• Please review our updated Terms and Rules here

Sol20 keyboard issue

falter

Veteran Member
Joined
Jan 22, 2011
Messages
6,556
Location
Vancouver, BC
I have a weird issue with my Sol.. when I first fire it up.. the keyboard doesn't work. The lighted keys like CAPS and Local react when pressed.. but none of the letter keys do. Initially it seemed the issue was the strobe not working properly. I did not get a pulse when a key was pressed. I discovered accidentally shorting pins 3 and 4 on u10 would produce twenty or so of the letter you'd previously pressed. I swapped U10 and it was working fine, until I shut it off for the night. Again when I powered up.. no keys. However I left it running while doing some reading and was surprised when I pressed keys that it was working.

I feel like it might be heat related.. like you power it on and after warming up some flaky component starts working. Just not sure what it could be..
 
The SOL-20 uses a capacitive keyboard and will almost certainly need refoamed. If you haven't done that, I wouldn't bother troubleshooting anything else until you refoam it.
 
I refoamed using new pads made by texelec. I'm a bit suspicious of their design... they don't have foil on the contact side and the foam is denser than what I got in my last set. After installing them about 1/4 of keys don't work. But none of them do on start up.. they only start working after the machine has been running for about 2-4 minutes. I've confirmed the behaviour is the same even if I have the keyboard keys off and am just making contact with my finger.
 
I can't comment on the replacement pads from texelec, I don't think I've bought anything ever from them. Perhaps you have a cold solder joint that's making contact once the board heats up and expands/moves. Try pre-warming it with a hair dryer or something to confirm it's definitely heat related.
 
Guys I just got in the replacement Pads from texelec (see the link Al posted above as texelec commented in the thread too) she is apparently using what seems to be mylar this time, every single pad she sent this time is working so it seems she got the bugs out. Also, reach out to her directly. I did and she was very helpful.
 
After you sort out the "pad" issue, I would check a few things.

Check that your Tantalums on the keyboard are good or simply replace them. Sometimes they can act up differently as they warm up.

Also carefully clean the contacts on the chips using an art eraser. Then make sure your sockets are tight. I have the crappy Apple-1 style TI sockets on one of my sol-20 keyboards. Had to rebuild one because it had intermittent contacts (I don't generally replace them since I have gotten good at this specific socket type)

Finally check your cable and the connector on both the motherboard and keyboard PCB. If you have a cold solder joint, this would also do it.
 
Thanks guys. I did reach out to texelec and she said she will replace all my pads with the new ones. I kind of wondered why she didn't use mylar in the first place. I also sort of question the type of foam she used.. it's that dense white stuff. When I last bought pads from another seller they were a lighter density green foam with mylar and they work awesome. Mistakenly thought she was the same seller.

I definitely think they are part of the issue.. but initially no.. I think possibly what Corey suggested. It feels like everything needs to warm up a while before keys will start working. The lighted keys always work right off the bat.. just everything else doesn't. Once it warms up they all work if you touch them with fingertips.
 
Falter, not sure about the SOL layout but the foam she uses works and feels fine for my compaq portables. And I dont know why she chose either material. All I know is of my first batch, 44 pads did not work. And its strange they either did or did not work. There were none that SEMI-worked. I assumed they were glued on upside down but who knows. I am happy with her current product and will probably buy again. I tried making a whole set myself, and it was many many hours of work and didnt come out well at all. So I think its worth it.
 
When I last bought pads from another seller they were a lighter density green foam with mylar and they work awesome.

Those were closer to the material KT originally used. You really want to use low-density polyethane foam instead of the polyurethane so they won't crumble over time, but
it is difficult to find. The main difference will be key travel and force required to actuate a key. I've made hundreds of them, an example is on the left here:
pad.JPG
Higher density, but the same thickness as the originals.

see:
https://groups.google.com/forum/#!topic/lisalist/ki86xMT7gww
"Making perfect Keytronic replacement keypads"
 
Last edited:
I've done a ton of checking around and still not sure what's going on.

I've removed and cleaned every IC on the thing, with no change in behavior. I've tried swapping ICs with known good, and testing the existing in other machines. No problem. I've checked every IC pin connection to the PCB (from top of chip leg to where the socket leg is soldered in). No problem.

One strange aspect is no matter how long you run the Sol for after the keyboard starts working (usually about 5-10 minutes), if you shut it off even for 10 seconds and then power back on.. you're back to square one. I would have thought if it was a cold solder problem or a component that only makes proper contact after being heated up, that it'd still work after a quick power off and restart, since the heat would not dissipate that quickly.

I note that certain functions like keyboard reset work no problem right off the bat.

I guess I'll check out the caps now (though I'm not great at that) and replace em.. see if that helps.
 
Do you have a scope or logic analyzer you can use to check if the keyboard provides data immediately from power up? With even just a logic probe you can verify the keyboard's behavior as far as its data strobing and special signals at J3. Now, what the data it strobes actually is might require more than just a logic probe or scope - maybe a separate test jig that can power the keyboard and latch it's data for checking?
 
I'm still mastering my scope but as far as I can tell, yes, data is being generated on power up. Take that with a grain of salt though. I've also checked for pulses as keys are being pressed, things seem to line up with the Theory of Operation, anyway.
 
If you're getting a good data-ready indication from the keyboard at power-up and nothing's being displayed at the monitor prompt I'd think the problem is on the mainboard. Obligatory question: are the power rails good from a power-up (and do they stay good)?
 
First confirm that this problem does not happen with a different monitor, it might not even be your SOL that is the problem. Next I agree test the caps related to your keyboard circuit using an ESR meter. I would not shot gun replace them though, you can disconnect from their circuit by removing one of the posts' connection and then test with an ESR meter. This is a point of discussion that comes up from time to time, replace ALL the caps or just find the ones that are known-bad. I guess it's up to you and your solder skills.

This one will work:
https://www.ebay.com/itm/Anatek-Blue-ESR-Capacitor-Tester-Meter-Complete-Kit-for-Assy/263722628688

There are cheaper ESR meters, but you don't want to go too cheap.

b
 
I may have found my problem while trying to solve another. I have the Sol connected to a power bar and use the switch on that to shut it down or turn it on as the switch on my Sol is getting unreliable. Anyway, I accidentally sort of half pressed it causing it to *very* rapdily cycle power which knocked out the Sol entirely and left me with a screen full of junk.

Anyway, I know I've got some issues with sockets in my machine so I reseated a bunch of chips and while doing so found the plastic part of the socket at U28 (74163) lifted right off the pins! Anyway, I did my best to reseat that and as I did, I noticed on power up I had both the normal prompt again (whew) AND the keyboard was working immediately! I shut down again, reinstalled the spare memory board... powered up.. nothing from the keyboard. Pressed on the 163 carefully but firmly.. voila.. typing starts again!

Not sure if it's the 163 or some kind of micro-fracture in the PCB. I don't know for sure but I don't recall U28 being involved with the keyboard. Anyway, pressing there seems to get the keyboard working again and it'll stay working for a long time.
 
U28 isn't in the keyboard path but it is in the display path, so a problem there might make it appear to be a keyboard issue. See drawing X-18: it's the character clock divider.

It give me the creeps every time the board flexes! I dread the day a trace fractures. At least we have nice, pliable, leaded solder elsewhere! ;)
 
Yeah. I kind of think it might be a trace or something in that same area that does have something to do with the keyboard and it's getting flexed in just the right way as I'm pressing on U28. I should investigate further but I really don't want to take the whole thing apart to get the motherboard out.
 
Maybe use the SOLOS command "SET O=1" to change the default output from the video section to the serial port and use a connected PC to display the output. That should show if there's actually a keyboard-input problem or (just) a video-section problem.
 
Back
Top