Some keys not working on Apple //c

Did these switches gradually die, over time, or did they all stop working at once?

You can order keyswitches. Which model do you have? There are three keyswitch types used on //c systems, depending on the model. The two latter switches are generally interchangeable (Alps Amber) with a rectangular aperture; the earliest model uses a + shaped male stem.

Note that repairing switches requires very deicate and patient desoldering and resoldering.

MacMark wrote:

It seems to be the + shaped male stem.

 

I did not use the //c for a while. After moving houses i have now enough room to use it again. At first it did not boot and i had to connect the cable for the Z80 card again to the motherboard like i did years before. The card sits below the keyboard and has just a black cable ending in a clip that snaps to the 2nd pin of the TMG chip like shown here:

https://www.macmark.de/apple2c/images/20.html

 The Z80 card is in the slot for the 65c02 and holds the 65c02 right next to the Z80. This is for CP/M, Turbo Pascal, dBase II and the likes.

During that reconnecting the Z80 i removed the keyboard too. But i don't think i damaged the keyboard's logic board on the back. 

 

Happy that the //c run again i tried my old programs and found those keys listed above do not work anymore. I remember having some not reliably working keys a few years ago already.

Open up the system again and check that the KB cable is firmly seated on both ends. 

Your other easy options are to send it to someone like me for a diagnostic, or to test it in another system. You can interchange that keyboard across all //c and //c+ models.

Z80 card now, that's unusual in a //c. 

if it used to work, but stopped working after been left in storage, it's probably corrusion in the keyswitch contacts.

apply some electrical contact cleaner and press the keys gratuitously before breaking out the desoldering tools.

dorkbert wrote:

if it used to work, but stopped working after been left in storage, it's probably corrusion in the keyswitch contacts.

apply some electrical contact cleaner and press the keys gratuitously before breaking out the desoldering tools.

That certainly could help, but because he has disassembled and reassembled it, I am cautious to give that advice. I am looking for a //c KB connector pinout to see what parts of the matrix are related. If all of the nonresponsive keys share a comon signal, then that is what I would suspect to be the fault. 

If they do not share a signal, then I would suspect the switches. 

This is what I have found:

Keyboard matrix

X0    1    2    3    4    5    6    7    8    9Y0    ESC    1    2    3    4    6    5    7    8    9

1    TAB    Q    W    E    R    Y    T    U    I    O

2    A    D    S    H    F    G    J    K    ;    L

3    Z    X    C    V    B    N    M    <    >    ?

4                            \    +    0    -

5                            ~    P    [    ]

6                            RET    UP    SPC    "

7                            DEL    DN    LT    RT

So, from what the OP said, if keys 7, U, K, P, +, Up, Down, and < aren't working., then signal X7 (Pin 9 on the 34-pin connector) is not connected, on one side or the other. I have only one immediate reference on the connector, so this needs verification. Comma is the same key as   '<'   , so that certainly sounds like X7 isn't wired.

This could be from the KB switches to the matrix, or from the IC to the cable header on the keyboard; or it could be a cold/broken joint on the mainboard for the other end of that connector; or a damaged trace from the mainboard connector to elsewhere on the mainboard. Spraying electrical cleaner into the switches won't solve those issues.

MacMark wrote:

I disconnected the keyboard (german layout, see pics in link below) from the motherboard and reconnected it. The keys are still not working. I verified again these are the affected keys 

• 7
• ‘ (this one is on the left side of the "delete" key in the top row:  8 9 0 ß ‘ delete)
• u
• p
• k
• , (this one is to the right of n and m and above the black apple key)

 

The keyboard cable can only be detached from the //c motherboard. It seems it cannot be detached from the keyboard.

 

I added new pictures of the keyboard, logic board and cables/connectors here:

https://macmark.de/apple2c/keyboard/catalog.html

The matrix is bound to be slightly different, due to the layout, but 7, U, P, K are likely the same signal on all layouts. If you enter basic, press the Down Arrow, and see if the cursor descends one line; then, press Exc, then press the Up Arrow to see if you can move it up one line. 

Your grave accent key takes the same position as the + key would normally, so it is also likely on the X7 signal. I would bet money that the up and down arrows do not work.

If that is the case, hen perhaps it is as simple as a quick reflow on the KB ribbon cable solder pads, or the mainboard KB cable connector pads. I do not see immediate trace damage in the photos, so you could have weakened/broken an already-cold joint when you disconnected it. 

Look also to see if any of the KB ribbon wires have any damage along both sides of the ribbon. If you can put a continuity tested on both ends of the ribbon, and check each connection from the ribbon termination, to the pad on the KB logic board, that will help to determine if there are any breaks.

P.S. I find it interesting that you have a Section symbol instead of £ on the 3-key. I suppose that makes sense. I show signal X7 as 'Pin 9' on the pinout that I found, but I do not know which side of the female connector starts at 1 on that pinout. If I can find more--I may need to see if there is a schematic in any of my papers--then I will inform you. 

Edit: Ah-ha, here is the connector pinout:

iic_keyboard_connector_pinout.png

MacMark wrote:

Timelord wrote:

 

The matrix is bound to be slightly different, due to the layout, but 7, U, P, K are likely the same signal on all layouts. If you enter basic, press the Down Arrow, and see if the cursor descends one line; then, press Exc, then press the Up Arrow to see if you can move it up one line. 

 

Your grave accent key takes the same position as the + key would normally, so it is also likely on the X7 signal. I would bet money that the up and down arrows do not work.

 

You're right. I tested it with Basic Mode and also with the menu list of "Introduction to Apple Logo". It says i could select a menu entry by typing a number or use the up and down arrows. Up and down arrows did not work. Left/right arrows do work.

Sadly, I am often right, but that isn't good news for you...

This verifies that the signal X7 is interrupted somewhere on its path. FWIW, you can check arrow keys from a system BASIC prompt, but thast is a matter for another time. 

(To do this, you can press Down Arrow at any time, and use the Esc key to enter editing mode, and press the Up Arrow.)

...so, NEXT:

You will need to examine every point, from the KB solder pad, through to the system's use of signal X7. This, believe me, is not fun. 

Do you have access to another //c keyboard, of any sort? I usually would test if it is the system, or the KB logic board with a part from my spares. Without that, it is a much larger job, as this could be in one of several places:

On the KB mech: 

 1. Between the cabe and the matrix controller.

 2. The KB cable itself.

 3. TThe solder pads for the Kb cable.

On the mainboard:

 1. The KB connector.

 2. Traces from the KB connector to elsewhere. 

Possibly some other places, but those are the common ones. 

If you have another //c keyboard--any version will work--then you can attach it and test to see if the issue exists. if it does, that narrows it to the KB logic board; if not, then it narrows it to the mainboard. I'm having a similar issue on a //gs board where signal Y1 isn't connected properly, and while I have  done a lot to narrow it down, adding a bodge wire along its trace soehow did not suffice. (I'm unsure why, as I bodged it to the via that the trace normally uses.) 

I would loan you a KB logic board for this test, but the freight back and forth to Germany would cost as much as just buying a spare, which would be absurd for you.

The inexpensive steps:

Reflow all of the solder joints on the KB mech, where the cable attaches to the mainboard. 

Check continuity between PIN 9 on the cable and the nearest trace point AFTER its connection. (Follow the traces and see if the female pin on the cable has continuity to the other side of the trace.)

If all is well on the KB mech, repeat these for the mainboard.

I do not have a KB mechanism with the appropriate language set-up for that machine.  The keyscps from a normal pin-switch KB will not mate, 1:1 between regions. Even PAL mainboards are getting hard to find. I have some spare NTSC boards, but I doubt that you would want to deal with that hassle. 

MacMark wrote:

Usually the //c keyboard has a strengthening bar below its logic board. This bar was removed from my //c to make room for the Z80 card. I guess that typing on the keyboard without that bar may have bended and damaged the keyboard's logic board. Good thought?

I suppose it is possible, but unlikely. If anything, extra heat from the Z80 card, or possibly even shorting between the two; or just the stress of removal and insertion could have damaged it. Do you have a hot air station, or a basic solering pencil?

In regard to finding a direct replacement, I would not know outside of ebay.de or your local classifieds websites. I would think that a Danish KB would be equivalent, but several EU region KBs may work. The UK version is similar, but keep in mind that you have hairpin keyswitches, not Alps, so if you but a KB for another region, and care about having the correct keycap symbols, that if you buy one for another region andit uses Alps caps, then you cannot swap them across. 

The Alps mechanism may also not fit with your Z80 card as it has a metal frame. 

The first thing to deduce is if the issue is the KB or the system mainboard. Any second //c mainboard or any second //c keyboard will allow you to test this. I do not know in what city you live, but there may still be fragments of an old user group nearby, to investigate. 

MacMark wrote:

I guess i can have access to a 2nd //c in my region and will try what happens when i exchange the keyboards.

I would also consider using an US keyboard because that's what i use on all my recent Apple devices because the US layout fits better my programing needs.

 

My 4yo recently "played" with the //c keyboard. Maybe a bit too rough. Just to complete the possible reasons list :-)

Brilliant. If you can plug your KB into another //c, you can check if the keys work on that, or not.

If they do not, then it is the KB mech.

If they do, then you know that is a mainboard fault. 

Please keep me updated. I am happy to continue to help, if possible. 

I personally hate, but I live with, US layouts. (I detest the locations of `@` and `"` on them.)

The fact that the keys are in the same row of the key matrix means that the fault was never with the key switches, but rather with the wiring. Timelord outlined the possibilities for where that could occur. If the fault is (currently) gone, that indicates strongly that a cold solder joint is to blame. Cold joints or broken joints can move inside the pad and break and make contact intermittently. The other possibilities are a broken wire in the flat flex cable making intermittent contact, or indeed a broken trace on the keyboard PCB with the same effect. You mentioned that the stiffener bar was removed and the PCB was able to flex, maybe more than desired when the toddler banged at it. I would suggest inspecting that PCB for any evidence of cracks.

Cold solder joints or cracked solder joints or hairline-cracked PCB traces will cause issues with temperature changes.  Often times when the PCB heats up the cracks sever.  Other times it's the other way around.

Not at all:

Applied Engineering sold many Z-RAM and Z-80 boards for the Apple //c.

The Z-RAM increased the available RAM for AppleWorks as well as provided a Z80 CPU for CP/M.

One of the two Apple //c's that I purchased a few years ago on eBay for parts, had an AE Z-80 card in it.

My original //c has a functioning Z-RAM card in it.

It was Microsoft that revolutionized the original Z-80 Softcard for the Apple ][e so that they could sell more programing languages, and in its day the Apple ][ because the largest number of computers that could run CP/M.

AE just carried that tradition forward with their cards for the ][e and //c.