
Joined
Sep 30, 2013
Messages
1,006
Location
Stockholm, Sweden
I am not sure if this has been mentioned: What about the status of AC LO and DC LO signals? A lot things need them to be high to work. Check those!
 

gnupublic

Experienced Member
Joined
Jan 23, 2019
Messages
129
Location
Germany, Berlin
Did you check that all voltages come to the boards? In one 11/05 I had a backplane with some blown wirewraps. There must have been a shortage once.
 

wh3016

Experienced Member
Joined
Jul 20, 2022
Messages
50
Location
Massachusetts
I am not sure if this has been mentioned: What about the status of AC LO and DC LO signals? A lot things need them to be high to work. Check those!
Are you referring to the AC LO and DC LO that originate on the power supply? I don’t think I ever recall seeing anything high on either when I was working on it. I hope my memory serves me correct. I’ll start it up later on today and report back.

Did you check that all voltages come to the boards? In one 11/05 I had a backplane with some blown wirewraps. There must have been a shortage once.

I haven’t checked the new set of boards yet, but on the original set, I checked several ICs and verified that they were receiving 5v +/-25%.
 

MattisLind

Veteran Member
Joined
Sep 30, 2013
Messages
1,006
Location
Stockholm, Sweden
Are you referring to the AC LO and DC LO that originate on the power supply? I don’t think I ever recall seeing anything high on either when I was working on it. I hope my memory serves me correct. I’ll start it up later on today and report back.
Yes. AC LO L signal is active low. So high when AC OK. Same goes for DC LO L. They both originate in the PSU. In one variant of the PSU there are a couple of Ge-transistors creating this signal. In others there is a FET doing the same work. Check the signal both in the PSU and in the backplane. The signal is OC of course so there has to be a pullup somewhere. I think the M930 module has the PU for those signals.
 

wh3016

Experienced Member
Joined
Jul 20, 2022
Messages
50
Location
Massachusetts
Measuring right off the backplane, DC LO reads 3.5 volts. AC LO reads 4.11 volts.

From the connector on the power supply, pins 3 and 6, BUS AC LO L reads 0.00 volts and BUS DC LO L reads 0.00 volts.
 

MattisLind

Veteran Member
Joined
Sep 30, 2013
Messages
1,006
Location
Stockholm, Sweden
Aha. That is indeed suspicous. These signals are OC so it looks like the signal path is broken somehow. It also means that we really do not not the actual value since there has to be a pull up on these lines.

A high signal is good but these signals are also used to reset things after the voltages stabilized. This could lead to strange behaviours.

You need to trace the signal and find where the chain is broken.
 

wh3016

Experienced Member
Joined
Jul 20, 2022
Messages
50
Location
Massachusetts
Aha. That is indeed suspicous. These signals are OC so it looks like the signal path is broken somehow. It also means that we really do not not the actual value since there has to be a pull up on these lines.

A high signal is good but these signals are also used to reset things after the voltages stabilized. This could lead to strange behaviours.

You need to trace the signal and find where the chain is broken.
Could this be the reason why nothing appears to be working? I found it odd that I’m getting measurements where the DC AC LO connect to the backplane and not on the power supply itself.

Speaking of the backplane, I couldn’t help but notice there appears to be some non-factory wiring on some wire wrap pins. I don’t know if this was done as a repair or modification. (If the position of the backplane looks odd, I shifted it over a few positions so I have room to connect IC test clips to the M7260)
 

Attachments

  • 2445D550-DD16-4BA1-A539-E80E3BC4B169.jpeg
    2445D550-DD16-4BA1-A539-E80E3BC4B169.jpeg
    3.3 MB · Views: 15

MattisLind

Veteran Member
Joined
Sep 30, 2013
Messages
1,006
Location
Stockholm, Sweden
Could this be the reason why nothing appears to be working? I found it odd that I’m getting measurements where the DC AC LO connect to the backplane and not on the power supply itself.

Speaking of the backplane, I couldn’t help but notice there appears to be some non-factory wiring on some wire wrap pins. I don’t know if this was done as a repair or modification. (If the position of the backplane looks odd, I shifted it over a few positions so I have room to connect IC test clips to the M7260)
Yes. It could be a reason for it not working. You have pullups on the M930. That is why they are high in the backplane. You somehow need to figure out where those signals is cut.

DEC used purple for DC LO and yellow for AC LO. Check the continuity from the PSU for those wires.

It can be that the backplane somehow is modified. It is hard to tell.
 
Top