Gold Label Omni Key 102 Questions

User avatar
Polecat

02 Jul 2021, 04:44

Evening update (typing on a real keyboard!). Dug out and opened up the early 102, and not surprisingly found a few surprises. First, this one *does* have a 24 pin EPROM. I didn't pull off the label, but it's a hand-written label, "Omni-102". It's in a 24 pin socket, so it must be an earlier PC board version than yours. PC board is labelled "Omni 102 P/N B0026 19880204". The controller chip is an 8049, not an 8039.

The labels on the bottom include a red Sefco label identical to yours, a silver Sefco label, a paper label, and an inspection number. No revision number on this one. Serial number (paper label) is 119235. Inspection number is 10208302. Date (paper label) is 07-15-1988. Not sure this one does us any good, as far as the EPROM goes, having a different controller and PC board version.

My two other gold label Northgates are from May and June of 1989, and have 28 pin EPROMs in 28 pin sockets. Rev 5.2 on both.

I have one other Gen1, which is a generic, unbadged one, very late (white Alps), with a 10xxxxx inspection number. No Northgate ID or labels, since these generic ones (I believe) weren't sold by Northgate. I'll pop it open and check the EPROM and PC board details, but it's newer than the others, so it's probably not helpful. Ugh, this stuff is never as simple as we want it to be!

I'm wondering out loud if some of the differences in these may be due to the way they were sold originally. Northgate included a keyboard with the computers they sold, but they also sold many of their keyboards separately, by mail order and through computer stores.

edit: Forgot to mention that there's a hand written note on the bottom of the early one that says "may not reset correctly on some AT". Might be my own writing, but if it is it's from so long ago that I don't remember or recognize it. That might affect compatibility with some converters, if true. I'll check that when I have more time.

edit2: Opened up the generic 102, and it has an early PC board (19881006) but has a 28 pin 2764 EPROM in a 28 pin socket. And an 8039 controller chip. To my surprise there's a Focus QC/inspection sticker on the top of the board. But no etched/embossed Focus logo like we've seen on some of these. The plate is stamped "OMNI/101A" (not 102!), and the EPROM has a silver sticker with OMNI printed in red, the same as my May and June 1989 102s. Again this one has no Northgate badge or stickers, and no SEFCO identification, just a warranty sticker and date sticker on the bottom cover. Sounds like your 102 is later than my very early one, but earlier than the other three. That would put it sometime in mid-1988.

User avatar
thefarside

06 Jul 2021, 20:34

Update. I followed jmaynard's advice and tested the SN74L chip, looking to see if pins 16 and 8 had +5 volts. For reference I included a picture of the SN74L chip below and marked pin 16 in a red circle and pin 8 in a blue circle. If I misinterpreted please let me know and i can remeasure.
SN74L chip
SN74L chip
IMG_2386.jpg (663.74 KiB) Viewed 2409 times
I measured the voltages by placing the common on the multimeter to the ground or "G" as indicated on the connector for the main (bottom right of the picture below) wire and placing the red terminal on each pin in the SN74L chip. The pictures below show my method and results for measuring each pin. I measured +5V on pin 16 and almost nothing on pin 8. Below are pictures of my measurements. Would the low voltage reading on pin 8 indicate a problem?
Pin 16 measurement
Pin 16 measurement
IMG_2380.jpg (4.26 MiB) Viewed 2409 times
Pin 8 measurement
Pin 8 measurement
IMG_2382.jpg (3.98 MiB) Viewed 2409 times
Willy4876 suggested checking out the solder joints and PCB and I examined the back and everything looked mostly normal. I did notice some scratches on the back that go across several connections, but the scratches don't look deep and I would think the scratches would only affect the keys with the broken trace. Below is a high resolution picture for reference.
Back of PCB
Back of PCB
IMG_2374.jpg (2.16 MiB) Viewed 2409 times
It occurred to be I'm using two adapters: an AT to PS2 which connects to a PS2 to USB, and I wonder if that could be the problem. I'm going to seek out another adapter to further troubleshoot, but I don't think it's the adapter based on another new observation.

I have two PS2 to USB adapters and while connecting in each one to see if there was any difference, one of them started showing the Scroll Lock light. Pressing the School Lock button would turn the light on and off, but none of the other keys or function lights would appear. After ~10 seconds the Scroll Lock light would turn off and the keyboard wouldn't respond until the adapter was reconnected. Below are pictures of the adapter setup I'm using on the Scroll Lock light.
Adapter Setup
Adapter Setup
IMG_2384-3.jpeg (132.93 KiB) Viewed 2409 times
Scroll Lock Light
Scroll Lock Light
IMG_2385-3.jpeg (254.06 KiB) Viewed 2409 times
I accidentally connected two pins with the multimeter terminal while measuring and noticed the indicator turned on. Sometimes all of them turned on or two out of the three. I'm not sure if that matters, or if I just ruined the chip... :!: Even though the lights turned on the keyboard still didn't respond.

Thanks again for all of the assistance! I'm hoping to continue to rule things out and determine the root cause. My next steps of action are:
1) Borrow an old PC with an AT connector to see if the keyboard works to rule out my converters.
2) Search for info sheets on each chip to find information on the proper voltages and see what info can be found to help debug.
3) Reach out to blindassassin to see if a spare PCB can be purchased.

If anyone has other suggestions just let me know!

User avatar
hellothere

08 Jul 2021, 17:52

Polecat asked me to check my OmniKey "blank label" keyboards that have a PCB that's fairly similar to OPs. Specifically, it's the chip that OP has black electrical tape on. Both of the chips have 28 pins.

User avatar
Polecat

09 Jul 2021, 04:46

hellothere wrote:
08 Jul 2021, 17:52
Polecat asked me to check my OmniKey "blank label" keyboards that have a PCB that's fairly similar to OPs. Specifically, it's the chip that OP has black electrical tape on. Both of the chips have 28 pins.
Thanks for checking! Sounds like we don't have a gold label 102 with a matching EPROM, unless by chance my very early one with the 8049 controller used the same version. Hopefully that isn't the problem.

User avatar
thefarside

01 Oct 2021, 13:07

I have a small update. I’ve gotten the data sheets on the controller chip and plan to perform as much testing as possible. Following another thread of a Focus keyboard issue it looks like I can use a logic analyzer to see the signals coming from the connector and verify the timing and commands are AT. I think I’ll need to know the AT protocol in order to do that and I plan to ask the author of that threat how to obtain the AT protocol. I have access to a logic analyzer at work so this might be an option, as it look like I can verify the clock and data signals coming from the board.

I don’t have access to a native AT computer so I purchased an AT > USB from orihalcon and that didn’t work, but did work perfectly on all my other keyboards, so I’m 99% sure the issue is either in the controller or part on the PCB….

I’m also starting to wonder if I can replace the controller with something new. I’m a novice but it sounds like people use a teensy or other type of small modern replacement controller to send the scan codes (might be the wrong description) over a USB interface. Does that sound like an option for me if I come to the conclusion the controller is the issue?

Thanks again for all of your help!

User avatar
Muirium
µ

01 Oct 2021, 13:18

Indeed, troublesome old controllers can be pulled out and replaced with something friendlier.

A resident expert at controller swaps and reverse engineering is MMcM. Take a peek of some of his work, here:

viewtopic.php?f=7&t=23229

Just one project that came up in a brief search. There are plenty more.

Oh, do check for diodes in your keyboard's matrix. Controllers can't do miracles and make a lousy PCB into NKRO. But if the diodes are there, it's game on.

User avatar
thefarside

16 Jan 2022, 17:19

I have a final update and good news. My Omni Key 102 is working and I am typing this post with it! I decided to reach out to "Northgate Bob" to see if he had any advice and he recommended sending it to him for evaluation. Bob determined the controller was bad and he didn't think it was the correct controller for the keyboard. Luckily he had a spare and replaced it. When I take the case off to clean and retrobright I'll post a picture of the controller he used. He also found a dead F10 switch and replaced the anti-slip pads.

My next steps are to clean and retrobright the case. I retrobrighted the keys and they came out great, so I'm hoping to get similar results with the case. Below are some before and after pictures of the keys. I have a lot of the liquid version of salon care so my plan is to completely submerge the case and use a UV light. To be safe I'm going to make a high resolution scan of the labels and cover them with tape just in case the retrobright process alters the colors.

Before retrobright:
Before retrobright
Before retrobright
IMG_3594.jpg (113.22 KiB) Viewed 1984 times
After retrobright:
After retrobright
After retrobright
IMG_3614.jpg (110.71 KiB) Viewed 1984 times
Thanks again to everyone for your help and information!

User avatar
Willy4876

17 Jan 2022, 15:09

I'm glad to hear it. Congrats and enjoy the keyboard.

User avatar
Polecat

23 Jan 2022, 04:34

thefarside wrote:
16 Jan 2022, 17:19
I have a final update and good news. My Omni Key 102 is working and I am typing this post with it! I decided to reach out to "Northgate Bob" to see if he had any advice and he recommended sending it to him for evaluation. Bob determined the controller was bad and he didn't think it was the correct controller for the keyboard. Luckily he had a spare and replaced it. When I take the case off to clean and retrobright I'll post a picture of the controller he used. He also found a dead F10 switch and replaced the anti-slip pads.

My next steps are to clean and retrobright the case. I retrobrighted the keys and they came out great, so I'm hoping to get similar results with the case. Below are some before and after pictures of the keys. I have a lot of the liquid version of salon care so my plan is to completely submerge the case and use a UV light. To be safe I'm going to make a high resolution scan of the labels and cover them with tape just in case the retrobright process alters the colors.

Thanks again to everyone for your help and information!
Congrats on your Northgate and thanks for the update. And three cheers for Northgate Bob, he's saved the day more than once. I'm still trying to determine how many PC board versions there were on the Gen1 102 models (and also the Gen2s), and what exactly is different about them? There was one on ebay a couple months back with the 24 pin EPROM in the 28 pin socket, but the PC board appeared to be a different version. I'll keep logging notes and photos in hopes of it all making sense some day. Strange that Bob thought the controller was the wrong one, being as it was soldered in place, unless he meant the EPROM? Anyway, I'm really glad to hear it's working and being enjoyed.

User avatar
thefarside

26 Feb 2022, 03:25

Polecat wrote:
23 Jan 2022, 04:34
Congrats on your Northgate and thanks for the update. And three cheers for Northgate Bob, he's saved the day more than once. I'm still trying to determine how many PC board versions there were on the Gen1 102 models (and also the Gen2s), and what exactly is different about them? There was one on ebay a couple months back with the 24 pin EPROM in the 28 pin socket, but the PC board appeared to be a different version. I'll keep logging notes and photos in hopes of it all making sense some day. Strange that Bob thought the controller was the wrong one, being as it was soldered in place, unless he meant the EPROM? Anyway, I'm really glad to hear it's working and being enjoyed.
Below is a picture of the bad 24 pin controller Bob removed and replaced with a 28 pin. I’m not sure if it’s completely dead or if it’s possible to repair. I think it’s the EPROM because it isn’t soldered in place. I don’t have any use for it and if you’d like it let me know.
Controller.jpg
Controller.jpg (497.25 KiB) Viewed 1657 times

User avatar
Polecat

26 Feb 2022, 04:17

thefarside wrote:
26 Feb 2022, 03:25

Below is a picture of the bad 24 pin controller Bob removed and replaced with a 28 pin. I’m not sure if it’s completely dead or if it’s possible to repair. I think it’s the EPROM because it isn’t soldered in place. I don’t have any use for it and if you’d like it let me know.
Controller.jpg
Yes, that's the EPROM. The controller is a 40 pin chip, either an 8039 or 8049. There's probably a jumper to select the EPROM size. They usually don't go bad electrically, but they can lose their programming if exposed to UV light with the label removed. That's how you erase them before reprogramming. Thanks for the offer, but the only reason I'd want it would be to save the programming, and if it's bad that wouldn't be possible. Good to know the 28 pin (2764) will work in place of the 24 pin (2732) (possibly with a jumper changed?).

Post Reply

Return to “Keyboards”