H89 in repair

Last edit Nov 8 2011 (c) Herb Johnson. See my general Web page on Heath-Zenith computers of the 1970's and 80's for more information about my interests. Also see my Web pages on computer restoration. - Herb Johnson


Notes and photos on H89 s/n J208G023. Various dates stamped on boards and cabinet, plus date codes on chips, suggest this computer was built in 1982. I tested the unit in 2009 and it initially worked. Then my notes say "June 2009 did work, then CPU less and less functional". That suggests some RAM failure? A second note says "June 2009 ROM Prompt OK, blew tantalum on center top of CPU". There's a smudge of soot where that cap burned in photo h89_1_2.jpg.

h89_1_1.jpg is photo of H89 from the front with lid up.

h89_1_2.jpg is photo of H89 looking in from the top.

h89_1_3.jpg is photo of H89 looking in from the left front.

h89_1_4.jpg is photo of H89 showing soft sector floppy controller "Z89-37 Disk Controller" card.

h89_1_5.jpg is photo of H89 showing the three cards (soft sectored floppy, serial/parallel, hard sectored floppy) and power supply, from front right.

h89_1_6.jpg is photo of H89 of the top of the power supply.

h89_1_7.jpg is photo of H89 showing the AC transformer area.

h89_1_8.jpg is photo of H89 keyboard. Note chipping on panel under the spacebar.

h89_1_10.jpg is photo of H89 back panel. Connectors for serial and parallel ports, one 5.25-inch floppy cable. The label on the back says "Z90-80 J208G023 LOT 44716".

h89_1_11.jpg is photo of H89 memory expansion card 85-22539-1, and shows CPU board ROM label 444-142.

h89_1_12.jpg is photo of H89 monitor board showing number 85-2553-2 and 082581.

h89_1_13.jpg is photo of H89 CPU and terminal board lables. The CPU label is 0000165-46866 181-3615. Terminal board label is 0000052-S31255 181-3383-2, and also has a date stamp "Jan 27 1982".


I powered up the H89 using a Variac, to monitor DC voltages and check for shorted capacitors. Everything looked OK, and when I gave it full AC voltage, I got the ROM monitor prompt on screen! This version of the monitor is limited. h89_1_14.jpg is photo of the View memory dump, for 0000H up. h89_1_15.jpg is photo of the dump at an apparent boundary of 5FFFH. The memory seems to "stop" at that point, for instance the photo h89_1_16.jpg is a photo of the "test" which tests memory only up to 5FFFH (it does not accept input values). Using the "substitute" command, I can't get memory to accept any values at 6000H and up. The monitor apparently uses RAM from the top of memory down, as changing those values crashes the ROM monitor.

RAM testing Reading the documents for the H89 - they are very extensive, written for rank beginners AND for technically advanced, Heathkits were soldfor technical education - I saw that memory size is set with jumpers. I probably set the RAM low due to apparent memory problems. So I rejumpered the memory setting to 32K and retested - all was well. I used a hair dryer to raise the RAM temperature to 120 degrees F - still ran OK. I repeated these hot-tests at 48K and 64K - no problems.

floppy testing I have a tested 360K half-height floppy drive, 40 tracks double sided. I put that on the H89's hard-sectored controller. The ROM monitor has a "Boot" command. Reading the manual tells me "boot" goes to the primary floppy controller, "boot secondary" to the secondary floppy controller. Then you follow the command with an optional number, for drive 0, 1, 2. With the hard-sectored controller, I had trouble getting the drive "head load" and "drive select" to function, I had to change both the drive's drive select and the command number, to get the drive to become selected (LED comes on, head moves and loads).

My next problem, was a lack of bootable diskettes. I could not locate more than a few disks, hard or soft sectored, that were "bootable" and also ones I could "sacrifice". Old drives and older diskettes, have nasty habits. If the head gets dirty from debris (mold, bits of magnetic oxide), it acts as a scraper and actually scrapes magnetic coatings from the diskette. Old diskettes can lose cohesion, and their media coatings get brittle. That's a bad combination that can damge subsequent GOOD diskettes.

So, I'm stalled from further testing, until I can get good boot disks, HDOS or CP/M, hard and soft sectored. I'm working on getting those from my colleages.

- Herb Johnson

Contact information:

Herb Johnson
New Jersey, USA
To email @ me, see see my ordering Web page.

Copyright © 2011 Herb Johnson