[KAYPRO 10] Boot Problems! Help?

[Patrick.B (TTR)]

now to figure out what diskettes are needed to restore a HDD ... its asking for diskette #2 of hload

 

[durgadas311]

Great news! Yes, please do try the ROM out on the K2 as well - provided it makes sense (depends on the symptoms, the ROM probably won't help trace smoke back to it's origin ;-) And keep providing feedback. Efforts like these will help write the "Troubleshooting Kaypros" document, let alone improve the monitor. I made the change to the T FLPY (et al.) output, version 2.3 is now out there. Hopefully that will help in the future, too.

 

[durgadas311]

Yeah, there's a whole series of diskettes needed, although I think the first one will get you a bootable harddrive(?). Maybe 2 is also required... I forget what I figured out going through that. Do you have the rest of the images? They should be "out there" someplace.

 

[durgadas311]

Regarding the "T CRTC" problem, I am also seeing that on my 2X. I obviously have something broken in that code, and will have to work on it some more. Stay tuned... but don't throw away your SY6545's yet! (yet another subtle difference between simulator and real hardware).

 

[Patrick.B (TTR)]

ah old hardware... got to reinstall a reload, parked the head to shut it down.... and then the hard drive fails on boot now BDos error , I think the TM502 has just lost its last breath.
I wonder if I can just slap another hard drive in it. or do I need to make a special process... ?

 

[durgadas311]

I don't have experience with that, but I hear that some other drives may be compatible with the controller. Of course, they have to use the same interface plus need to have >= geometry. There's also a Kaypro FreHD solid-state replacement.

 

[durgadas311]

FYI, I found the problem with the T CRTC command. Turns out, the 6545 might trigger the Update bit multiple times during a single retrace, and I was assuming that happened only once. So, I needed to keep resetting the Update bit while waiting for it to go "off" (stay off). It now works on my 2X, ran the command many times (90+), and with a larger update count (6), and saw no hangs. Version 2.4 is out there.

 

[durgadas311]

I have started work on a Diagnosing Kaypros Guide here: http://sebhc.durgadas.com/kaypro/diags.pdf. The scope of the document is after you've verified power supplies, capacitors, etc and are able to run the monitor.

The latter half of the document is just copy-over from the monitor document - still to be written. But it's a start. Let me know what you think, additions, corrections, etc.

 

[Peter z80.eu]

@durgadas311 - May be you have an idea where/how to start with a non-working Kaypro 10... you can switch the computer on, but only a small cursor appears on the screen, nothing else.
There is no ROM message, no disk drive action (drive select or similar), no hard disk action (no LED blinks).
The power supply delivers all voltages as expected. Someone checked all TTL and other chips with a cooling spray before I got it.
The connectors were all disconnected and reconnected.
The first step would be getting a ROM message I guess - but what should I check first in the described situation ?

 

[durgadas311]

Depends on the ROM you have, but for the 81-302 ROMs the CRT controller is initialized (cursor appears) and then the HDD controller (WD-1002) is reset. It is possible that the reset operation hangs, due to a faulty HDD controller. between CRTC init and message displayed it resets the WD-1002 and then initializes high memory. I don't think the high memory init can hang.

 

[durgadas311]

I take that back, I don't think the reset WD-1002 can hang. So, perhaps it is trying to "print" the message to the screen but the CRT circuitry is not functioning correctly and it is hung waiting for the ready signal to write data to CRT RAM. If you are going to replace the CRTC, be certain to get the correct version. The 6545 or the 6845E ("E" is vital).

 

[Peter z80.eu]

Thank you for your fast answer. I would prefer to be sure that the CRTC is defect. Are there any kind of test procedures or measurements for that ? I mean, I do not have a 2nd Kaypro 10 to compare voltages and oscilloscope curves, so this would be very helpful to get more than only the schematics. I own a Kaypro IV, too, but there exist no CRTC in the IV for a quick exchange test.
For example, how can I be sure that the Z80 is working ? Are there any easy test procedures to recognize this ?

 

[durgadas311]

There is some evidence that the CPU is at least partially functional, since the CRT chip is getting initialized. If the CPU could not run at all, the CRT chip would never get initialized and there wouldn't be any video at all. Any further testing of the CPU probably has to be by running software.
If you can burn EEPROMs/EPROMs, and can hook up the serial port to a PC, you can try and run the monitor ROM I provided links to earlier. Documentation is in http://sebhc.durgadas.com/kaypro/monitor.pdf and http://sebhc.durgadas.com/kaypro/diags.pdf. There are procedures documented there on how to check out basic operation of the CRT chip. I also have a standalone ROM image for a memory test, which is a good thing to try if even the monitor ROM won't run correctly.

 

[Peter z80.eu]

Will try to burn the monitor ROM into a 2764, stay tuned ;-)

 

[Peter z80.eu]

Yes, it took more than a year, but finally I did it. I burned the content of the emergency monitor ROM file into a NEC D2732A EPROM, verified the content, and replaced the original boot ROM 81-278 with it.
Unfortunately after power on, it showed instantly only a pattern with an ugly character, but only this character repeated, not random chars.
See the attached pictures. After I replaced the EPROM with the old ROM, I got the old picture, only a blinking cursor.
So I am again clueless about what goes wrong, except that I may have to create the serial cable as described in the monitor PDF file. Not sure I can use a standard Null Modem cable.

 

[durgadas311]

So if the monitor ROM shows that pattern on power-on, it indicates that the CRT controller chip did not power-on correctly. It should have powered-on disabled - no display. I'm not sure what to think of that. The fact that you got a cleared display (and blinking cursor) with the normal ROM indicates that the CRT controller and display RAM were initialized. However, you are not getting the ROM signon string. The display RAM chips are static, and I'm not sure what they might do on power-on. My 2X (universal mainboard) has random characters, but does not power-on with the display enabled (I only see the random pattern after I enable the display using the monitor command).

The serial port hookup can be daunting. A NULL modem cable/adapter is only required if the signals need to be reversed, and it depends on what you are connecting to on the other end. Are you using a USB-serial adapter, or do you have a native serial port on your PC? Or are you connecting to a terminal?

 

[Peter z80.eu]

The description of the "Emergengy Monitor" ROM says in http://sebhc.durgadas.com/kaypro/monitor.pdf :

"The purpose of the monitor is to help debugging a broken Kaypro. Replace the normal boot/BIOS
ROM with an EPROM created from this monitor image. The Kaypro will now interact using the serial
port (“Serial I/O” on early models, “Serial Data” on later models). No other hardware will be
initialized, which means the display and keyboard will not work, nor will the disk. The display will
usually be filled with random characters, or at least will not be cleared."

So I thought "filled with random characters" seems to be correct, and the serial port must be used (instead). Wrong ?

 

[Peter z80.eu]

Is it a good idea to replace the crtc 6545 ?
I have a second Kaypro 4, at least I can compare measurements, if this make sense.

 

[durgadas311]

Perhaps I should have said something like "the display characters are unpredictable" - although even that is not what you are seeing: it appears your display is filled with *predictable* characters (the same one character). I can't see the pattern clearly since the photo is over-exposed, but it looks like maybe it isn't one single character but a consistent repeating pattern of maybe two characters. Although, we don't know why the display is on at all. It certainly is possible for a bad CRTC to present with multiple symptoms (e.g. failure to power-on disabled, plus hang when storing characters). But,
You could swap the CRT controller chips between your good and bad Kaypros (don't put the suspected bad one in your K4, though), to see what changes. Be careful to follow good ESD practice as these parts are scarce and you don't really want to risk damaging a good one. It would be fairly easy to determine if it is bad before swapping, but you'd need to get the serial port hooked up.

Your donor Kaypro 4 must be a 4/84, though. The 4/83 ("IV") does not have a CRT controller chip.

You mentioned that your original ROM was "81-278". I don't see any reference to that part number, in ROM image archives or Kaypro documentation. There is an 81-277 and an 81-478.

 

[Peter z80.eu]

See attached file, I just read the wrong ROM sticker (81-278 seems to be the char ROM). It's of course the 81-277. My 2nd Kaypro is a Kaypro 4, not a Kaypro IV. Will try to replace the 6545, not sure where to get a new (or at least used but working) one.

Edit later: I reseated both, char ROM and boot ROM (with the inserted emergency monitor ROM), and the display is now totally blank (no cursor, no random chars). Also, the floppy drive LED blinks now (?).