• Please review our updated Terms and Rules here

IBM 5110/5100 video Display card repair (possible diode fix)

Another application of a diode in conjunction with the the sort of resistor values nearby is to mix the sync pulses and video together to create the composite video.

So it could be that is the diode's application. In general the H & V sync pulses are OR'ed or XOR'ed together and the video (graphic) signals are mixed resistively where the sync value ends up being about 1/3 the amplitude of the overall waveform and the video component the other 2/3, a diode can assist this mix.

It would be really interesting comparing the two boards to make a scope recording of the signals on the diode and the nearby resistors comparing the two cards. I'm guessing that IBM would have used a 3rd party VDU which accepts composite video.
 
Well, which part is the capacitor?

There are several "951" parts (I call them "black monoliths") which are 2-pads wide. Most capacitors I'm familiar with are round with "+" sign caps at the top.

Indeed, that's my suspicion. IBM has ICs and resistor packs that look like nobody else's, so why shouldn't they have their own caps as well. They do seem to have a polarisation mark on them, and the fact that there's one next to each of the four connectors at the bottom, each at the same place (which makes me suspect "power rail") just smells like a capacitor to me.

Is there a kind of metallic putty that can be used to make temporary connections like this?

I've never heard of such a thing myself.

I hadn't heard back from the bitsavers host, but it occurred to me that github might be an OK place to archive these for a while.

Thanks for sharing those. Archive.org is another popular spot for material that awaits a glorious emigration to Bitsavers. That said, I think Bitsavers is unlikely to accept documents in jpg format --- see their homepage under "The PDF Document Format". In any case, github is likely fine too.

2) The display has its own means to turn the Alarm on/off, and turn on the Machine Check light, so the fact this happens at the "4th executive ROS instruction" may be just coincidental. I did enable STEP mode (with the GOOD Display inserted), and it does STEP 1 full opcode at a time. The PC (program counter) at register 1 went 0x000A (expected startup), 0x000C, 0x000E, 0x0010... (2 bytes per instruction for PALM).

Interesting and valuable information. Unfortunately I don't have time this evening to delve into your other insights!

5) At the bottom center, I've never fully understood the +6V stuff. I recall reading somewhere in the MAP and Diagnosis stuff about it. I assume "CE" there means Common Earth? And I think the idea was you can check S02 and your equipment should be getting 6V. So I think it was there to help calibrate a techs equipment, as opposed to the 6V actually being used for anything. But, shouldn't this mean somewhere on this board are a couple 240ohm resistors?

You are correct about this usage: the +6V is for calibration. And CE? It stands for... Customer Engineer. That's the tech's job title :)

I don't think more resistors are necessary for generating this voltage. Instead, there's a zener diode that skims 2.5V off of the 8.5V power to leave +6V (through the 200 ohm resistor to limit current). You don't need to use a voltage divider here if that's what you had in mind for the resistors.

Hang on a second... isn't that diode in your picture sitting right next to a 200 ohm resistor? Could that diode simply be our zener, generating the CE's +6V?
If so, not of much help in solving our problem, but at least we've reverse-engineered a small fraction of the board...
 
I've never heard of such a thing myself.

I recall my father had used something like this, something like:

My idea is I could use a very tiny amount on the small pads at the backside of the Display board -- and it would have enough tack to hold a monitoring wire (for the oscope). Then afterwards, it cleans off with no visible impact to the Display card. For just 1 or 2 pins it might be ok. Any actual solder I think will leave a mark (disrupt the existing solder pads).



Thanks for sharing those. Archive.org is another popular spot for material that awaits a glorious emigration to Bitsavers.

Quite welcome! That actually cost (kind of) a lot, renting time on a large scanner -- which I did before realizing bitsavers had it in PDF format. But it's done and we have a couple formats from independent sources, which is good. I'm not sure what to do with the physical copy anymore - I recall reading on bitsavers, I think they shred most thing after the digital scan, just since it is so impractical to warehouse such an extensive amount of reference documentation.



You are correct about this usage: the +6V is for calibration. And CE? It stands for... Customer Engineer. That's the tech's job title

Oops, I saw that ground symbol and was sleepy and I didn't think that through properly on what it stood for. So, yes it would be rather uneventful if the measurement across that diode is a constant ~6v. Away from the equipment for a few days, but I'm up for trying to scope a few of these pins.
 
Sure, with some low-temp solder you can tack a fine wire to any pad. Use just enough solder to tack the wire to the pad. Clip the wire off after you're done--don't try to un-solder it. I use AWG 30 wire-wrap wire for this purpose (kynar insulation; silver plated).
 
I recall my father had used something like this, something like:
https://www.amazon.com/Sciplus-Elec...ing-Wire/dp/B000Z9H7ZW/ref=asc_df_B000Z9H7ZW/

I think I just find this weird because it seems unfamiliar. Will it really clean off well, though? If not, suppose you did have to replace a component one day: would residue of this old glue make your life difficult?

I'm not sure what to do with the physical copy anymore - I recall reading on bitsavers, I think they shred most thing after the digital scan, just since it is so impractical to warehouse such an extensive amount of reference documentation.

If you would like to post them to an address in London, I would be happy to DM you :)
Either way, I would hold on to them, but that's my style.
Thanks for shelling out for the scans!
 
Yes that diode could well be a zener.

I think realistically to start on the faulty card repair, a card extender of sorts will have to be made (If say the Huntron is not tried first). There are numerous circuit points which require examination with the meter & scope.
 
I'm still away from the equipment for a few days, but I've been reading in the IBM 5100 MIM (Maintenance Information Manual) to look for any additional insights about the Display card.

The highest resolution image of an IBM 5100 Display card that I can find offhand is here (which has a lot more DIP chips than the 5110 version, but does share "that one diode" aspect in common):

And the IBM 5100 MIM section 5 basically includes the System Logic Manual (which is broken out as a separate manual for the 5110).

Here is a portion (from 5-34) of the Display card from the 5100 (where it is explicitly calls out a Zener Diode):


1655270846321.png
(is FE field engineer?)
 
BTW, the difference between the symbol displayed (in this BAD Display card) between NORMAL and REGISTERS mode is subtle: (the examples below are from when the BAD Display was in STEP mode)

1655274666572.png

When we flip to REGISTER mode, the display is "forced" to display RWS content starting at 0x0000 (where it shows the first 512 bytes of memory, which each byte converted to a 2 character hex value). In NORMAL mode, the display is "forced" to display RWS content starting at 0x0200 (but each byte is converted to a single cell display symbol, so it can show a full 1024 characters, 64x16).

I've poured over the APL and BASIC (and the Async Comm) reference manuals, to try to find what this "two circles" symbol is while in the Display Registers mode (again, for this "BAD Display"), and I can't find anything. It's not an 8, or at least I don't think it is (e.g. maybe Line register is skipping a row - but don't think so since it's not in the NORMAL "0" case). The 5100/5110 is also "famous" for its "overstruck characters" - a concept from old typewriters, where you can emulate a symbol by typing a key, backspace, and press a different key, and the compound print results in a different symbol than the two keys individually would. (and Language reference manuals have a section on the keys for this). I don't think this symbol is one of those overstruck keys.




Also, I I found this interesting tid bit.... Those jumpers on the Display card just initialize the DEFAULT character set. You can override that Default using the HOLD and SHIFT 1-8. Norbert's online 5110 emulator didn't seem to replicate this feature, but I'll give it a try when I get back home (I expect the symbols to update as soon as you switch character sets). And apparently you can update this programmatically using WRITEFILE FLS command.

1655275114208.png
 
1655280984345.png
So this Logic diagram is helpful....... It implies I check M05 and M09 to see if there is a BUS OUT ERROR or DEVICE ADDRESS ERROR (or Both). I assume all that gets latched into some DIP? I'd have to follow all those pins (D12...U04).


My quick question is, what is "TP" ? (test point?)

A1H2 on the left is the BASE IO card next to the Processor, which I'm assuming is working OK (since the GOOD Display card is working with that same BASE IO).
 
The highest resolution image of an IBM 5100 Display card that I can find offhand is here (which has a lot more DIP chips than the 5110 version, but does share "that one diode" aspect in common):
Interesting. It seems quite possible to me that one or more of the 5110's big metal can ICs is consolidating the logic of a lot of those DIPs on the 5100.

At this point I would bet reasonable money on "that one diode" being the zener for the +6V supply.

I've poured over the APL and BASIC (and the Async Comm) reference manuals, to try to find what this "two circles" symbol is
I noticed that symbol too. It's too normal-looking to be corrupted data, I would think. But I can't find it anywhere either. The 5100 MIM's character table (PDF pages 336-337) doesn't have anything like it; not too surprising perhaps given that it has completely different characters.

My quick question is, what is "TP" ? (test point?)
Seems like a good guess to me. I was hoping that the Logic Symbol Legend on PDF page 256 of the 5100 MIM would offer a hint, but no such luck. It looks like the 5100 MIM just uses the word "test point" in its diagrams, as on e.g. PDF page 272.

Also, I I found this interesting tid bit.... Those jumpers on the Display card just initialize the DEFAULT character set. You can override that Default using the HOLD and SHIFT 1-8. Norbert's online 5110 emulator didn't seem to replicate this feature, but I'll give it a try when I get back home (I expect the symbols to update as soon as you switch character sets).
Interesting --- so every 5110 has all characters for all Roman-based character sets? (I noticed somewhere that Katakana needs an expansion card.)

So this Logic diagram is helpful....... It implies I check M05 and M09 to see if there is a BUS OUT ERROR or DEVICE ADDRESS ERROR (or Both). I assume all that gets latched into some DIP? I'd have to follow all those pins (D12...U04).
This sounds like a great thing to check --- it could potentially narrow down the space of problems we could be facing.

And the IBM 5100 MIM section 5 basically includes the System Logic Manual (which is broken out as a separate manual for the 5110).
There are some differences, it looks like. The System Logic Manual includes the A1 netlist as its very last page --- what a treasure that is! You can basically probe any signal on that backplane with ease. I wish we had anything like it for the 5100.
 
Last edited:
Interesting --- so every 5110 has all characters for all Roman-based character sets? (I noticed somewhere that Katakana needs an expansion card.)

Confirmed, on the physical IBM 5110 Type 1 that I have. You press HOLD, then use SHIFT-NUMBER (0-9) to change character sets. A simplest example is using "$" dollar sign, but various other symbol characters changed as well. The key "trick" is that you have to use the numbers on the NUM PAD - the "regular" numbers along the top didn't respond to changing character sets.

I'm not sure yet how to get a BASIC program to output all the character symbols (a kind of "ASCII Chart" for the 5110). That might be a good exercise to do in the PALM machine code - load registers with values 0 to 255, and write them into RWS starting at 0x200 (screen buffer) at various offsets. With even just 30 per row, just need about 8 rows.
 
There should be a way to print out the character set under 5110 BASIC. The 5110 BASIC manual has a note showing that symbols can be entered in hexadecimal format. See page 84 of the 5100 BASIC reference where X'0201130903' would be printed as BASIC. I haven't fired up the emulator yet. Little awkward since it doesn't look like one can create a loop to generate hexadecimal constants. But manually entering all 255 values should be possible.
Also, any 5100 programs that use a hexadecimal
constant to generate characters will generate different graphics with the 5110.
Tried with the emulator
Code:
10 A$ = X'CCCDCE'
20 PRINT A$
gave a result of
soy 5110.png
 
Last edited:
Thanks @krebizfan I wrote a machine language version first, but then realized when the DCP is being used to run PALM instructions, it doesn't have the logic running to interpret the Language character set changes (i.e. of pressing HOLD and SHIFT 0-9) -- so you can't SHIFT character sets when using that approach (and I didn't want to change Jumpers on the Display card and reboot to see the new symbols).

So I used your X'xx' example, placed the full sequence in Excel to essentially auto-generate the code, then pasted that over to the physical 5110 using my KBD5110 serial connection. Yes, there is probably a smarter way to code it up, but this brute force way works also. Both the assembly and BASIC version are now in the VEMU5110 repo. Specifically:




I talked more about that in another thread. But pertinent to here, I've found the mysterious symbol that looks like a disjointed 8. It is character #0 in all the symbol sets of the 5110 - and in hindsight, that should have been obvious. Shown below is the entire character set (on good working Display card) from 0x00 to 0xFF for (most) 5110's (except the Katana, or Japanese, configured ones?). I used WHITE background mode, since it has less reflections in the photo and I think it shows the symbols more clearly. (again for the more "international" character sets, just a few symbols are changed - mostly vowel accents and currency related).


IMG_2518a_white_SMALLER.jpg


This doesn't really help solve any particular issue with the BAD Display card. But while in STEP mode, it shows a full 64x16 grid of that first (0x00) symbol -- which suggests to me there might not a Sync issue with the display (although it is a bit jittery in RUN mode), but it may be more related to the Device Address decoder -- if the parity bit fails, maybe it defaults to showing the first (0x00) symbol always. And that Decoder should be one of the IC DIP chips. May be a while before I can test the pins further (other activities ramping up) - I'm not opposed to an experienced Veteran checking out the hardware, if anyone has time over this summer :D
 
In lieu of a proper riser card - can't I just use some really long (100cm) male-to-female jumper wires? It's 4 groups of 24-pins (96 total), so I can run the Display card sitting about 2 feet out of the system. That'll make it much easier to access both the components and pads side (while flat rather than standing vertically). The pins on the A1 board and the 5110 "expansion" slots are a tad thicker than modern typical pins (good because they don't bend as easily), but I've found the standard Arduino style jumper wires still fit on those pins just fine.



But then the next issue is -- what pin is what in the slots of the A1 board? So I itemized out all the pins described for the Display card in the SLM (system logic manual) of the 5110, and I observed that the pins are categorized with letters B D G J M P S U. "BD" is used for the two rows on many other connectors, but what about the other letters?

AND THEN... in the 5100 MIM, I spotted the following for the "A1" board that is in the external tape unit -- it described the connectors grouped as SU MP GJ BD (the same set of letters I identified for the main A1 board from the SLM documents). Curiously, the 5100 Aux Tape Unit diagram is inconsistent with itself ("C1" isn't repeated in all the other slots, like it is for "B1/B2/B3/B4" and "A1/A2/A3/A4") -- and what's even more curious is that in the 5110 MIM, this appears to be "corrected" and it list C1/C2/C3/C4 (with "BD" in each, instead of "SU MP GJ BD"). It's like they tried to hide what letters are used for what connectors on these A1 boards?

1655705874352.png

[also -on the internal tape unit, the two connectors on the back of that device are described using group letters "BD" (one end) and "SU" (at the other end); that's consistent in both the 5100 and 5110 MIM's ]

So, putting the SLM info and this connector labeling info together -- and realizing that backside of the A1 board is essentially a built-in riser board for all the cards -- I think the Display I/O 4x24 pin connector pin out should be:

1655702304996.png

I've verified most of these (the power pins and the front-panel pins, continuity across all the marked GND). Simplifying the above chart a bit, I end up with this:
(note that when the A1 board is "closed" {folded into the bay}, the order of pins is reversed, so PIN2 is now on the LEFT)
(note J10 is also "NC" not connected, I missed that in the above version)
(attached PDF may be easier to read)

1655706175835.png
@stepleton recall you noticed little capacitors along each of the 4 connector pins of the Display card, and notice the bottom PIN3 of each of these connectors is 5V (D3, J3, P3, U3). Coincidence?

Anyhow, that's neat, maybe a chart like this for all the cards might be useful. But only lets us study the INs and OUTs of the card... But the problem here is still more at the component level, one of those IC chips not incrementing (ROW or COL) or not decoding address lines correctly, something like that -- and that'll require "rising" the board out of the slot, to get better access to them.
 

Attachments

  • 5110_slotG_display_pins.pdf
    413.4 KB · Views: 3
Last edited:
So I didn't setup the o-scope, but I did do the following test:

Now knowing the pinouts on the A1 board, and knowing that the fault happens at the '4th' instruction (startup in STEP mode, click 4 STEP 4 times)... I did that for both the GOOD/WORKING and BAD/NON-WORKING cards installed (stopping at that 4th instruction), since both setups should be in the same state at that point (all the other cards are identical)...

Below are the results in measuring each of the pin voltages in that state.

First, a couple notes on the WORKING DISPLAY (on the LEFT)... GREEN cells are the expected GND points, which all check out. BLACK is the expected "NC" (not connected) points, but many of them actually had some voltage, so I marked those in YELLOW cells. It's not an error, and the Display does work. I know "NC" are floating (right?), so they don't have to be at 0; still, I was a bit surprised to see 2-3V on these NC pins.

Now, for the NON-WORKING DISPLAY (on the right), there is ONE main interesting point. But first some other notes:
- I tried to focus just on values that were very different than the working state (marked with "/quote). It was a little subjective on what that "difference" threshold was (sort of around 5-10%).
- When the BAD Display "goes bad", the Display is turned off, so it makes sense that the DisplayOff bits would invert between these two runs (5v going to near 0v; although I thought 0.2V was the threshold for 0, the DisplayOff bits go to 0.38V ? ). Another oddity is that I'm not sure if the Display can turn itself off, or if it has to be Commanded so from the Executive ROS? That may be pertinent, because of the next note...
- The BusOutError (M5) and DeviceAddressError (M9) basically have the same value between these two trials (below 0.2V). This makes me suspect that this isn't the cause of the Audible alarm and the MachineError, in which case -- well, then is the Executive ROS declaring the error? (and yet, it doesn't declare any error with the WORKING Display Card)
- The ALARM pin (P5) appears to be blaring using 8.32V. The alarm is fairly loud, perhaps to accommodate outside conditions (and that it's buried behind a CRT and next to a couple noisy fans or belt). I'm not sure how to induce a constant ALARM on the GOOD configuration to verify that the speaker should be hit with 8.5V (I guess I could just test continuity between one of the pins on the speaker and the known 8.5V line? or actually, from BASIC there is some command like WRITE FLS that can cause the speaker to turn on briefly, I could just measure the speaker leads during that). That said, maybe isn't not voltage that is necessary for the volume of the speaker?
- What really gets my attention is the voltage differences between P10 (EvenBit6), P11 (BusOut3), and especially S5 (EvenBit0). [more on this below]
- DARK RED are differences that I think I can explain. BRIGHT RED are differences that concern me. ORANGE cells are differences that are suspect, but not a big concern as of yet.


1655715404905.png


S5 in particular gets my attention since it was reading over 8V. I'll double check these particular numbers tomorrow when I'm more refreshed (and just double check that both machines were in that "4th instruction" step). But for sure, EvenBit0 shouldn't be reading over 5V. And now that I know the A1 layout better, I can check the source of EvenBit0 as it come from the Processor (A1J2 S02). If the processor is indeed feeding 8.3V there, then I don't have an explanation for that -- if not, and the Display card is just reading a very high value, well that might narrow down the problem? (the value may not go directly from Processor to Display, but the Display might buffer it up into some IC DIP first??)
 
BTW, what does MCC and C2/C4/C5 mean in these pinouts? I was able to associate those with something that I recognized.
 
BTW, what does MCC and C2/C4/C5 mean in these pinouts? I was able to associate those with something that I recognized.

I didn't know, but I had a guess. With several things that have the same name that starts with C and has no other hint, it must be something fundamental, something so common that they'd expect a technician to guess right away. I traced it back in the 5100 logic diagrams to its source:

1655755121102.png

Sure enough: C is for clock. Specifically, we have five separate clock phases. IBM explains the symbology on PDF page 264 of the 5100 MIM:

1655755276214.png

The probing you've done is pretty interesting stuff, and I think it's bound to reveal a lot! One of the things that occurs to me is that some of the intermediate voltages you're seeing could be because the signal on a pin you're measuring is doing a lot of switching between 0V and +5V. A scope of course will tell the tale.

Some remarks based on what you've said:

although I thought 0.2V was the threshold for 0, the DisplayOff bits go to 0.38V

The 5100 MIM says you're logic low if you're below 0.8V (PDF page 256):

1655755796437.png

I'm not sure if the Display can turn itself off, or if it has to be Commanded so from the Executive ROS?

I don't know either, but I do know that the display can be commanded to go off by software. By default, the BASIC and APL interpreters disable the display while they're running a program, since the Display Card will suspend the processor whenever it needs to access RAM to redraw the screen. When the display is off, the processor does not get suspended. The 5100 will only turn the display back on while a program is running if the program prints something; otherwise, the 5100 assumes that you want your result as fast as it can compute it, and it also assumes that you don't want to see it think.

There may be other things besides the processor that can disable the display, potentially including logic on the Display Card itself. Maybe the mechanism can be found in the logic diagrams somewhere.

The BusOutError (M5) and DeviceAddressError (M9) basically have the same value between these two trials (below 0.2V).

That's a good puzzle. However, it looks like several cards in the 5110 can pull down -Machine Check as a result of a +BusOutError or a +DeviceAddressError. Maybe it would make sense to examine the outputs of other cards, identify the pins that correspond to those other error types, and spot the card that's actually triggering the halt?

The ALARM pin (P5) appears to be blaring using 8.32V.

I think you can rest assured that this behaviour is correct. Look at diagram 420 in the 5100 System Logic Manual. +Alarm On is connected to 8.5V through a 240 ohm current-limiting resistor.

But for sure, EvenBit0 shouldn't be reading over 5V.

Extremely weird. You may have noted that this line is an input to the Display Card, not an output.

As a pin on the storage read/write bus, this pin can be driven from several places. The RWS cards can drive the bus; so can the Controller and the Executive ROS card and many others. One of the few cards that can't drive the pin is the Display Card.

Shenanigans like multiple cards trying to drive the bus at once could cause this voltage to go too high like this.

We've already established that it's the Display Card as a primary suspect in the fault --- we know that if you replace the broken card with another one, then the whole machine works.

I think there are a couple of categories of problem. Category 1: the bad Display Card is messing with its inputs somehow.

Maybe a buffer chip got fried and something that should be an input is now driving the line. Maybe consider taking ALL the cards out except for the Display Card. Then check voltages on the pins for the storage read/write bus lines. Don't be surprised if they are high at first: floating inputs do strange things. But use a reasonably chunky resistor (say 200K?) and try pulling the pin down to ground through the resistor. If the pin stays high or won't be pulled down, then I'd say that this is suspicious.

You may want to carry out this experiment with your homebrew power supply. It's possible that without the load of the usual complement of logic cards in the 5110, the PSU will generate abnormally high voltages; your custom job probably won't.

Category 2: the bad Display Card is sending outputs that cause the problem.

The Display Card doesn't have too many outputs, and the ones that go to the display we can almost surely ignore, since the display can't cause Machine Checks. You've basically got the bits it sends out on the storage address bus, +Printer Clock Pwrd, +Display Req Pwrd, and -I/O Display Off. How could those things conspire to upset the entire machine?
 
Wow, and MCC is Master Control Clock? I think I saw "MCC" on some clock in one of the MIM diagrams.

And thanks for verifying the voltage ranges - I thought I had seen 0.2V somewhere, I skimmed back through the MIM documents, but nope, just the reference you pointed out about 0.8V.

Well after a 2nd review with a more fresh set of eyes: I transcribed the BAD-CARD "S" row partially incorrectly. My handwritten notes were correct (where I was writing down what the DMM reported), but I must have dozed off -- notice I borrowed the same values from "P" (both the 8.32 @ PIN 5 and the 1.635 @ PIN 11) going into S. As punishment, I reviewed them all again from the handwritten notes.... Looks like "S" was the only set of pins that went off track (I recall getting a little excited after verifying the 6V CE reference @ S02).


So, what is Printer Clock Pwrd? I did put that one on an o-scope, and it does oscillate quite a bit.

Here is the revised/corrected chart:

1655801202866.png

Out of time for now - this is at least narrowing down which pins of interest to scope.



EDIT: One thing to add -- I've tried this same BAD Display card in my three other 5110 Type 2's (each with otherwise known good cards), and this Display card causes the same exact behavior in each of them (audible alert and RED light on startup). That's just evidence to me that it is the card, and not a coincidental incompatibility with the system I'm trying to get it work in.
 
Last edited:
Whew, it's a relief to see logic lines back at logic voltage levels. +8.3V is pretty hot for TTL!

Let's talk about cycle steal.

I mentioned earlier how the Display Card suspends the processor when it wants to collect video data to display on the screen. This suspension is called "cycle steal" (it uses clock cycles that the processor would otherwise use to do processing), but these days we might call it DMA.

Anyway, in order for the Display Card to steal some cycles, it initiates a handshake with the Controller Card (i.e. the processor) that eventually results in the Controller Card giving the Display Card two bytes of display data on the Storage Read/Write bus. You can read all about this handshake in the 5100 MIM PDF page 246. (I suspect the system is not so different for the 5110, but you may want to double-check.) Here's the important bit:

The' -display request' line is used by the display adapter to request a storage cycle steal when the adapter is ready to receive the next two bytes of data in the display data register. The controller activates the '-stolen cycle next' line during the storage cycle that precedes the requested stolen cycle. The' -stolen cycle next' signal deactivates the '-display request' line and limits the display adapter to alternating the cycle steal activity with a controller storage cycle.

During the stolen storage cycle, the controller activates the' -stolen cycle' line and puts the two types of data addressed by the storage address bus onto the storage read/write bus. The' -stolen cycle' line also gates the data from the storage read bus into the display register on the display card.

It seems likely to me that if the Display Card is no longer doing this handshake in the correct way, then it could cause the processor to become upset. I notice that your -Display Request, -Stolen Cycle, and -Stolen Cycle Next voltages differ by about 20%, which (if we assume they are logic signals) may suggest a different pattern of activation to the normal card. I might investigate those signals in particular if I were troubleshooting this problem.

However, the differences you're observing in your tables could also simply reflect the fact that you're making observations here (I think?) after the processor has decided to shut down the display owing to the error, and that shutdown would disrupt the amount of cycle stealing the card is doing anyway. If -Display Off (bottom left below) is active (low), then (...follow the traces...) the leftmost OR gate is active and its output goes low, which means the two "Char or Cursor" AND gates are inactive, which means the OR gate to their right is inactive, which means the AND gate at top centre is inactive, which means -Display Request stays inactive (high) too. There are no cycles being stolen anymore.

1655841222698.png

So, to compare like with like, perhaps what I would do then is scope the cycle steal control lines on the good card and the bad card and compare their behaviour just before the fault. See if there's any difference; if so, that might be important.

Please note that in the above, I've been reading the 5100 MIM because it's more familiar to me. It looks like you can find a similar description in the 5110 MIM on PDF page 117.

So, what is Printer Clock Pwrd? I did put that one on an o-scope, and it does oscillate quite a bit.

I'm afraid the 5110 MIM doesn't turn up search results for "Printer Clock", but that could be due to OCR failure. In the System Logic Manual, however, you do see it get buffered and go into the printer (PDF page 8), where it goes into a box called "Print Emitter Timing". It's worth saying though that this signal emerges from the "Character Counter" on the Display Card (PDF page 6), and that same line also enters a Display Card box called "Cycle Steal Logic". So this is probably another signal that might be worth comparing in your study of the Cycle Steal handshake.

I mentioned that the 5110 System Logic Manual abbreviates more detail than the 5100 MIM Logic section. The "Cycle Steal Logic" is a great example of this: for the 5110, it's a box; for the 5100, it's logic gates like what you see above.
 
Back
Top