[back to collection]

Bare Card Cage #1

(last updated: 21-Sep-2008)

Operational Status

Configuration

The current configuration is in its original 'as acquired' condition.

Major Events

Still To Do

Description

Acquisition

20-Feb-2006

This unit was part of a large pickup in Longmont, Colorado.

Cleanup

12-Mar-2006

Very dusty! Interestingly, this cage has very long (3') wires/cables attached for the power. Obviously this cage went into some special cabinet. This may be an OEM system. The cage has 4 slots.

The CPU board has dust caked on that will not vacuum away!

I removed the cards, vacuumed each one, photo'd each one, and then the backplane. I reinserted the cards and re-wrapped the backplane in foil.

The boards were photographed in sequence from slot 1AB to slot 4CD.

Operational checkout

21-Sep-2006

Fortunately, Dworkin included a nice little Sigma Information Systems power supply in all of the pickup materials. It has the +5v and +12v supplies that the backplane needs, so it should work fine, as long as it can carry the load. It looks up to the job, though there is no label indicating the ratings.

So, upon review of the cage, I have a RX01 floppy controller, some memory, a DLV11 SLU, and the terminator/boot board. This last, I suspect, is in the wrong position. I believe that it needs to be the last card in the cage. I'll keep that in mind as I go along.

On other thought is that I might want to remove two of the three memory cards, just to simplify the whole thing, but I don't know right now which is addressed where, so I'll probably leave them all for now.

First thing, then, is to find a suitable cable for the DLV11 SLU. I could steal one from my Darien 11/23+, but I hope not to do that. I now realize it may not be as simple as connecting at VT100. This SLU board can do 20ma current loop connections to. I'd better check it carefully for its configuration (and the other boards).

Well, well, this is an interesting configuration. For one thing, the terminator board is not at the end of the bus. Second, the memory boards are not in adjacent memory banks, which seems strange to me. Sure I can imagine that someone might have software that does that, but note that there are not any odd-numbered memory banks. There are 0, 2, 4, and 8. So there is a 4KW gap between each bank for some reason. Hmm. Plus, all memory boards have refresh reply enabled. I wonder if this cage ever was used this way, or were the boards just popped in any old order when it came time to give it away?

Undoubtedly I will need to reorganize the board locations, and change the memory bank and refreash reply configurations. I think I'll remove all three extra memory boards for the first power up. Now that I know what's addressed where, I can safely remove them, I think, without creating any problems. The CPU board has memory at block 0, and that's all that I need.

So, for my first power up, I'll use: 

- M7264 : LSI-11 CPU
- M7940 : DLV11 Async SLU
- M9400-YA : Bus terminator, boot ROM, memory refresh

I don't know if the CPU board with refresh enabled will conflict with the M9400 with it's own memory refresh enabled. We'll find out.

Now, gotta get a cable from the DLV11 to something...perhaps I should hook up a VT100, since that's a known qauantity, something that I know works. I'd love to hook up a DECwriter or the ASR33, but I haven't cleaned them up yet, or have knowledge of their functional state at this point.

Perparations for power up

27-Sep-2006

During the past few days I have cleaned and checked out the LA36 DECwriter II, which uses a 20ma serial interface as does this system. With that unit now functional, I have a suitable device to act as the console for this system. Luckily, it included the 20ma cable, which should allow direct connection with the DLV11 in this cage.

I have powered up the Sigma Information Systems power supply and checked the voltages. The +5V reads at 6.02V, so that's high. The +12V reads at 14.45V, and that's high, too. I don't know if this is due to the no-load configuration at the moment, or if the voltages really are out of line and need adjustment. It is a switching style of power supply, so it shouldn't depend on a load. The power supply does have some adjusting potentiometers, so I can make those adjustments.

Reviewing the LSI Bus specification in the Microcompuers And Memories 1981 handbook (pg. 360), it states that the +5V has a tolerance of +/-5%, and the +12V has a tolerance of +/-3%. So the values I'm reading from the power supply are way out of range.

I left the power supply on for about ten minutes, just so that it could settle in, and made the adjustements. Unfortunately, I can only go so far. I move both the +5V adjustements to eheir stops. The +5V now reads 5.33V, and the +12V reads 13.97V. So I can get the +5V into spec, but not the +12V.

(I also noted that there is an empty AGC fuse clip on the power supply PCB. After a little probing, I see that it protects the +24V line, which I wasn't getting any readings from. There is no indication of amperage the fuse should be, and since it serves only the +24V line I believe I can proceed without any action on it.)

I removed all of the boards from the backplace, except the CPU, SLU, and Termination boards. With the serpentine configuration of the backplane (1AB->1CD->2CD->2AB->...), I believe that this is the correct configuration:

1- M7264 : KD11-F : LSI-11 CPU w/ 4KW RAM  ->
2- M9400-YA : REV11-A : 120-ohm term.,     | M7940 : DLV11 : Serial Line Unit

Operational checkout (cont.)

29-Sep-2006

I wanted to confirm, once and for all, what was the bus grant chain flow in this backplane. Looking at some past posting on classiccmp.org, I see that Allison states that the backplane has this order:

    A  B C  D
1- |--->|---v|
2- |v---|<---|
3- |--->|---v|
   (v--------)
4- |--->|---.|

But this seems kind of strange...why swap positions 7 and 8? In the end I removed all boards and cables from the cage/backplane, and traced the bus grant lines myself. I now can confirm, at least on this specimen, that the order is:

    A  B C  D
1- |--->|---v|
2- |v---|<---|
3- |--->|---v|
4- |.---|<---|

So, I should put the boards in this order, after all:

1- M7264 : KD11-F : LSI-11 CPU w/ 4KW RAM  ->
2- M9400-YA : REV11-A : 120-ohm term.,     | M7940 : DLV11 : Serial Line Unit

Based on this now known correct configuration, here are the things that deserve checks: - is the LA36 communciations channel working? Can I try a loopback to itself? - is there a problem with both the CPU and the Terminator board doing Refresh? - is the DLV11 working at all? How can I test this? In another system? Loopback? - or, is the Boot Mode jumper on the CPU a problem?

I've kind of convinced myself that it might be the last. There isn't any device to boot from, and perhaps it never drops to ODT, but just loops looking for the device. It might be better to set the Boot Mode to 1, which is directly to ODT. Unfortunately, someone used a soldering iron to make the jumper connection on W6. I hate to take my soldering iron to the board, not knowing if that's the real problem, but at least I can wire-wrap the connections the way I want them, without resorting to solder.

Well, I did it. But, no change in behavior! Too bad! I found some more informatoin on the web, about how the backplane my not require additional termination unless the loads are high, so I pulled the REV11 out just to see if that made any difference. No, it didn't.

I'm a little nervous still, about the high 12V line. It's still reading at 13.8V or so. This might make the chips stall by saturating some switching componenents. I'll try to find a better supply.

Power supply checkout

03-Oct-2006

Turns out that the battery in my multimeter was going dead, and I was getting erroneous readings from it. After replacing the battery in the multimeter I double-checked the output voltages on the Sigma power supply. Sure enough, they were not correct. I had the 5V too low, at about 4.3V and the 12V line was also low at about 11.5V. So, now that these voltages are re-adjusted to their correct values, I might try to power up the cage again.

Well, I tried to boot again, with only the CPU and DLV11 boards in the system, connected to the LA36 DECwriter II. No go. I did confirm that the voltages are now in the proper ranges. I just hope that when I tried the system with the voltages being too low, that it didn't damage any of the electronics. I can't imagine that it could, but I don't know...

Anyway, I'm trying again with the Terminator board installed, too. Nope, no difference.

I think the next thing to try is to replace the DLV11 that came in the cage, with a spare that I have from eBay. I want to leave the original DLV11 as it is for now, so I'll use my spare to connect a known working RS232/EIA terminal. Hopefully, that will reduce the number of variables that I'm dealing with.

Well, I reconfigured the DLV11-J from eBay, and tried it, using my PC with HyperTerm as the terminal. Nothing. No output seen. So I still don't know where the trouble is. Could be this new board, too! For the record, I tried with and without a null-modem adapter, and I tried both port 0 and 3 in both configurations. I should be seeing output on port 3 at 9600 baud, but I'm not.

DLV11 checkout

04-Sep-2006

Trying to reduce the number of variables, I thought to put the DLV11 20ma configured board in the 4' PDP-11/23 system. More details in the Longmont LA36 DECwriter II notes, but the short story is that the DLV11 and the connected LA36 work just fine! So, the trouble lies with the cage itself, the backplane, the power supply, or the CPU board.

[back to collection] [top]

Copyright © 2006-2025 Jared Blaser. All rights reserved.