H89 problem: no boot prompt

[Roland Langfeld]

Dear members of this group,

I need your help: my H89 stopped working from one day to the other: after power on there is only one beep, no "H:", just an underscore as cursor.

the configuration is 

- 64 kB RAM, 

- H37 floppy controller, 

- memory decode ROM U517 444-66
- JJ501 - JJ504: 1 | 1 | wired to pin17 P508  |  0 (=B)  setting 64k RAM OK
- IO-decode ROM: 444-61
- code ROM: 444-84 for MTR-90
- sec address decoder: 444-83 for MTR-90
- JJ 505 - 508: 1 | wired to  pin14 P508   |  1  |  1 (=B) setting MTR-90 OK
- ROM 444-19,
- RAM 2112 on U23 and U25 

This configuration worked well for a few years

What I did/checked:

- all supply voltages are OK

- all RAMs tested and OK

- 444-84 and 444-19 seemed to be OK, exchanged with EPROM-copies from the web, no difference

- IO-Chip 8250 and Z80 exchanged: no change

- checked all logic signals on the CPU-board with a scope - all signals seem to be OK

- CPU Clock is OK, activity on address- and data lines

My last attempt: 

checking the CPU activity with a logic analyzer. After RESET the CPU executes commands from adr. 0000 from 444-84 and finally gets stuck in an infinite loop - see my disassembly (attachment)

What is going on ? Any help is highly wellcome !

Roland

[Joseph Travis]

You mentioned the supply voltages, but not the outputs of the regulators on the CPU board.  Three of them are TO-220 packages, two are TO-92.  The two SRAMs at U523 and U525 should be 2114.  Note:  It is easier to troubleshoot the CPU board with the various I/O boards removed (to ensure they're not part of the problem).  In the case of the Z89-37 board, you'll have to replace the ICs that were displaced on the CPU board.

1 - Check the inputs and outputs of all of the regulators.  There are four along the top and one along the left side.

2 - Place SW501-5 to the right and power on / reset the H89.  This should execute the built in RAM test.  The LWA should be 377377.  If not, then there is / will be a RAM failure.  In case of failure, the PASS number is the value that was being written and the error message will display the address of the failure and the data which was read.  You can deduce the bad bit(s) and replace the RAM(s) accordingly.

3 - If you weren't able to execute the RAM test, remove the 16K memory board and the top two banks of RAM.  Set the memory jumpers according to the H88/H89/Z89/Z90 Configuration Guide for 16K and repeat Step 2 above.

4 - Once the computer is able to pass the memory test, you can add another bank and repeat Steps 2-3 until all memory is working.

5 - If none of the above works, let us know.  BTW - A closeup photo of the CPU board can be helpful to us.

Good luck!

Joe

--
You received this message because you are subscribed to the Google Groups "SEBHC" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sebhc+un...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/sebhc/ab936db9-a55f-4c0a-aeec-09a16e01d9f3n%40googlegroups.com.

[Roland Langfeld]

Dear Joe,

thankyou for your immediate response !

1) Sorry, I was not clear:I checked all voltages (-12V, -5V, ...) on all voltage regulators: voltages are in spec, and there is no ripple/noise on the output

2) I already tried the SW501-5 switch setting for a  mem-test: no response on the screen, just the blinking cursor.

3)I checked each 4116-chip individually on a chip tester and found all good. Anyhow, I will reduce te RAM on the board 10 16 k and will report...

There is (beside the RAM-expansion) only one additional board attached to the CPU board: the softsector floppy controller. Because the controller is involved in some kind of address-decoding I decided to keep it on the board. I do not have the chip U557 that fits into the position of the flatband cable going to the floppy controller to run the board without the controller.

Pictures of my board are attached

Roland

[Joseph Travis]

I'd prefer a single photo of the front and back, with the boards and cables removed (leave the individual jumper wires in place).  It looks like you have a DIY 16K memory expansion board and it appears to have ICs piggybacked on it.  Front and backside photos of it (removed) would also be helpful.  The two ICs that were displaced on the CPU board can be found on the Z89-37 board.  Use a multimeter to verify that SW501-5 is actually shorted when switched to the right.  The DIP switch is 40+ years old and it may be necessary to switch it back and forth several times to ensure good contact (I encountered this recently).

I'm curious as to which IC tester you used for the 4116s.  The reason I ask is that I've encountered some that don't fully test the DRAMs (don't do refresh tests) and give a false PASS indication.  Another one gave false failures on some TI branded DRAMs.

BTW - You can remove the 2114s at U523 and U525 while testing, just to ensure they aren't part of the problem.

Joe

--
You received this message because you are subscribed to the Google Groups "SEBHC" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sebhc+un...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/sebhc/f6c13ca6-3170-4c26-a6f1-13b6dc6af76bn%40googlegroups.com.

[glenn.f...@gmail.com]

Roland: you say it’s in an infinite loop, but can you be more specific? At what address range is the loop?

 

Do you have a copy of the MTR90 manual? It has the full source listing. If not, it is here:

https://sebhc.github.io/sebhc/documentation/hardware/HZ89/MTR-90_Ref_ROM.zip

 

your disassembly corresponds to the console UART initialization code. See attached snapshot from the listing.

 

If you can be more specific on the range where it is looping that would help us further diagnose the problem.

 

If the ROM startup code is not getting past this point, then it will never examine the DIP switches and run the memory test.

 

• Glenn

--

You received this message because you are subscribed to the Google Groups "SEBHC" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sebhc+un...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/sebhc/ab936db9-a55f-4c0a-aeec-09a16e01d9f3n%40googlegroups.com.

[Roland Langfeld]

Juhu, my problem has been solved !! Thankyou to Joe for his very systematic approach !

- removing the floppy controller, extracting the 2 chips and putting them in place on the CPU board (I didn't know that they have been used there !)

- the problem still exists, now...

- set switch 501-5 to start with the memory test, power-up again

- memory test came up, indicating a problem in pass 200 at 40000 pattern 000

- checking the corresponding chip with my chip tester gave me an "OK" in 5 testruns of 6. (the tester seems to be unreliable as Joe expected)

- replacing the chip (it is in the first 16K)

- being optimistic: setting switch 501-5 back

result:  double beep at start, H:-prompt is there

-running T(est) now from the monitor-prompt - no problems after 377 passes !

- successful boot from my Gotek-Drive

I'm happy ! Thankyou again !

Glenn: to answer your question: just after 30 instructions from start (RESET) the CPU looped at address hex 0428 - 0429. 

Roland

[norberto.collado koyado.com]

It is stuck polling the 8250, to print the H:. Check the 8250 clock to be correct. Replace the 8250 with a 16550 to see if any improvements.

Norby

Get Outlook for iOS

---

From: se...@googlegroups.com <se...@googlegroups.com> on behalf of glenn.f...@gmail.com <glenn.f...@gmail.com>
Sent: Tuesday, February 3, 2026 1:22:53 PM
To: se...@googlegroups.com <se...@googlegroups.com>
Subject: RE: [sebhc] H89 problem: no boot prompt

 

To view this discussion visit https://groups.google.com/d/msgid/sebhc/059b01dc9553%24423efa40%24c6bceec0%24%40gmail.com.

[Roland Langfeld]

Norby: the 8250 has been checked before: clock is OK, and I testwise replaced the chip. The bad RAM at o40000 seemed to cause the problem (see previous post)

Roland

[Joseph Travis]

Congrats!  'Glad you got it going!

Joe 

To view this discussion visit https://groups.google.com/d/msgid/sebhc/fe4af3c3-e93e-427f-bf2a-9d075f892259n%40googlegroups.com.

[norberto.collado koyado.com]

Thanks and great debugging process!

Get Outlook for iOS

---

From: se...@googlegroups.com <se...@googlegroups.com> on behalf of Joseph Travis <jtravi...@gmail.com>
Sent: Tuesday, February 3, 2026 2:11:18 PM
To: se...@googlegroups.com <se...@googlegroups.com>
Subject: Re: [sebhc] H89 problem: no boot prompt

 

To view this discussion visit https://groups.google.com/d/msgid/sebhc/CAGQDgBCmqpJ3_Ke%2B4XPEw5zCiw%2B7wkBwPFdORw2mxtvenn0eAw%40mail.gmail.com.

[Glenn Roberts]

Ironically that was my first suspicion. Bad ram at 40.000 will cause the monitor to not come up. But since you had tested the ram I didn’t say anything. I recall Joe lamenting that the ram tester is not always reliable. Glad you’re up and running!

Sent from my iPad

On Feb 3, 2026, at 5:11 PM, Joseph Travis <jtravi...@gmail.com> wrote:



To view this discussion visit https://groups.google.com/d/msgid/sebhc/CAGQDgBCmqpJ3_Ke%2B4XPEw5zCiw%2B7wkBwPFdORw2mxtvenn0eAw%40mail.gmail.com.