The Tale of the Good Kaypro and the Bad Kaypro

[k6whp]

This just cropped up a couple of days ago and I am working my way through it. Situation is Kaypro 2X units. Unit #1 functions correctly, unit #2 does not. Unit #2 was function correctly and then failed as described below. Both have Epson SD-521 drives.

Symptoms:
(1) Boots O.k.
(2) A DIR command will display the A: drive contents.
(3) A -D (9-D.COM directory program) will not.
(4) A DIR B: command only results in the "DIR B:" command appearing on the screen, the B: drive lighting up and spinning but nothing further.
(5) Any other attempt to execute a program in the A: drive or the B: drive will result in similar results to #3 above.

Attempts to resolve the problem:

(1) Steps taken to determine problem by swapping components.
(A) Floppy cables swapped;
(B) Mother board swapped.
(C) FDDs (A: and B: FDDs) swapped. (Included resetting the jumpers and resistor plug as appropriate.)
(D) FDDs swapped in units -- A: for A;, B: for B:.
(E) All above resulted in Unit #1 failing as it did and Unit 2: behaving correctly.

(4) Measurements done.
(A) Both units showed +11.5 VDC at PS main board plug. Both showed little/no fluctuation on main board plug when above symptom tests done.
(B) Scope put on pin 6 of U43 (Z80 MCU) and measured 4.000 MHz as expected. No measurement done with symptom tests.
(C) Scope put on pin 24 of U44 (1793 FDD controller) and measured 1.000 MHz as expected. No measurement done with symptom test.
(D) Will measure clock stuff above during symptom tests. (Do not believed they will vary).
(E) When I figure out how to jury rig the FDD cable, power connector, will measure voltage, will measure same.

Conclusion so far.

It appears to NOT be mother board, FDD cable, power. Maybe the ROM chip? Both are (I think) 81-292s - I could swap them?. At this point I am just taking a W.A.G.

Wonder if anything occurs to anyone. Thanks in advance for suggestions and/or sharp eyes.

 

[durgadas311]

Might help to hear more about the symptoms of #3 above. The DIR B: hang might be no index pulse coming from the drive. (or is #4 NOT a hang?)

It's not clear exactly what happened after swapping drives. If A: <-> B: still failed for drive B:, perhaps it is the cable or IDC connector(s) - working off the assumption that INDEX or some other critical signal is not getting through. I'm assuming when you swapped A: <-> B: you physically moved the drives on the connectors? rather than simply changing DS0/DS1 jumpers and leaving everything else the same?

 

[durgadas311]

If both mainboards are the same vintage (same ROM versions) you should be able to swap the EPROMs.

 

[k6whp]

Damn! No matter how much I proofread these things! The answer to your question is:

(1) The "DIR:" does not hang. Doing repeated "DIR:" commands do not hang.
(2) The "DIR B:" command hangs.
(3) The "-D" (the -D.COM program) hangs.

..as to swapping the ROM, I will try that. Ironically, it was the ROM who lost its label -- as we discussed a while back.

Question: is it possible to burn a new ROM?

Be right back with the ROMulan swap answers.

Thanks Doug!

 

[durgadas311]

You can burn a new ROM. There are images of the ROMs all over the internet. You'll need an EPROM programmer, not sure if all the modern ones will work with those EPROMs (depends on the programming voltage required). I believe there are modern EEPROM (flash) devices that can be substituted, but I have no experience with that.

I know others have quoted statistics and physics related to "uncovered" EPROMs actually fading, but it always makes me feel queazy to see that and I have heard tales of faded EPROMs. It doesn't need to be totally erased in order to be flaky (erase time is not the same as "fade" time), so I'm not really sure just how much exposure is required to cause problems.

That being said, I'm not sure how a faded ROM could work for drive A: and not for drive B: (the same code is executed for both). But, stranger things have happened.

 

[k6whp]

Question: is it possible to burn a new ROM?

Be right back with the ROMulan swap answers.

Thanks Doug!

Turns out burning a new ROM is not an issue. Swapping them did not reveal the problem. That is, when the good 2X received the bad 2X's ROM it behaved correctly. I hesitate to put the good 2X's ROM into the bad 2X for fear of glitching it.

Oh, by the way, doing an ERA *.[whatever] (or an ERA command in general) works on the A: drive but hangs up -- as you might expect -- doing an ERA command (ERA B:*.* or similar). REN works the same way as well.

So, as far as I can see -- which ain't very far -- it may be in the drives or the power supply to the FDDs. However, I am going to -- once more -- verify the operation of the two FDDs in the good machine and then swap them into the bad machine and see what happens.

After that, it's back to reading the ratty schematics in that bogus "Technical Service Manual" Kaypro put out in September 1985. Let's ' just say it ain't up to the same scratch as a Tektronix scope manual scope manual. If this were a Tek 485 or a 7704A, I'd be home for supper by now!

 

[durgadas311]

Can you put a probe on the 1793 INDEX signal and see if it is still pulsing when "hung on drive B:"? (U75 pin 35) I seem to recall that commands on the 1793 would stall if there was no index pulse (The RDY input, pin 32, is tied high so the 1793 thinks the drive is always "ready"). The INDEX pulse is what is used to timeout commands, so if you were getting INDEX I'd expect you to get errors rather than a hang. Since you see drive B: LED light, it would appear that the select signal is getting to the drive, and the same code should be executed for drive A: or B: (just a different bit set in the control port), seems there must be something specific to drive B:. If drive B: has always been the same physical connector on the cable (e.g. the one at the end of the cable), then it might be pointing to a problem with that connector.

Or have you already swapped the cables between the two units?

 

[k6whp]

Swapped cables? I have! (See 1A above.)

But you bring up a good point and the main reason I am in this exercise is to learn. It is surprising to find the datasheet is still available for that chip. (I have attached it for any who might stumble on this thread needing help in the future.)

In order to scope the chips (they're U43 for the Z80A and U44 for the 1793 on my model/according to the September Kaypro 1985 tech spec, page 6--27), I'll have to do some cartage and swap units. For the stuff I got today, I was stretched out like Rose Mary Woods was when she created the 18-minute gap on the Watergate tapes! First I'll swap the drives (yet gain, yet again) and make sure the first test was correct.

For the rest of the evening, I'm going to read up on the 1793 and squint at the hoary tech spec schematics to see what else occurs.

Thank you, Doug, for bringing this up. Appreciate your assistance and suggestions ever so much!

 

[durgadas311]

Oh, right, you have 2x and for some reason I was still looking at the K10 schematics. But, basically the same hookup for the 1793. Pin 35 is the INDEX pulse, which is what I was wondering about. FYI, the best datasheet for the 1793 (IMHO) is the original Western Digital one, which includes flowcharts of the commands. Much more information about the operation of the chip. Probably more than you need to debug this, but worth saving.

 

[haglebu]

I recently had problems with my Kaypro II. I measured and checked a lot but found nothing. Long story short: it was the data separator.

Some reading:
Micro Cornucopia, Number 23, April-May 1985, page 35

 

[k6whp]

Thanks for the tip! Alas, that was not the case.

Noting that the WD9216 was an 8-pin DIP, I simultaneously rejoiced and dreaded the prospect that it was not socketed. God bless Adam Kay and his minions for they soldered no chip, no matter how trivial, onto the board. Seeing that, I swapped the 9216 from the bad board to the good (not wanting to glitch a possibly good chip by going the other way) but the good 2X functioned as designed. Reluctantly, swapping the "good" 9216 to the bad board cured nothing as well.

I will search further but, if you read the initial post, you will find that I swapped motherboards with no change -- i.e., bad remained bad and good remained good. I am loathe to swap the WD1793 chip (it is a virtual 40-pin centipede) at least until I exhaust other alternatives of discovery.

Proves what a valuable asset the .pdf copies of Micro-C can be! I am very grateful for your extensive research and suggestion.

 

[DeltaDon]

I don't understand something or am missing something. Based upon the following:

"Situation is Kaypro 2X units. Unit #1 functions correctly, unit #2 does not."

"Attempts to resolve the problem:

(1) Steps taken to determine problem by swapping components.
(A) Floppy cables swapped;
(B) Mother board swapped.
(C) FDDs (A: and B: FDDs) swapped. (Included resetting the jumpers and resistor plug as appropriate.)
(D) FDDs swapped in units -- A: for A;, B: for B:.
(E) All above resulted in Unit #1 failing as it did and Unit 2: behaving correctly."

So #1 was working, #2 had a problem. You moved stuff and now #1 has the problem. Seems to me that something moved from #2 did cause the problem since #1 is now failing.

 

[k6whp]

"So #1 was working, #2 had a problem. You moved stuff and now #1 has the problem. Seems to me that something moved from #2 did cause the problem since #1 is now failing."


As I wrote:

"This just cropped up a couple of days ago and I am working my way through it. Situation is Kaypro 2X units. Unit #1 functions correctly, unit #2 does not. Unit #2 was function correctly and then failed as described below. Both have Epson SD-521 drives."

The status of both machines remains the same: #1 functions correctly and #2 does not. NOTHING I did by way of substitution or re-configuring altered the original condition of either machine. Sorry for the confusion.

 

[k6whp]

Well ,a little late in following up on this but the real world collided with some frustrating testing and uncertain results.

Anyway, a couple of problems were found and -- at first -- they seemed to resolve the issue but continuous testing saw the bad machine lurch back into its realm of misbehavior. Specifically:

(1) During testing, it was noted that "light, inadvertent jiggling" of the connector from the DC power supply supply to the motherboard and the floppies caused the machine to blink on and off. (That is, power down and back up rapidly.) I inspected the PS board and found evidence of some bad stress joints of the connector. (Pics below.) All of the joints were re-flowed and, needless to say, the pins were polished and corrosion removed. The Molex connector was cleaned with contact cleaner and a toothpick was used to scour the "holes".





Further vigorous stressing the connector did not re-occasion the intermittent behavior.

(2) After the testing, the machine behaved normally; that is, I was able to log into the B: drive, execute programs from either drive without incident -- for a while. Gradually, the problems returned. I was able to execute a DIR command on the A: drive successfully but no longer to do a "DIR B: command or execute programs on either drive. Basically, it appears as though the drive light turns on indicating it is selected but the drive does not read. Also note that testing discontinued (that is the machine was allowed to "cool down") but immediately upon resuming testing, the problems evidenced themselves again.

(3) At this point, it's no more Mr Nice Guy. I measured the PS voltages to the motherboard and the drives and noted:

(a) Motherboard at 11.8 VDC, and 4.8 VDC.
(b) A: FDD essentially the same.
(b) B: FDD essentially the same.

The low voltages are puzzling (i.e., NOT 12 VDC and 5 VDC.) and perhaps these should be higher. Does the PS have problems? Dunno. Did not measure the FDD voltages with the drives running but there has to be some draw down with the drives sucking current to operate.

..so, attention turned to the PS and the wiring harness for tomorrow. No need to respond; I am just updating the thread to be helpful for others who may need help with similar problems further down the road.

 

[acgs]

One way to check and see if they are "low voltages" is to test the "Good Kaypro" and compare the results. The PS likely still has issues, but maybe not with the voltage. Regarding after the cool down, I'm not sure I read you: did the problems immediately resume or come back as the system warmed up?

 

[k6whp]

After I "fixed" the 2X, the problems returned eventually. The remained. That is I would turn off the machine, go watch a sandwich, come back, switch the machine on, and the problems were still there.

 

[acgs]

OK. So the problems went away after your "vigorous testing" only once, for a short time. Now everything is back to abnormal with the bad machine. Got it, thanks.

 

[k6whp]

...oj,and the "low voltages" were the same on these good 2X. I read somewhere that +/-5% is OK. So they're well within the operating range.

Still in all, the B: drive isn't being read. Also, the A: drive does not read/execute any programs.

Not a wizard when it comes to the WD1793 chip, but Doug up-thread recommended I look at that. Will scope that one out...

I think I can muster four channels on my 7704A.

 

[acgs]

Have you tried disconnecting the B: drive and running the machine with only the A: drive? If that allows you to run programs and continue to read the disk, maybe a problem developed in the select line jumper(s) / trace(s) or the resistor plug.

 

[k6whp]

I have done so. The plugs on the FDDs are set correctly. A: is #1 (as in 0,1, 2, 3..etc. and B: is #2. The resistor plug is in drive B: per requirement.