[KAYPRO 10] Boot Problems! Help?

[Patrick.B (TTR)]

so I have acquired a Kaypro 10 (see the main board image attached).\

I have attempted to boot it, but the HDD-ready light never comes on though the HDD does "spin up"

When the system is turned on, I have a power light on the screen that displays a " Kapro 10 v1.9e " floppy light turns on, and I have a floppy disk system diskette 2.2F in the drive.
Nothing more happens than the light on the floppy drive continues to remain on. I hear no drive-head activity.

So I switched the floppy cable for a Gotek unit with an HFE image of the same system Diskette 2.2F and the LCD display that shows me when the system tries to "access tracks/read/writing"
and in the case of the Kaypro, the drive is never actually trying to load anything.

I disconnected the HDD suspecting it may be waiting for it, but I get Identical results.

Any thoughts? what I should try first? and no I have not one idea what ROM it is either as shown in the image the label is missing (not found inside the case)

 

[cj7hawk]

Cable wrong way around? That would account for the disk drive being selected without a signal from the Kaypro.

Otherwise, check the chip that sends the signal and confirm the input signal and the output signal. It is most likely a TTL open collector chip.

I'm not familiar with the Kaypro to suggest more than this sorry -

 

[durgadas311]

If you have the ability to burn EPROMs, you might try creating this monitor ROM and see if you can poke around at the hardware and find out what is not functioning. http://sebhc.durgadas.com/kaypro/monitor.pdf

 

[Patrick.B (TTR)]

I did get lots of suggestions, just a little update the original ROM was a 81-302B (1.9e) now it is updated to ROM 81-302C v1.9E but I get identical results. with 2.2F and 2.2H system disks.

If you have the ability to burn EPROMs, you might try creating this monitor ROM and see if you can poke around at the hardware and find out what is not functioning. http://sebhc.durgadas.com/kaypro/monitor.pdf

I will do that now and report

 

[Patrick.B (TTR)]

okay so ran the diag rom

the keyboard test
all keys work without hanging

testing the crtc
and it hangs during the test

tested Hd
it initially and was about to read from it without a problem.

so seems the 6545 chip or video ram is faulty?
The documentation is .. vague to say the least on "how-to" do and understand results.

 

[durgadas311]

I probably need to see what you're trying for the CRTC testing. I thought you were getting the screen initialized with 81-302 ROMs, so it is probably just a procedural issue.

If the HDD is functioning and reading the boot sector, then I don't think the ROM will ever try the floppy. Doesn't explain why disconnecting the HDD didn't boot the floppy, but the hardware might still look to the ROM like it is working enough to hang. The 81-302 ROM probably doesn't allow for the HDD being absent, unlike the "universal" ROM 81-478 which configures itself for HDD+floppy or dual-floppy automatically. I'm not suggesting trying the 81-478, as that is intended for a different mainboard and different CP/M versions.

 

[Patrick.B (TTR)]

Durgadas311 I probably need to see what you're trying for the CRTC testing. I thought you were getting the screen initialized with 81-302 ROMs, so it is probably just a procedural issue.

I am talking about the Kaypro Monitor test.

The CRTC 6545 test looks like this: in the manual, it says it should have "Update" three times and exit, mine only says it twice and hangs until I press break.

: N CRTC
: T CRTC
Wait... Update Update Update
:

mine does
: N CRTC
: T CRTC
Wait... Update Update

Durgadas311: If the HDD is functioning and reading the boot sector, then I don't think the ROM will ever try the floppy.

well the monitor test for the HDD read the boot sector and almost look identical to the paperwork for testing. so I was encouraged to test other things
which led me to the CRTC testing... and SY68545A-1 will take 7-10 days before I can find another. unless I can test the one from a Kaypro 2?

 

[durgadas311]

If the 81-302 ROM was showing the CRTC working, I'd forget about the CRTC tests for now. Let's focus on what we know is broken.

So, it's possible that the OS image on the HDD is corrupt. I'd have hoped that the controller would give a read error, but I'm not sure what the ROM does in that case, either (it might just loop indefinitely). We could walk through reading more of the boot track, to see if any errors arise. You could try that by doing "O 83 01" (and substituting 01..10 for each try) and reading the sector. That might require also doing "O 82 01" to reset the sector count each time - I'm not sure exactly how the controller handles that. I'm assuming that your HDD read command came back without errors? The final byte after the "T HDD 20" command (error register) should be "00", and the last byte before that (last status register sample) should not have bit-0 set.

The HDD commands are still a work in progress, as I only have a simulator to use for checking those. We may come up with improvements as we work on this problem.

I also need to study the 81-302 code and get a better idea what it is actually doing. I can get to that later today.

 

[durgadas311]

From what I'm seeing in the 81-302 ROM code, the floppy drive is being checked for a diskette first, and only if there is no diskette then will it try to boot off the HDD. The way it determines whether there is a diskette is by turning on the motor, selecting the drive, etc and then looking for the INDEX pulse. If there is no INDEX pulse in a "reasonable" amount of time, it then assumes it should boot from HDD.

So we may want to test the floppy controller to make sure it is working correctly (that the ROM can see the INDEX pulse when a diskette is in the drive and spinning). There is a test for that in the example, which should work for this case (the floppy on a Kaypro 10 is "A" in the sense that it is bit-0 of the sysport). If there is an INDEX pulse occurring every 200mS, then we can check further to see what might be happening.

I take it the system did not boot when there was no floppy in the drive? Which might not be surprising if the HDD has failed. It could additionally be that there's a problem with the floppy, yet-to-be-detected, preventing any boot.

The ROM code is fairly convoluted, and we don't have source code (I'm disassembling it), so finding more detail on the decisions being made at boot time is going to require more-strenuous investigation.

 

[Patrick.B (TTR)]

what you are saying is falling into what's happening, the drive lights up but never ": actuates" I pulled it and it is "selected" but never activates, and neither does a gotek for confirmation so seemingly I have an FDC issue on my hand that never replies to the process with "no disk in the drive to boot get HDD started", and disconnecting the floppy doesn't do anything either, so likely something is being pulled down or up all the time instead of fluctuating to show yes or no floppy in the drive. I did swap the SY6545 from the Kaypro II I have into the Kaypro 10 and the testing was even worst for the CRTC it was only 1 "update" showing before locking up, and nothing was on display when I put the 81-302C from in at all, so the CRTC test is on about something as well, we may have more than one issue simultaneously

Here are the HDD tests you requested: Please note on the emergency monitor ROM, the " HDD Ready led lights up when turned on by "N HDD" were has with the normal 81-302C ROM it never comes on.

I also tried to test the floppy drive.. and it doesn't look good.

 

[durgadas311]

Regarding the CRTC, I'm assuming you meant that you swapped the SY6545 from a Kaypro 2/84 since the Kaypro II has no CRTC. Assuming that the CRTC works when in the Kaypro 2/84, I would expect it to work in the Kaypro 10.... probably (there are some differences). If that "Update" test is hanging, then it does mean that it is possible for the ROM to hang when trying to put characters to the screen. But I thought that you were getting the full "* KAYPRO 10 v1.9E *" on the screen - although it is possible there is a problem that only happens after warm-up and the ROM is able to print that message before it fails. There is a test for read/writing CRTC registers, "T CRTR", which you run *before* (without) initializing the CRTC. That might tell us if there is a problem getting the CRTC initialized. There's also a test for the Vertical Retrace signal, which also depends on the CRTC operating correctly. That test ("T VRT") requires initialization, but will not hang if there is no signal.

The "KAYPRO 10" message is put on the screen immediately before trying to boot, so if that were to hang then you'd never get to the boot routine. However, if that message completed without hanging, you would not hang until CP/M tried to print the signon message.

One thing to note about (DD/DS) floppy boot is that the heads don't move off track 0 - all the boot sectors and the CP/M directory are on track 0 (side 0 and 1), so there will be no stepping when you reach the point that CP/M signs-on (which might hang if the CRTC circuitry were faulty).

 

[durgadas311]

And, yes, the monitor documentation is pretty skinny. What I need to do is write a second, longer, document about "troubleshooting techniques" ("How to Troubleshoot a Kaypro"), the monitor document is really just meant to be an overview and reference guide.

 

[durgadas311]

Oh, I missed the floppy testing. Yes, that doesn't look good. I'm wondering if the FDC is actually "there" at all. You might try reading the status register ("I 10") and see if you get something other than "FF". If all registers (10, 11, 12, 13) return "FF" that would indicate that the FDC /CS isn't working, or the chip is just fried.

The wait for INDEX code does timeout, and so won't hang. But, the code before that that does the RESTORE command might actually hang if there is no FDC (if the FDC never completes the RESTORE command).

 

[durgadas311]

It is also possible for the chips to "loosen" in their sockets, or otherwise have connection problems that might be solved by re-seating the chip (carefully, of course - those 40-pin DIPs are tough to get out and bending or even breaking pins is a possibility).

 

[Patrick.B (TTR)]

It is also possible for the chips to "loosen" in their sockets, or otherwise have connection problems that might be solved by re-seating the chip (carefully, of course - those 40-pin DIPs are tough to get out and bending or even breaking pins is a possibility).

Yeah I have re-seated all the chips already, someone suggested a ram chip may be at fault as well causing the hanging after the initial " *Kaypro 10 v1.9E* "

There is a test for Read/writing CRTC registers, "T CRTR", which you run *before* (without) initializing the CRTC. That might tell us if there is a problem getting the CRTC initialized.
There's also a test for the Vertical Retrace signal, which also depends on the CRTC operating correctly. That test ("T VRT") requires initialization, but will not hang if there is no signal.

:T CRTR
Passed.

:T VRT
A0 03E7
80 044A
A0 0A58
80 0ABB
A0 10C9
80 112C
A0 173B
80 179D
A0 1DAC
80 1E0F

I'm wondering if the FDC is actually "there" at all. You might try reading the status register ("I 10") and see if you get something other than "FF". If all registers (10, 11, 12, 13) return "FF" that would indicate that the FDC /CS isn't working, or the chip is just fried.

I 10 - C1
I 11 - FF
I 12 - 01
I 13 - FF

Am I interpreting those 2 FF errors to indicate something awry with the floppy disk controller area of the motherboard?

 

[durgadas311]

I just tried this out on my 2X, and got:

Input 10 = 00
Input 11 = 00
Input 12 = 01
Input 13 = 00

This is after power-on, without doing anything to the FDC (or sysport) at all. If yours had C1, FF, 01, FF after power-up, that looks wrong - especially since "C1" would indicate that the FDC is busy with a command. That value also implies that the FDC "RDY" input is grounded, which should never happen since Kaypro ties that pin to Vcc. If I select drive A: and turn on the motor (no diskette in drive), I get:

Input 10 = 44
Input 11 = 00
Input 12 = 01
Input 13 = 00

Again, this is without issuing any command to the FDC.

 

[durgadas311]

Another test might be to see if you can alter the contents of registers 11,12,or13. Those should generaly just read back the last value you wrote (provided you don't run any commands).

 

[Patrick.B (TTR)]

f yours had C1, FF, 01, FF after power-up, that looks wrong - especially since "C1" would indicate that the FDC is busy with a command. That value also implies that the FDC "RDY" input is grounded, which should never happen since Kaypro ties that pin to Vcc[/quote

yeah, I agree so I will see if I have a spare 1793 on hand and see if that stops.

 

[durgadas311]

I'm going to improve the FLPY test (and others that use the same method) so that when there's no change to the status register for the entire 65536 iterations, it places a single sample containing the (final) value of the status register. This might have helped in this case, as your initial "T FLPY" runs would have immediately shown the bogus status register contents.

 

[Patrick.B (TTR)]

so based on the result of the "I" command I decided to replace the 1793 chip this morning., no cost I have tons of them.

The bugger booted right up, the HDD was likely corrupted as it wanted a system diskette in drive A: but the HDD led lite right up after the initial "kaypro v1.9E"
and the floppy in question 2.2.H is currently installing the OS on the HDD ..

Thank you and your " emergency ROm was really useful. I have some K2 to test next and If you need more help testing the Rom I am volunteering cause it was indeed a time saver.