Tektronix 4052 intensity issue

[nikola-wan]

What other options do you think you have? There is another sticker inside the computer - on my 4054 monitor that lists options installed at the factory.

 

[nikola-wan]

In particular - have you taken the ROM backpack off your 4052? One of the options is an Extended Memory board - which provides RAMDISK performance. It was only available on 4-slot ROM backpacks. Only available with the 4-slot ROM backpack.

 

[nikola-wan]

Cool.

Have you taken the 4-slot backpack off your 4052? There is an option 27 or 28 with an Extended Memory File Manager (RAMDISK with 4907 file commands) that would be visible when you remove the backpack.

 

[forty-2]

I don't believe I have any other options, the sticker inside only mentions option 3, and option 24 (64k RAM, which must have been removed at some point?)
I just removed the backpack, disassembled it and reassembled it, see photos below.

CALL "RATE" does work, but I'm guessing that's a part of the internal ROM, since there isn't an extra ROM associated with the Communication Board.

On reassembling, I still don't seem to get any calls from RAM Packs, and any Pack installed in the first slot (the one without the key) causes the machine to hang as before.

 

[nikola-wan]

Great backpack photos! You are correct - no Extended Memory board in the backpack.

I believe the COMM option ROM is now on the MAS board in the bankswitch slot 0 instead of being located in the ROM backpack.

If you get a 4052 Diagnostic ROM Pack from Jos - you can run the 4052 ROM Checksum program and see if you have the latest 4052/4054 v5.1 ROMs. The other way to check is if the Patch ROM sockets are empty on the MAS board.

Do you have a photo of your MAS board? That photo would indicate whether the socketed optional DRAM bank is populated or empty.

It could be a bad DRAM in the optional bank is causing the system to only detect 32KB of DRAM.

Jos' 4052 Diagnostic ROM Pack is able to pinpoint bad DRAMs. I've used it to repair several 4052 and 4054 computers.

 

[forty-2]

I reached out to Jos to get a diagnostic pack, I just hope it works given my lack of success with any of those slots in the Backpack.

I burned new EPROMs to deal with the ROM rot, and those were the v5.1 images that Jos shared somewhere.
Looking at the pics from the last time I had the board out (to swap the ROMs) it does look like I only have 32k installed.

I'll start looking for replacements, but I think I'm going to wait to poke at this any more until I get that diagnostic pack.
oh, and I did hear back from Tek museum, they're going to look into those microfiche scans tomorrow - thanks for the suggestion!

 

[forty-2]

I also found a picture of the anti-static strap I improvised to handle the boards (I've since bought a real one)

 

[nikola-wan]

I reached out to Jos to get a diagnostic pack, I just hope it works given my lack of success with any of those slots in the Backpack.

I burned new EPROMs to deal with the ROM rot, and those were the v5.1 images that Jos shared somewhere.
Looking at the pics from the last time I had the board out (to swap the ROMs) it does look like I only have 32k installed.

I'll start looking for replacements, but I think I'm going to wait to poke at this any more until I get that diagnostic pack.
oh, and I did hear back from Tek museum, they're going to look into those microfiche scans tomorrow - thanks for the suggestion!

I think we can try to diagnose whether backpack slots 51, 61 or 71 work.
Try just plugging in the R11 ROM Pack into slot 51 with the other backpack slots empy and turn on the 4052.

CALL "LHOME" with no parameters is an R11 Character&Symbol ROM Pack command
CALL "LETTER","Hello World" should draw Hello World in vector text font.

Having Slot 41 not working will be a slight issue for the Diagnostic ROM Pack - but you will still be able to use it in a different slot by typing commands manually. For example - my 4052 Checksum program uses the CALL "CRC" command in the Diagnostic ROM Pack and will be able to checksum all the BASIC system ROMs, Communication ROMs and ROM Packs that are plugged in.

 

[forty-2]

Well this is getting weird.
With the R11 ROM Pack in slots 51, 61, or 71, calling "LHOME" or "LETTER" gives me
CALL NAME INVALID IN IMMEDIATE LINE - MESSAGE NUMBER 32

R11 is the only card that will not lock up in slot 41, but running CALL "LETTER","HELLO" does nothing (just get the wait light for a moment)

Running CALL "LHOME" Gets:
SYSTEM ERROR - 87FE 27 4C 7FE9 7FEF A0 0000 6A

All other cards still lock up in slot 41, except for R11, and an empty card.

This feels like a few problems, one of which may be bad cards, and maybe some bad RAM? And then something else entirely with slots 51, 61, 71...
Jos is going to see about getting me a diagnostic card, but it might not be until September.
I somehow don't have any 2732A EPROMS in stock, or I'd just burn a new card and test that theory. I may take my chances and order some from China, because the domestic sellers prices are laughably high.

 

[forty-2]

As for the slots - the only difference between them is the pair of chip select lines, right? and those are coming from U175 a 74138, which is driven by U251 the MC6821 PIA Controller. I'm tempted to desolder both and socket them so I can swap in replacements.

 

[nikola-wan]

It does sound like bankswitch decoding is not working right. That could explain why slot 41 is behaving differently than the other three.

The R11 in 41 System error for CALL "LHOME" sounds like backpack addressing is not working right.

If you unzip my latest FlashDrive files:
https://github.com/mmcgraw74/Tektronix-4050-GPIB-Flash-Drive/blob/master/FlashDrive27Feb2023.zip

and look at my 4052 CRC program (File 4) in the Root folder, the DATA statements starting at line 131 contain a string of ASCII Hex characters for each ROM or Option ROM slot that is checked. Line 500 contains the Diagnostic ROM Pack CALL "CRC" with A$ as the input parameter.

Jos has posted on his ftp site the Tektronix Diagnostic ROM Pack manual that describes all the commands and Appendix B-2 contains the CALL "CRC" parameter description:



The "tt" bits select particular "banks" and with "zz"=00 the address bits select the specified ROM space.

The "tt" table is incomplete as each of the four option slots could contain a 4050E01 (or Jos' Multifunction 8-slot ROM Pack) as you can see in my DATA statements for Left and Right Expanders. Since you are the first person I've found with a 4-slot backpack, we will need to add slot 61 and 71 to the data statements to check all four slots and possible ROM Expander slots.

If you put a logic analyzer on the bankswitch output pins - during power up you should see all the combinations of bankswitch numbers plus the additional combinations for slots 61 and 71. If I'm not mistaken the option ROM bankswitch decoding is done in the backpack. If you follow the bankswitch logic to the four option slots - each one should have a pin connected to the bankswitch to enable the ROM Pack address decoder. Somewhere is an issue with your backpack.

 

[nikola-wan]

A command to try with the 4052 File Manager is:
CALL"SETTIM","9-AUG-23 9:30"
or
CALL "TIME",A$

Since you may not have a 4907 connected - these commands will likely show an error - no 4907 or GPIB error.
But the CALL might work and not cause a System Error.

 

[jdreesen]

I reached out to Jos to get a diagnostic pack, I just hope it works given my lack of success with any of those slots in the Backpack.

The DRP you'll receive is tested, so if it does not work at your end you'll have to debug the 4052 end.

I burned new EPROMs to deal with the ROM rot, and those were the v5.1 images that Jos shared somewhere.

Important : which images exactly ? I posted 2 versions, the original ones ( can be found on bitsavers ) and a transposed one, intended for wsi57c49 EPROMS ( slightly dirrefent pinout) . You are using MCM68766, hence you will need the original set.

Looking at the pics from the last time I had the board out (to swap the ROMs) it does look like I only have 32k installed.

Indeed. Just add 4116's when all the rest ist functional !

Jos

 

[forty-2]

I just hooked up my logic analyzer. Each slot has 2 bank switch lines, and in slot 41, 51 & 71, one of them isn't doing anything.
A clue!
I'm going to socket U175 and swap out the 74LS138.

 

[forty-2]

Important : which images exactly ? I posted 2 versions, the original ones ( can be found on bitsavers ) and a transposed one, intended for wsi57c49 EPROMS ( slightly dirrefent pinout) . You are using MCM68766, hence you will need the original set.

I'm pretty sure I used the original ones (I didn't even know about the WSI57C49 alternative). Presumably I wouldn't have gotten this far with the wrong images?

 

[jdreesen]

I had not read the whole thread, sorry. But yes, you are right w.r.t. the images.

 

[forty-2]

The DRP you'll receive is tested, so if it does not work at your end you'll have to debug the 4052 end.

Jos - Oh yeah, I didn't mean the DRP was going to be bad, I'm just worried that if I've got bad slots, I can't use it.

A command to try with the 4052 File Manager is:
CALL"SETTIM","9-AUG-23 9:30"
or
CALL "TIME",A$

Monty - the file manager is another ROM Pack, ya? Calling those with my machine in it's current state gives the same "CALL NAME INVALID" error.

I did some more shots with the logic analyzer (an HP 16500B). Here's power-on. It takes about a half a second to see data on the bank switch lines, but it doesn't look like I'd expect, in terms of them all being in sync (the ones that work). It looks the same at higher sample rates.
The inactive lines also don't correlate with one of the 74138 select lines being bad.

I'm worried that without a working backpack, I'm not going to be able to check the RAM. I'm also not sure what to suspect next, besides the MC6821PIA. I do have one I pulled out of a Tektronix 7D01 (in all it's beautiful purple ceramic glory)... worth swapping it out?

 

[nikola-wan]

Jos - Oh yeah, I didn't mean the DRP was going to be bad, I'm just worried that if I've got bad slots, I can't use it.




Monty - the file manager is another ROM Pack, ya? Calling those with my machine in it's current state gives the same "CALL NAME INVALID" error.

I did some more shots with the logic analyzer (an HP 16500B). Here's power-on. It takes about a half a second to see data on the bank switch lines, but it doesn't look like I'd expect, in terms of them all being in sync (the ones that work). It looks the same at higher sample rates.
The inactive lines also don't correlate with one of the 74138 select lines being bad.

I'm worried that without a working backpack, I'm not going to be able to check the RAM. I'm also not sure what to suspect next, besides the MC6821PIA. I do have one I pulled out of a Tektronix 7D01 (in all it's beautiful purple ceramic glory)... worth swapping it out?

I'm not sure where you are measuring the backpack bankswitch signals.

The Option 3 Communication Board contains the backpack ROM bankswitch signals (since you have Option 3 COMM backpack):

This is pdf page 439 in the 4054/4054A Parts and Schematics: http://dvq.com/tektronix/Tek_070-2839-03_4054-4054A_parts_and_schematics.pdf



These Bankswitch signals are connected to pin A22 of each of the four option slots on the backpack following this table in the 4052/4054A Technical Data service manual:

Slot 41 = BS20
Slot 51 = BS28
Slot 61 = BS00
Slot 71 = BS08

Notice that the ROM Expander Bankswitch BSX addresses are NOT THE SAME for those four slots - as they are latched on a ROM Expander to select Expander slots 1-8, and are connected to pin A20 on each of the four slots.




Here is the schematic of the Editor ROM Pack showing the Address lines and bankswitch bit used to decode the EPROMs into the 0000-3FFF bankswitch 4052/4054 ROM space:
Address A15 must be low in addition to slot pin A22 being low to decode the specific option slot ROMs.
If A15 is high - then PIA's or other IO Devices like the Printer Interface ACIA are being addressed.


I hope this helps figure out your backpack slot addressing.

For 4052 ROM Packs - you will only need to check pin A22 on each slot.
For 4052 ROM Expanders like 4050E01 or Jos' 4052/4054 Multifunction ROM Pack, check pin A20 on each slot.

Monty

 

[forty-2]

Ah, thanks for explaining the difference between A22 & A20

Yeah, I've been referring to those pages in the manual, and I've labeled the traces on the LA accordingly, measured right at A20 & A22 of each card edge connector.

From those points, you can see I'm getting:

Slot 41 / J91
BS30/A22 - goes high, then pulses low
BS20/A20 - goes high, stays high

Slot 51 / J92
BS28/A22 - goes high, then pulses low
BS38/A20 - goes high, stays high

Slot 61 / J94
BS00/A22 - goes high, then pulses low
BS10/A20 - goes high, then pulses low

Slot 71 / J95
BS08/A22 - goes high, then pulses low
BS18/A20 - goes high, stays high

So to get basic ROM Packs working, I only need to worry about A22 for now, right?


Based on that, I'm still suspicious of those traces. I would expect them to go low sequentially, rather than all at once, that's what you meant by this:

during power up you should see all the combinations of bankswitch numbers plus the additional combinations for slots 61 and 71.

right?

 

[forty-2]

Manuals for R07 & R08 are linked on the Wiki now
4052R07
4052R08