Gold Label Omni Key 102 Questions
- thefarside
- Location: United States
- Main keyboard: IBM 4704 F107
- Main mouse: Old, boring Logitech
- Favorite switch: Buckling spring
I’m the new owner of an Omni Key 102 and I wanted to ask a couple questions. Overall it’s in great shape. There’s some yellowing but it’s clean the switches feel great. Here’s a picture:
1) It doesn’t power on (nothing on the lock lights and no key response) when I plug it in using a variety of working converters. I tried passive and active. Besides checking the capacitors is there anything else I can check?
2) The switches feel great and are very clean, but I was curious if I could do anything to ”help” them. I know to leave clean alps alone, but was curious if blowing air on them from the outside or inside would help clear our old dust. Another picture for reference: 3) Has anyone retrobrighted an Omni Key 102 case? I’m curious if I should leave it alone or if retrobright would be a safe restoration method. I’ve only done one retrobright on an FK-2001 case and keys and it seemed to have a marginal improvement with no impact on the case or labels.
Any advice is appreciated. Thanks in advance!
My questions are:1) It doesn’t power on (nothing on the lock lights and no key response) when I plug it in using a variety of working converters. I tried passive and active. Besides checking the capacitors is there anything else I can check?
2) The switches feel great and are very clean, but I was curious if I could do anything to ”help” them. I know to leave clean alps alone, but was curious if blowing air on them from the outside or inside would help clear our old dust. Another picture for reference: 3) Has anyone retrobrighted an Omni Key 102 case? I’m curious if I should leave it alone or if retrobright would be a safe restoration method. I’ve only done one retrobright on an FK-2001 case and keys and it seemed to have a marginal improvement with no impact on the case or labels.
Any advice is appreciated. Thanks in advance!
- Bjerrk
- Location: Copenhagen, Denmark
- Main keyboard: Cherry G80-1800 & Models F & M
- Main mouse: Mouse Keys, Trackpoint, Trackball
- Favorite switch: IBM Buckling Springs+Beamspring, Alps Plate Spring
Wait, who told you to leave Alps switches alone?
I'd say the number one priority is to prevent them from being used in an at all dirty state. Or, put differently: if you're gonna use them, why not give them a good dusting off first?
You can just open them in-plate and give their insides a good dust removal (either with pressurised air or a brush).
Other than that, I probably wouldn't do anything to them, really.
I'd say the number one priority is to prevent them from being used in an at all dirty state. Or, put differently: if you're gonna use them, why not give them a good dusting off first?
You can just open them in-plate and give their insides a good dust removal (either with pressurised air or a brush).
Other than that, I probably wouldn't do anything to them, really.
- TNT
- Location: Germany, Karlsruhe
- Main keyboard: Ellipse Model F77 / Zenith Z-150
- Main mouse: Logitech G203 Prodigy
- Favorite switch: It's complicated
- DT Pro Member: 0250
Will send pics later, retrobrighting works great with those (cream method)thefarside wrote: ↑28 May 2021, 23:19
3) Has anyone retrobrighted an Omni Key 102 case? I’m curious if I should leave it alone or if retrobright would be a safe restoration method. I’ve only done one retrobright on an FK-2001 case and keys and it seemed to have a marginal improvement with no impact on the case or labels.
Any advice is appreciated. Thanks in advance!
- thefarside
- Location: United States
- Main keyboard: IBM 4704 F107
- Main mouse: Old, boring Logitech
- Favorite switch: Buckling spring
Thanks! Regarding the switches, my thought was to get rid of dust to prevent any damage or binding from any dust or dirt that would be in there. So it’s ok to open the top housing and give the inside a blast of compressed air?
Below are some more pictures for reference. I wasn’t sure if the factory lube could come off if I hit it with air.
I have a very dirty blue alps Focus FK-555 and I’m hoping to avoid that with this.
The top housing looked pretty clean: Bottom Housing: Stem with factory lube:
Below are some more pictures for reference. I wasn’t sure if the factory lube could come off if I hit it with air.
I have a very dirty blue alps Focus FK-555 and I’m hoping to avoid that with this.
The top housing looked pretty clean: Bottom Housing: Stem with factory lube:
- Polecat
- Location: Downstream from Silicon Valley
- Main keyboard: Monterey K104 Industrial Gray
- Main mouse: Logitech Optical
- Favorite switch: Early Alps SKCM
- DT Pro Member: -
Check the XT/AT switch, or just move it back and forth several times in case the contacts are dirty or oxidized. Presumably you want it set to AT, but even if it's set to the wrong position you'll usually get something on the lock lights at least.thefarside wrote: ↑28 May 2021, 23:19
1) It doesn’t power on (nothing on the lock lights and no key response) when I plug it in using a variety of working converters. I tried passive and active. Besides checking the capacitors is there anything else I can check?
Also check the cable. It's not uncommon to have one or more broken wires on Focus or Northgate keyboards. If you have an ohmmeter check that each wire on the internal connector connects to a pin on the plug.
- TNT
- Location: Germany, Karlsruhe
- Main keyboard: Ellipse Model F77 / Zenith Z-150
- Main mouse: Logitech G203 Prodigy
- Favorite switch: It's complicated
- DT Pro Member: 0250
My cam really doesn't catch how yellow it was when I got it, but trust me, it was really bad.
Before:
After:
It's still waiting for its blue Alps to arrive and make it whole
Before:
Spoiler:
After:
Spoiler:
- thefarside
- Location: United States
- Main keyboard: IBM 4704 F107
- Main mouse: Old, boring Logitech
- Favorite switch: Buckling spring
Thanks TNT it looks great! You used the cream method? I purchased a few bottles of salon care developer 40 clear thinking I could completely submerge the parts and leave them in the sun. When I did this with a lightly yellowed FK-2001 it seemed to work on the keys, but didn’t seem to do much with the case.
For the Northgate I was worried submerging the case might ruin the labels, but I’m open to using the cream method if that’s safer.
For the Northgate I was worried submerging the case might ruin the labels, but I’m open to using the cream method if that’s safer.
- Rami
- Location: United States
- Main keyboard: Apple Extended Keyboard. Model M0115
- Main mouse: Mo-159U
- Favorite switch: Orange Alps
- Contact:
I've retrobrighted key caps but never a full chassis, for the keys I used the submersion method. As for the electrical issues it might be the AT/XT switch as others said. It could also be a leaky cap, those can cause problems but as long as they don't ruin any traces it's an easy fix. Excellent pick-up! This is one of the best keyboards in my opinion!
- TNT
- Location: Germany, Karlsruhe
- Main keyboard: Ellipse Model F77 / Zenith Z-150
- Main mouse: Logitech G203 Prodigy
- Favorite switch: It's complicated
- DT Pro Member: 0250
Labels should be fine, I was more concerned about the screw inserts since those are brass and metal stuff really shouldn't come into contact with H2O2...thefarside wrote: ↑29 May 2021, 13:21Thanks TNT it looks great! You used the cream method? I purchased a few bottles of salon care developer 40 clear thinking I could completely submerge the parts and leave them in the sun. When I did this with a lightly yellowed FK-2001 it seemed to work on the keys, but didn’t seem to do much with the case.
For the Northgate I was worried submerging the case might ruin the labels, but I’m open to using the cream method if that’s safer.
- TNT
- Location: Germany, Karlsruhe
- Main keyboard: Ellipse Model F77 / Zenith Z-150
- Main mouse: Logitech G203 Prodigy
- Favorite switch: It's complicated
- DT Pro Member: 0250
- thefarside
- Location: United States
- Main keyboard: IBM 4704 F107
- Main mouse: Old, boring Logitech
- Favorite switch: Buckling spring
Thanks I was excited to get this and I agree, the quality of the chassis and switch make for a great experience. I also like the F keys on the left and feel it’s more ergonomic for me. I’ll be sure to take pictures of my progress and update this thread as I go. Thanks again to everyone for the advice!Rami wrote: ↑29 May 2021, 16:25I've retrobrighted key caps but never a full chassis, for the keys I used the submersion method. As for the electrical issues it might be the AT/XT switch as others said. It could also be a leaky cap, those can cause problems but as long as they don't ruin any traces it's an easy fix. Excellent pick-up! This is one of the best keyboards in my opinion!
- thefarside
- Location: United States
- Main keyboard: IBM 4704 F107
- Main mouse: Old, boring Logitech
- Favorite switch: Buckling spring
I’ve carefully disassembled it and overall it’s pretty clean and the cables, connectors and PCB look good:
My last question is regarding whether to disassemble each switch and spray with compressed air or is it ok to spray in the gaps between the slider with the switches assembled? My concern was whether I’d be blowing any lube from the stems or housing. Below is a picture for reference:
BTW the internal connector for the cable was very difficult to disconnect! It looks similar to the one on the Focus FK-2001. Has anyone else experienced issues with the internal connector?My last question is regarding whether to disassemble each switch and spray with compressed air or is it ok to spray in the gaps between the slider with the switches assembled? My concern was whether I’d be blowing any lube from the stems or housing. Below is a picture for reference:
- Polecat
- Location: Downstream from Silicon Valley
- Main keyboard: Monterey K104 Industrial Gray
- Main mouse: Logitech Optical
- Favorite switch: Early Alps SKCM
- DT Pro Member: -
Have you found and fixed the problem yet? I would recommend that before cleaning the switches. It looks very clean already in the photos. I *strongly* recommend putting something over the window on the EPROM! Stray UV or even sunlight can erase the contents with it uncovered. Hopefully that isn't what's wrong. A piece of opaque sticker material or black electrical tape is all you need there.
- thefarside
- Location: United States
- Main keyboard: IBM 4704 F107
- Main mouse: Old, boring Logitech
- Favorite switch: Buckling spring
Thanks for the EPROM warning! It’s been by the window for a day but not in direct sunlight but I’ve covered it to be safe. Hopefully that’s not the issue. I haven’t figured out the issue yet, but my plan was to do the following:
1) Test both ends of the cable with a multimeter.
2) Test the capacitors.
3) Assuming 1 and 2 tests pass, plug the keyboard in to my computer using passive and active adapters, flipping between the AT/XT protocols.
If there’s anything else I can test just let me know. It looks like there’s a lot of places where a connection could fail. I’m going super slow with this, so I don’t mess anything up or lose anything. I’ll leave the switches alone and appreciate all the help!
1) Test both ends of the cable with a multimeter.
2) Test the capacitors.
3) Assuming 1 and 2 tests pass, plug the keyboard in to my computer using passive and active adapters, flipping between the AT/XT protocols.
If there’s anything else I can test just let me know. It looks like there’s a lot of places where a connection could fail. I’m going super slow with this, so I don’t mess anything up or lose anything. I’ll leave the switches alone and appreciate all the help!
- Polecat
- Location: Downstream from Silicon Valley
- Main keyboard: Monterey K104 Industrial Gray
- Main mouse: Logitech Optical
- Favorite switch: Early Alps SKCM
- DT Pro Member: -
Yes, those are the tests to do first. Bad connections are usually caused by moisture/corrosion or physical damage, and there's no reason to suspect either from what I can see. As clean as that keyboard looks I wouldn't be surprised if someone had cleaned it already. Personally I would just brush off any visible dust with a clean paint brush or similar and call it good. Compressed air seems excessive to me.
- thefarside
- Location: United States
- Main keyboard: IBM 4704 F107
- Main mouse: Old, boring Logitech
- Favorite switch: Buckling spring
Small update. I tested the cable for continuity and everything passed. I connected one probe of my multimeter to each terminal on the connector and put the other probe on each wire on the other end and there were no issues. I wiggled the cable to be safe.
Below are pictures of my testing of four capacitors and the resistance measurements, which were consistent. I’m guessing they need to be desoldered to be tested individually, but I’m open to any suggestions. I also noticed a broken trace. I’m hoping this is the root cause of the board not working and my plan is to solder a wire to restore the connection. Below is a picture with the broken trace highlighted in red.
I then tested the capacitors, which visually looked ok. There are blue cylinders, small pads and a silver metal square version. I put the multimeter into Ohms to test resistance, hoping to charge each capacitor until it was no longer providing resistance. I noticed each one was maintaining a constant level of resistance and could not be “filled”. I think they can’t be tested while connected to the PCB, since they are still connected to other circuits. Does that make sense? Is there a way to test capacitors while connected to an electrical path in a circuit board? I’m a novice and any advice is appreciated. Below are pictures of my testing of four capacitors and the resistance measurements, which were consistent. I’m guessing they need to be desoldered to be tested individually, but I’m open to any suggestions. I also noticed a broken trace. I’m hoping this is the root cause of the board not working and my plan is to solder a wire to restore the connection. Below is a picture with the broken trace highlighted in red.
-
- Location: US
- Main keyboard: IBM M122, 3270-style
- Main mouse: Apple Magic Trackpad 2
- Favorite switch: IBM 3278/3279 beamspring
I wouldn't be so quick to assume the broken trace is the cause. That looks like a configuration jumper, and may or may not be supposed to be there.
I'd plug it in and make sure there's +5 between pins 16 (across from the dimple) and 8 (on the same side as the dimple, on the other end) of that SN74L-whatever that chip is in between the two caps.
I'd plug it in and make sure there's +5 between pins 16 (across from the dimple) and 8 (on the same side as the dimple, on the other end) of that SN74L-whatever that chip is in between the two caps.
- Polecat
- Location: Downstream from Silicon Valley
- Main keyboard: Monterey K104 Industrial Gray
- Main mouse: Logitech Optical
- Favorite switch: Early Alps SKCM
- DT Pro Member: -
Agreed that the caps can't be tested in-circuit with an ohmmeter. There are in-circuit cap testers that apply a very low voltage signal, but a good one would cost you more than the keyboard.thefarside wrote: ↑29 Jun 2021, 04:37Small update. I tested the cable for continuity and everything passed. I connected one probe of my multimeter to each terminal on the connector and put the other probe on each wire on the other end and there were no issues. I wiggled the cable to be safe.
I then tested the capacitors, which visually looked ok. There are blue cylinders, small pads and a silver metal square version. I put the multimeter into Ohms to test resistance, hoping to charge each capacitor until it was no longer providing resistance. I noticed each one was maintaining a constant level of resistance and could not be “filled”. I think they can’t be tested while connected to the PCB, since they are still connected to other circuits. Does that make sense? Is there a way to test capacitors while connected to an electrical path in a circuit board? I’m a novice and any advice is appreciated.
Below are pictures of my testing of four capacitors and the resistance measurements, which were consistent. I’m guessing they need to be desoldered to be tested individually, but I’m open to any suggestions.
I also noticed a broken trace. I’m hoping this is the root cause of the board not working and my plan is to solder a wire to restore the connection. Below is a picture with the broken trace highlighted in red.
An old photo of my 102 shows no jumper in that position:
download/file.php?id=54738
jmaynard is probably correct about it being a configuration jumper. Possibly something like internal or external EPROM for the controller IC. Strange that it looks like it did have something there at one time. Also strange that mine has tantalum or monoceramic caps instead of the electrolytics. Far as I know it's original, as I've owned it almost since new.
- Willy4876
- Location: Ottawa, Ontario, Canada
- Main keyboard: Focus FK-2001 with Alps Doubleshots
- Main mouse: Logitek MX Master 2S
- Favorite switch: Blue Alps
- DT Pro Member: -
Is this the silver component that you're talking about? If it is then that's not a capacitor. That's the oscillator for the microcontroller. It is possible for crystal oscillators to go bad but I've never seen it.thefarside wrote: ↑29 Jun 2021, 04:37I then tested the capacitors, which visually looked ok. There are blue cylinders, small pads and a silver metal square version.
Spoiler:
Sometimes there are a couple of capacitors that are used either for timing, or in an oscillator to generate the clock for the controller. These tend not to be electrolytic capacitors (the type in cans that tend to leak and short). While the values for these capacitors are critical, they don't normally fail.
In my experience most of the time I have a keyboard that isn't working it's either a bad solder joint or physical damage. Look at the solder joints. If they look a little black, or cracked, or even just different from the rest, then you can use a soldering iron to reflow the solder. Look to see if there is any corrosion on the traces of the board, or if any traces are broken. Look at the components and make sure that none of them have broken legs or other damage.
The uncovered window on the EPROM is really not a good thing, but nothing to be done about that now. If the data on it is corrupted or erased then the microcontroller will not know what it's supposed to be doing, and thus won't do it. In that case, your options would be to replace/reprogram the EPROM with one with the proper data on it, or to replace the controller all together. I'm not sure if the image (data) of the EPROM is known, but if it's not then someone with one of these boards might be able to get an image of it for you.
- Polecat
- Location: Downstream from Silicon Valley
- Main keyboard: Monterey K104 Industrial Gray
- Main mouse: Logitech Optical
- Favorite switch: Early Alps SKCM
- DT Pro Member: -
All good advice. I have an old DOS-based EPROM reader/burner here someplace, and three or four gold label Northgates with socketed EPROMs, so assuming one of my old DOS computers still boots up I should be able to read the code from one of mine. I'm not sure if the file format is the same to translate to another EPROM burner, but I'm willing to give it a try if the EPROM turns out to be the problem.Willy4876 wrote: ↑29 Jun 2021, 22:40Is this the silver component that you're talking about? If it is then that's not a capacitor. That's the oscillator for the microcontroller. It is possible for crystal oscillators to go bad but I've never seen it.thefarside wrote: ↑29 Jun 2021, 04:37I then tested the capacitors, which visually looked ok. There are blue cylinders, small pads and a silver metal square version.
Other users have already mentioned that there isn't a practical way to test capacitors in circuit, but you can check if the capacitors are shorted (have started acting more like resistors). You can check this by measuring resistance across the capacitors. When you first touch the probes to the capacitor legs you should see a very low resistance that will climb to a high resistance. As long as the resistance doesn't stay low the caps aren't shorted. If they're not shorted, the caps probably are not your problem. This is because most of the time, on keyboards the caps are there to make the 5 V supply to the chips more stable. How much capacitance there is is not critically important. Without properly working caps you might get instability, but the board would generally work for the most part.
Sometimes there are a couple of capacitors that are used either for timing, or in an oscillator to generate the clock for the controller. These tend not to be electrolytic capacitors (the type in cans that tend to leak and short). While the values for these capacitors are critical, they don't normally fail.
In my experience most of the time I have a keyboard that isn't working it's either a bad solder joint or physical damage. Look at the solder joints. If they look a little black, or cracked, or even just different from the rest, then you can use a soldering iron to reflow the solder. Look to see if there is any corrosion on the traces of the board, or if any traces are broken. Look at the components and make sure that none of them have broken legs or other damage.
The uncovered window on the EPROM is really not a good thing, but nothing to be done about that now. If the data on it is corrupted or erased then the microcontroller will not know what it's supposed to be doing, and thus won't do it. In that case, your options would be to replace/reprogram the EPROM with one with the proper data on it, or to replace the controller all together. I'm not sure if the image (data) of the EPROM is known, but if it's not then someone with one of these boards might be able to get an image of it for you.
- thefarside
- Location: United States
- Main keyboard: IBM 4704 F107
- Main mouse: Old, boring Logitech
- Favorite switch: Buckling spring
Thanks to everyone for your help! I’ve also been wondering if the EPROM is the source of the problem and if not it’s definitely a single point of failure. I like the idea of having a backup of the EPROM image and would appreciate getting a copy of possible.Polecat wrote: ↑30 Jun 2021, 05:44All good advice. I have an old DOS-based EPROM reader/burner here someplace, and three or four gold label Northgates with socketed EPROMs, so assuming one of my old DOS computers still boots up I should be able to read the code from one of mine. I'm not sure if the file format is the same to translate to another EPROM burner, but I'm willing to give it a try if the EPROM turns out to be the problem.Willy4876 wrote: ↑29 Jun 2021, 22:40Is this the silver component that you're talking about? If it is then that's not a capacitor. That's the oscillator for the microcontroller. It is possible for crystal oscillators to go bad but I've never seen it.thefarside wrote: ↑29 Jun 2021, 04:37I then tested the capacitors, which visually looked ok. There are blue cylinders, small pads and a silver metal square version.
Other users have already mentioned that there isn't a practical way to test capacitors in circuit, but you can check if the capacitors are shorted (have started acting more like resistors). You can check this by measuring resistance across the capacitors. When you first touch the probes to the capacitor legs you should see a very low resistance that will climb to a high resistance. As long as the resistance doesn't stay low the caps aren't shorted. If they're not shorted, the caps probably are not your problem. This is because most of the time, on keyboards the caps are there to make the 5 V supply to the chips more stable. How much capacitance there is is not critically important. Without properly working caps you might get instability, but the board would generally work for the most part.
Sometimes there are a couple of capacitors that are used either for timing, or in an oscillator to generate the clock for the controller. These tend not to be electrolytic capacitors (the type in cans that tend to leak and short). While the values for these capacitors are critical, they don't normally fail.
In my experience most of the time I have a keyboard that isn't working it's either a bad solder joint or physical damage. Look at the solder joints. If they look a little black, or cracked, or even just different from the rest, then you can use a soldering iron to reflow the solder. Look to see if there is any corrosion on the traces of the board, or if any traces are broken. Look at the components and make sure that none of them have broken legs or other damage.
The uncovered window on the EPROM is really not a good thing, but nothing to be done about that now. If the data on it is corrupted or erased then the microcontroller will not know what it's supposed to be doing, and thus won't do it. In that case, your options would be to replace/reprogram the EPROM with one with the proper data on it, or to replace the controller all together. I'm not sure if the image (data) of the EPROM is known, but if it's not then someone with one of these boards might be able to get an image of it for you.
I started looking in to copying the image file from the EPROM and it sounds like a common method is to connect to the JTAG interface on the EPROM (if it has one and is accessible) and from there you can use a program to copy the contents of the image. Here’s a link to the process:
https://embeddedbits.org/2020-02-20-ext ... sing-jtag/
I put electrical tape over the window so hopefully that will protect the EPROM from future UV exposure.
My next course of action is to:
1) Measure the voltages between pins 16 (across from the dimple) and 8 (on the same side as the dimple.
2) While plugged in test the XT/AT switch by switching quickly to see if that helps it initialize.
I’m wondering if my adapter isn’t really active so I’m going to look into getting a high quality active version. If anyone has recommendations please let me know.
It also occurred to me that I’m testing a circuit board while not discharging myself or being grounded. I’m also going to invest in a grounding strap and decide if I should get an anti static mat. It would be a shame for me to ruin a keyboard by zapping it. I think the risk is low but I’d hate to learn the hard way.
- Willy4876
- Location: Ottawa, Ontario, Canada
- Main keyboard: Focus FK-2001 with Alps Doubleshots
- Main mouse: Logitek MX Master 2S
- Favorite switch: Blue Alps
- DT Pro Member: -
That board is old enough that it really isn't likely to have JTAG. That's almost certainly a parallel EPROM. Parallel EPROMs use a much simpler interface. Generally, the minimum setup that it takes to read one is to set the output enable pin to output, and set the chip enable pin to enable, and give it an address on its address lines. It is possible to read the entire EPROM this way but would take a ridiculous amount of time. If you use something like an Arduino though, it's simple enough to make an EPROM reader or to make an EPROM programmer. With older EPROMs you often need higher voltages (I've seen them range from 7V to 22V) when you're programming them. If you determine that the EPROM is the issue you should try to find a datasheet for it to see what you would actually need. Normally, you can just google the writing on the chip and find one without much trouble.thefarside wrote: ↑30 Jun 2021, 13:49
...
Thanks to everyone for your help! I’ve also been wondering if the EPROM is the source of the problem and if not it’s definitely a single point of failure. I like the idea of having a backup of the EPROM image and would appreciate getting a copy of possible.
I started looking in to copying the image file from the EPROM and it sounds like a common method is to connect to the JTAG interface on the EPROM (if it has one and is accessible) and from there you can use a program to copy the contents of the image. Here’s a link to the process:
https://embeddedbits.org/2020-02-20-ext ... sing-jtag/
I put electrical tape over the window so hopefully that will protect the EPROM from future UV exposure.
My next course of action is to:
1) Measure the voltages between pins 16 (across from the dimple) and 8 (on the same side as the dimple.
2) While plugged in test the XT/AT switch by switching quickly to see if that helps it initialize.
I’m wondering if my adapter isn’t really active so I’m going to look into getting a high quality active version. If anyone has recommendations please let me know.
It also occurred to me that I’m testing a circuit board while not discharging myself or being grounded. I’m also going to invest in a grounding strap and decide if I should get an anti static mat. It would be a shame for me to ruin a keyboard by zapping it. I think the risk is low but I’d hate to learn the hard way.
Yes, checking pin 8-16 is definitely a good test and it's never a bad idea to try a couple of converters. I have a couple of boards that are picky about which converter I use, so I just keep a couple of generic ones on hand.
As to the ESD strap, it's never a bad idea to have one. They're relatively cheap and give some peace of mind. I'm generally too lazy to get mine out, and I haven't had any issues (knock on wood), but it's never a bad idea to wear one when you're working on this type of thing. As to the antistatic mat, If you get one look into how to properly ground them. I haven't ever set one up, so I can't give you any specific advice on that. Just do a quick youtube/google search on it and you should be fine. I think it's relatively simple.
- Polecat
- Location: Downstream from Silicon Valley
- Main keyboard: Monterey K104 Industrial Gray
- Main mouse: Logitech Optical
- Favorite switch: Early Alps SKCM
- DT Pro Member: -
Definitely no JTAG on these keyboards. It's a standard (parallel) EPROM, and my reader/programmer handles all of those. But it's a major effort to set it up, so please do the other testing first to hopefully identify a simple problem. Note that yours is a 24 pin version (2708/2716/2732) while mine is a 28 pin (2764/27128/27256/27512). I don't remember which mine is, and I can't read the number on yours. Your EPROM is properly installed, shifted away from the notch (24 pin in a 28 pin socket). The missing jumper wire might have something to do with the EPROM size/capacity. Is there a version number on the bottom label of your Northgate?thefarside wrote: ↑30 Jun 2021, 13:49
Thanks to everyone for your help! I’ve also been wondering if the EPROM is the source of the problem and if not it’s definitely a single point of failure. I like the idea of having a backup of the EPROM image and would appreciate getting a copy of possible.
I started looking in to copying the image file from the EPROM and it sounds like a common method is to connect to the JTAG interface on the EPROM (if it has one and is accessible) and from there you can use a program to copy the contents of the image. Here’s a link to the process:
https://embeddedbits.org/2020-02-20-ext ... sing-jtag/
I put electrical tape over the window so hopefully that will protect the EPROM from future UV exposure.
My next course of action is to:
1) Measure the voltages between pins 16 (across from the dimple) and 8 (on the same side as the dimple.
2) While plugged in test the XT/AT switch by switching quickly to see if that helps it initialize.
I’m wondering if my adapter isn’t really active so I’m going to look into getting a high quality active version. If anyone has recommendations please let me know.
It also occurred to me that I’m testing a circuit board while not discharging myself or being grounded. I’m also going to invest in a grounding strap and decide if I should get an anti static mat. It would be a shame for me to ruin a keyboard by zapping it. I think the risk is low but I’d hate to learn the hard way.
I don't think the XT/AT switch will work as a "hot swap". As far as I know you have to reboot the computer with the switch in the desired position for it to work. Nothing is saved in the keyboard, but it has to recognized by the computer when booting in its proper mode. Most converters (Soarers is the exception) will need the switch in the AT mode for the keyboard to be recognized.
-
- Location: Des Moines / Cedar Falls, IA, USA
- Main keyboard: IBM Model F107
- DT Pro Member: 0190
Very interesting thread! If With all that has already been tried, if it was me, I'd look to source a working PCB from a later model with the same layout (which looks like you have the most common one). Lots of these have been parted out over the years, and I would bet some some users still have those PCBs that are just collecting dust. Only thing that a later PCB might not have is the AT/XT switch. You would have to desolder all your switches, but that should take far less time than chasing down the gremlins that have so far been hiding.
- thefarside
- Location: United States
- Main keyboard: IBM 4704 F107
- Main mouse: Old, boring Logitech
- Favorite switch: Buckling spring
No worries regarding the image file. Patience is one of my virtues, which is why this project has taken me so long.
Attached below are some high resolution close ups of all of the chips. I hope this helps and I have a stupid question. For each chip is there a standard image/firmware file that enables it to function? I was under the impression a company purchases the chips for their hardware performance characteristics and writes their own firmware to enable the features required.
I do see most vintage keyboards have an “Intel 1980” chip and it would make sense that Intel made a chip to perform a specific function for something like a keyboard and the companies purchase that same chip to meet their keyboard requirements.
It would also make sense that chip manufacturers provide purpose built chips and it’s less likely each keyboard manufacturer is writing their own custom firmware, other than keyboards that have special features.
Chip closeups:
Attached below are some high resolution close ups of all of the chips. I hope this helps and I have a stupid question. For each chip is there a standard image/firmware file that enables it to function? I was under the impression a company purchases the chips for their hardware performance characteristics and writes their own firmware to enable the features required.
I do see most vintage keyboards have an “Intel 1980” chip and it would make sense that Intel made a chip to perform a specific function for something like a keyboard and the companies purchase that same chip to meet their keyboard requirements.
It would also make sense that chip manufacturers provide purpose built chips and it’s less likely each keyboard manufacturer is writing their own custom firmware, other than keyboards that have special features.
Chip closeups:
- thefarside
- Location: United States
- Main keyboard: IBM 4704 F107
- Main mouse: Old, boring Logitech
- Favorite switch: Buckling spring
Thanks! I considered getting a replacement PCB and thought it would be hard to source one, but would be happy to purchase one that would be compatible. I considered reaching out to “Northgate Bob” but thought I’d see how this pans out first. I’m learning a lot and my plan is to continue troubleshooting to determine if it’s the PCB and if so, if it can be fixed. If not I’ll try to replace it. I like having spare parts so I might pursue a replacement PCB just to be safe, as I plan to keep this keyboard.orihalcon wrote: ↑01 Jul 2021, 13:20Very interesting thread! If With all that has already been tried, if it was me, I'd look to source a working PCB from a later model with the same layout (which looks like you have the most common one). Lots of these have been parted out over the years, and I would bet some some users still have those PCBs that are just collecting dust. Only thing that a later PCB might not have is the AT/XT switch. You would have to desolder all your switches, but that should take far less time than chasing down the gremlins that have so far been hiding.
- Polecat
- Location: Downstream from Silicon Valley
- Main keyboard: Monterey K104 Industrial Gray
- Main mouse: Logitech Optical
- Favorite switch: Early Alps SKCM
- DT Pro Member: -
The two 74LSxxx chips are just generic TTL chips, no firmware or programming. The 8039 controller is likely also not specially programmed being as there is an external EPROM. The EPROM is a 2732, and contains firmware to support the controller. That one will be specific to the Northgate, and there were different revision levels. Two of my Gen1 102s have a bigger EPROM than yours, as evidenced by them being 28 pin chips. Probably 2764, but I haven't confirmed that. That's why I asked if there is a version number on the bottom label. Likely to need to change jumper wires for the 2764 version, and possibly the controller chip itself. I can check my other Gen1 102s to see if they have a 2732.thefarside wrote: ↑01 Jul 2021, 13:21No worries regarding the image file. Patience is one of my virtues, which is why this project has taken me so long.
Attached below are some high resolution close ups of all of the chips. I hope this helps and I have a stupid question. For each chip is there a standard image/firmware file that enables it to function? I was under the impression a company purchases the chips for their hardware performance characteristics and writes their own firmware to enable the features required.
I do see most vintage keyboards have an “Intel 1980” chip and it would make sense that Intel made a chip to perform a specific function for something like a keyboard and the companies purchase that same chip to meet their keyboard requirements.
It would also make sense that chip manufacturers provide purpose built chips and it’s less likely each keyboard manufacturer is writing their own custom firmware, other than keyboards that have special features.
The PC board is different on all the later Northgates. Gen2 Northgates had a separate PC board for the controller electronics, connected to the main board with a ribbon cable. Gen3 had a single board again, but neither Gen2 or Gen3 are dimensionally compatible with Gen1. The spacing between key blocks is different, as are the mounting holes. And the layouts are slightly different. See Blindassassin's thread for much more on the Northgate PC boards. There were at least two or three different PC boards used in the Gen1 Northgates, but they were *probably* all physically compatible.
- thefarside
- Location: United States
- Main keyboard: IBM 4704 F107
- Main mouse: Old, boring Logitech
- Favorite switch: Buckling spring
Thanks for the information Polecat. Attached below are pictures of the bottom label. I didn’t see a version number but I included the full bottom and a closeup up the label in case I’m missing it.
- Polecat
- Location: Downstream from Silicon Valley
- Main keyboard: Monterey K104 Industrial Gray
- Main mouse: Logitech Optical
- Favorite switch: Early Alps SKCM
- DT Pro Member: -
Thanks for the photos. The red SEFCO label suggests it's an earlier model, which agrees with it having what's probably an earlier EPROM version (smaller). Unfortunately there's no paper label or "inspection" number on yours. The Wiki shows examples of those:thefarside wrote: ↑01 Jul 2021, 21:18Thanks for the information Polecat. Attached below are pictures of the bottom label. I didn’t see a version number but I included the full bottom and a closeup up the label in case I’m missing it.
wiki/Northgate_OmniKey:_Gen1
The paper labels were probably added by Northgate after the keyboards were delivered. They are often missing (fragile...) and might not have been used on some earlier versions. The serial numbers on those labels do not necessarily correspond to features such as switch type (blue or white Alps). The inspection number was probably put there by the manufacturer. Two or three number formats were used on those (102xxxxx, 80xxxxxx, 10xxxxx), and they also do not appear to be a valid way to predict features. I have one very early 102, which is likely to have the smaller EPROM. I'll try to dig it out ASAP. In the mean time checking for power on the chips would be a useful test, as would be trying a different converter and/or booting up the one you have with the keyboard in AT mode. My Northgates are not picky about converters, but I don't think I've tried the very early one.
edited for typos...I hate laptops!