FK-9000 not working - any ideas?

[chyros]

Hey guys,

So I got this FK-9000 a while ago, and it's a very interesting board (hence why I'm looking for another or similar keyboard). Unfortunately, it doesn't work. At all. Just says "no keyboard present" when booting up, and the lock lights stay on. The built-in calculator only displays rubbish, but does actually respond to some keypresses. The battery is corroded over and I asked recently (here) whether the battery is mandatory for it to work but apparently that's not the issue.

The board is in a fairly bad state electronically speaking, I think. The PCB has blackened areas over it; I can't tell whether that's the electronics corroded (?) or something else, but it looks nasty (see below).



I've heard this type of keyboard is quite unreliable, but I haven't been able to find out exactly why. Does anyone have any idea if this keyboard is salvageable, and if so, where to start?

(interesting side note: it uses Alps clones, as well as a few (11, randomly distributed) actual, genuine Alps switches! In the picture below, of the numpad area, you can see FOUR different types of switch!)

[CPTBadAss]

Have you tried cleaning the PCB with say isopropyl alcohol? Just to see if it helps?

Also sometimes Alps switches have the logo on the bottom of the switches.

[suicidal_orange]

If it's not showing up to the computer that means there's a problem between the cable and the controller - any chance of a pic of that area?  Is the cable damaged?

If you can see two points connected by a PCB trace you should get a continuity beep using a multimeter, if you do it's not corroded enough to be bad, but it would be tedious to test every trace but you could try a couple.

If the controller has failed you could reincarnate it by wiring in a Teensy instead, you would loose the calculator though.  Unless you had the keys wired up to two Teensys, and a switch to choose which they were connected to...

Do you like the fact that it has a calculator or the switches and layout?

[chyros]

If it's not showing up to the computer that means there's a problem between the cable and the controller - any chance of a pic of that area?  Is the cable damaged?

If you can see two points connected by a PCB trace you should get a continuity beep using a multimeter, if you do it's not corroded enough to be bad, but it would be tedious to test every trace but you could try a couple.

If the controller has failed you could reincarnate it by wiring in a Teensy instead, you would loose the calculator though.  Unless you had the keys wired up to two Teensys, and a switch to choose which they were connected to...

Do you like the fact that it has a calculator or the switches and layout?

Sure, here's some pics, front and back:




This area actually appears to be relatively unspoilt. The areas around the big processor and the calculator connector similarly appear to be quite intact. Is it physically possible for the keyboard to not be detected by defects between the PCB traces? I'll try a multimeter next Monday when I'm back at work.

I kinda like the fact that it has both, but the calculator is the more interesting feature of the two IMO.

Have you tried cleaning the PCB with say isopropyl alcohol? Just to see if it helps?

Also sometimes Alps switches have the logo on the bottom of the switches.

I just tried to clean it a bit with some EtOH but it really doesn't do much. Gets off a bit of the white powdery stuff that's around here and there but otherwise the black bits are not cleaned with it at all.

The other switches are definitely Alps clones, I think type OA2. I really don't know why there's a mix of the two in the board, maybe 1994 (that's when it was made) was when they started transitioning.

[suicidal_orange]

That all looks very healthy.  I'm thinking that if the power lines are connected from the cable to the controller chip the LEDs could light up, but if the data lines aren't connected it won't register as a keyboard.  Doesn't look likely that's the case but it's a single sided PCB so you should be able to follow the four connections - only one connects in the pic.

[chyros]

Sorry mate, you totally lost me there xD . I don't know all THAT much about electronics xD .

[orihalcon]

If you are willing to ditch the calculator function (which sounds like may already be shot), then you can just completely replace the original controller.  Alps boards are nice in that you don't actually need a physical PCB to be there due to the metal mounting plate in most cases (in my experience anyway).

Xwhatsit controllers work well even for non-capsense boards.  See my post here if interested: https://geekhack.org/index.php?topic=69394

[chyros]

Hmmm, that looks interesting, but could it really handle all these strange, nonstandard functions? There's as many as 129 (!) keys on this board... Also, it would be really cool to get the calc going too - the PCB in that area looks pretty good and the separate calc PCB looks fairly good too. Do you think it's possible that it's just corrupted processors maybe?

[suicidal_orange]

Sorry mate, you totally lost me there xD . I don't know all THAT much about electronics xD .

I just went to highlight what I was talking about and noticed a mark I somehow missed last night - at the end of the 4 solder points near the middle (the cable connector?) there is a mark going fully across one trace.  Is that a break?  That could be the problem for the keyboard side if it is.

The controller recommended above is designed to replace failed ones so it must happen often,129 is not a lot of keys and you could program them to do whatever you want

Can you work out how the calculator works?  You mention a separate PCB so it might be completely separate from the main controller, in which case it might be repairable.  More pics please )

[chyros]

Yeah, the 4 solder joints are the cable connector. There is a glob of what appears to be resin on it, but it doesn't look like a break. Here's another picture of it:



full keyboard, front side (as well as my glorious crotch ):



full keyboard, back side:



the area of the keyboard that the calculator switches are in:



a chip:



another chip:



calculator connector:



calculator, front side:



calculator PCB, back side:



calculator PCB, front side:

[Daniel Beardsmore]

Also sometimes Alps switches have the logo on the bottom of the switches.

Alps SKCL/SKCM always have a logo on the bottom. The clue though is the tabs that rest on the plate:



Four small tabs indicates a clone: Alps never used this design.
Two long tabs generally means genuine Alps, but some clones did have the long tabs.

[Daniel Beardsmore]

The other switches are definitely Alps clones, I think type OA2. I really don't know why there's a mix of the two in the board, maybe 1994 (that's when it was made) was when they started transitioning.

Keyboards and switches by year — Focus isn't even listed here yet. It's an interesting question — I don't know whether Focus transitioned from genuine Alps to clones or, like with Ortek, ran both switch types in parallel. The only way to be sure is to record examples in that table and see what pattern emerges. I already have confirmation of OA2 in 1993 FK-2001 keyboards from Sandy, but I've yet to prove that T1 replaced OA2 in 1995. I don't yet have any confirmation of T1 existing before 1994 or so, though.

To confirm OA2 you do unfortunately need to open a switch. It's easiest done on a de-soldered switch as they're far easier to reassemble that way.

[chyros]

The other switches are definitely Alps clones, I think type OA2. I really don't know why there's a mix of the two in the board, maybe 1994 (that's when it was made) was when they started transitioning.

Keyboards and switches by year — Focus isn't even listed here yet. It's an interesting question — I don't know whether Focus transitioned from genuine Alps to clones or, like with Ortek, ran both switch types in parallel. The only way to be sure is to record examples in that table and see what pattern emerges. I already have confirmation of OA2 in 1993 FK-2001 keyboards from Sandy, but I've yet to prove that T1 replaced OA2 in 1995. I don't yet have any confirmation of T1 existing before 1994 or so, though.

To confirm OA2 you do unfortunately need to open a switch. It's easiest done on a de-soldered switch as they're far easier to reassemble that way.

Oh I already have, first thing I did xD . That's how I suspected it was OA2. It also confirmed that the board is a mix of clones and genuine Alps (the Alps ones have the complicated switchplate). They're a ***** to reassemble though.

Here's a picture of the insides of one of the clone switches:

[Daniel Beardsmore]

There's very little if any confirmed reliable dating information for these keyboards. I try to go by the controller chip if I can, or the case mouldings, but photos of anything bearing a date is very rare for Focus keyboards.

You're saying 1994 for yours, which corresponds with the 1993 dates on the chips. Are you getting the 1994 year from the case or the serial number? What serial number is on yours?

[chyros]

I dated it using one of those dial daters in the moulding of the keyboard - says Jan 1994. Serial number 940106227. The dial goes back all the way to 1989 but I doubt they were being made at the time.

[Daniel Beardsmore]

Hm, so your serial number and case date are indeed a match. FK-2001 goes back to 1988, but the FK-9000 only goes back to 1992 (based on FCC ID registration dates).

See, Sandy has an FK-2001 with white Alps and a 95 serial and 1992 PCB date, and without any photos of the chips, it's not clear which date is correct, if either:

http://sandy55.fc2web.com/keyboard/focus2001.html

If the serials do give the year of manufacture, and there's tentative evidence that they do, then there was no cutover from genuine Alps to clone. With that said, his "95" keyboard has bamboo switches, so maybe Focus were willing to buy cost-reduced Alps switches.

It would also indicate that Simplified Alps Type III is from 1996, which would be odd, since that post-dates the apparent EOL of OA2, but it's consistent with the use of Windows keys on a keyboard. It also contradicts the 1990 dates on the keyboard (DSI Modular Pro) where the bad moulds used in that switch originated. The implication would be that the bad moulds were stored somewhere for six years!

It's a weird world.

[chyros]

The Alps world can be a bit confusing from time to time, for sure xD . I guess it adds to the excitement though ^^ .

Most of all I just can't believe they're no longer being made.

[Daniel Beardsmore]

I don't know who's paying for the Focus website, as the site's still there, but the company seems to have gone out of business. Even Tai-Hao—their keycap supplier—lost contact with them. I still get the same error for their sales mailbox that I've been seeing since December 2012 or so:

00:12:22 > RCPT TO:<sales@focus.com.tw>
00:12:23 < 550 Mailbox quota exceeded

It's a shame that Alps switches are no longer sold. Even the Alps clone switches they used aren't sold any more — Hua-Jie confirmed to me that the AK-CN2/DN2 series went out of production, as did AK-C5/D5 (the reduced metal series), leaving only AK-CN2 (2)/DN2 (2) which are almost the same as Xiang Min switches (I have no idea who copied who there).

So many keyboard and switch product lines are all gone now.

My hope was that we could get a comprehensive list of keyboards and switches by year, from which we could determine patterns and understand who made what and when.

[chyros]

Tested it with a multimeter (one at least as old as the keyboard xD) and it would appear that even the most blackened PCB traces show good conductivity. All the chips seem to be conducting as well. Good or bad news? xD

[keyhopper]

Looking at your front full keyboard picture of the PCB, I see a switch looking thing, what could that be?
At first I thought it could be a toggle between XT and AT protocol, but the keyboard date is too new
Cheers!

[keyhopper]

There is a video in youtube from a guy with a FK-9000 that doesn't work either. At some point in the video he says he maybe will try to replace the electrolitic capacitors in it. (at least that is the fix with some old keyboards).

      "1992 Focus KeyPro FK-9000 keyboard" 
     

Cheers!

[chyros]

Looking at your front full keyboard picture of the PCB, I see a switch looking thing, what could that be?
At first I thought it could be a toggle between XT and AT protocol, but the keyboard date is too new
Cheers!

It is an XT/AT switch; yes, I tried it in the AT position. 1994 is a bit on the new side for an XT/AT switch, I agree, but it would appear they tried to cover all bases with this keyboard. Which is charming, really/

There is a video in youtube from a guy with a FK-9000 that doesn't work either. At some point in the video he says he maybe will try to replace the electrolitic capacitors in it. (at least that is the fix with some old keyboards).

      "1992 Focus KeyPro FK-9000 keyboard" 
     

Cheers!

Yes, this guy appears to have THE EXACT SAME ISSUES AS ME, STEP FOR STEP Oo . Almost unnerving. I don't know which ones the capacitors are or if items like these are easy to order though Oo .

[keyhopper]

[...]
Yes, this guy appears to have THE EXACT SAME ISSUES AS ME, STEP FOR STEP Oo . Almost unnerving. I don't know which ones the capacitors are or if items like these are easy to order though Oo .

IBM Model M2's are famous for having an issue with capacitors:

       "Repairing the IBM Model M2"
       https://geekhack.org/index.php?topic=5065.0

in that thread, the keyboard's LEDs lighted up but the keyboard did nothing. After replacing the caps it worked fine. But those were SMD caps prone to fail.

The electrolytic capacitors are the ones that look like little batteries. They are specified by capacitance and max voltage, for instance 2.2uF 63v
When finding a replacement, you can use one of higher max voltage, but the capacitance in uF must be the same one.

They have a polarity, so take pictures of where the (-) lead was soldered to before desoldering the old one.

Good luck!

[chyros]

There's only one of the little battery capacitors on the board, says 10 uF 25V. I'll see if the electrical workshop has one of them lying around for me . Cheers!