Atari 400 Keyboard - Shift, Ctrl, Break not working.

[slacker]

Hi All,

 

I've been working on fixing up an Atari 400 my friend gave me. It started with just getting a BSOD at boot. After attempting to clean and re-seat all the chips on the main board, CPU board, and RAM board with no luck, I picked up a 48k RAM upgrade board. After installing that, the computer booted! All keys on the keyboard worked and was able to use my BASIC cart for a while without any issues.

 

I then I ran into another problem then where the joystick ports didn't work. (Just to note, these didn't work from the start, they didn't stop working after the computer booted for the first time) I was able to get those working by installing a new PIA chip (G65SC21P-2). Things seemed to be fine at this point. I played some games but then had issues with keyboard suddenly missing keys. Surprisingly, pulling and re-seating the PIA chip fixed this even though I was under the impression the keyboard was controlled by the POKEY and the two multiplexer chips.

 

Anyway, long story short, I'm at the point where everything works great except the SHIFT, CTRL, and BREAK keys. These used to work as I was able to program a bit in BASIC earlier but now they're suddenly not working.

 

I should also note, I have a 5200 that I swapped the CPU, GTIA, ANTIC, and POKEY chip back and forth into the 400 and it didn't make any difference to anything. I also swapped around the two keyboard multiplexer chips without any change.

 

I'm wondering if anyone has any ideas or if you guys know what pins I can jump on the keyboard connector to simulate pressing "SHIFT+2" to get a double quote or something. I've been able to jump some pins to get it to display a bunch of characters but I'm not sure what combo is needed to simulate pressing SHIFT. The pinout images I've found online don't seem to match up what I'm seeing on my computer. Basically, I'm trying to rule out if somehow I borked the keyboard taking things apart 1 million times or if some other trace, chip pin, or something is broke or making a bad connection on the main board.

 

Thanks in advance!

 

Edited August 21, 2022 by slacker

[xrbrevin]

try this for reference:

 

[TGB1718]

I believe SHIFT/CNTL/BREAK are the only keys connected to pin 9 on the ribbon cable

I would check there first for a bad connection

[slacker]

Thanks for the replies. I guess I'm confused on how to tell what pin number is what. In the attached picture, you can see my keyboard plugged in on the bottom right.

 

If counting from the top down, if I jump pin #2 and I think maybe pin #15 or so (I don't have the computer in front of me right now) I was able to get a 2 to print on the screen. If i moved up to pin #14, it would print 1 and so on. The 2 seem to line up with the map above, but doesn't mention what other pin needs to be connected to get it to display a 2. Moving the one side to pin #1 or #3 would generate a different set of characters.

 

Would pin #9 be the 9th from the top? and if so, what pin would the other side go to?

 

Thanks again.

 

[TGB1718]

This should help:-

Pin 1 at the top of your photo.

Also attached the service manual in case you don't have it.

 

Atari_400-800_Service_Manual.pdf

[slacker]

Thanks! Downloaded!

It looks like if only some keys don't work the Atari guide basically specifies "replace keyboard". I haven't had a chance to dig more into the manual to see if there's other options, I'll do that later tonight.

 

I did make a little headway. I noticed that when some of my keys don't work on the keyboard, if I press a bit down on the POKEY chip, they work again until I lift my finger. I'm guessing it's not making a good connection in the socket. The socket when I got the computer was in bad shape with a lot of corrosion. I'll try an soldering in a new socket and see if that at least fixes that issue.

Pressing down on the POKEY doesn't seem to change anything with the SHIFT, CTRL, BREAK buttons though.

 

I tried jumping pin 8 & 9 to each other and to various other pins while I had a jumper stuck in the other pins printing a constant stream of "2"s and wasn't able to get shift to work that way. I'm thinking maybe that's a keyboard issue? It's hard to tell because I'm not sure I'm jumping the pins for shift correctly.

 

[Rybags]

I noticed the pins on the 400 kb connector get plenty of oxidation.  You could remove the cable and replace a few times to help clean it.  Though your cable socket looks different, I'm fairly sure my one has actual pins, not an insert connector like yours.

Also with Pokey and the kb multiplexor ICs, pull them up about 1 mm then reseat.

Though the multiplexor ICs are probably OK, there's usually more widespread issues if one goes bad.

[slacker]

The keyboard connector on mine has a socket on the main board side but on the keyboard side, there's a bunch of pins at the end of the flex cable.

Unfortunately no luck with pulling ICs or swapping the multiplexers. I even hit the pins with some very fine sandpaper and used basically 2 cans of Deoxit spraying out the sockets and the pins.

 

Attached is an image of the start state of the socket the POKEY was in (I don't seem to have an after picture handy). I cleaned it up as best as I could. It no longer looks like it's bad but the fact I can press on the chip in the socket and the the keys (minus SHIFT, etc) start working leads me to believe that socket is done for.

 

I was going to just pull one of the sockets from my 5200 and solder it in but with how cheap a pack of them are, I decided to just get some new ones. That should be here sometime tomorrow so I'll see if putting in a new POKEY socket solves anything.

 

Edited August 23, 2022 by slacker

[TGB1718]

With corrosion that bad, it may have even gone through the pins below the socket where you can't see,

replacing is a must when it's that bad, good luck removing the old one, at least with the 400 the PCB

is much sturdier than XL/XE models

[bfollowell]

1 hour ago, TGB1718 said:

With corrosion that bad, it may have even gone through the pins below the socket where you can't see,

replacing is a must when it's that bad, good luck removing the old one, at least with the 400 the PCB

is much sturdier than XL/XE models

 

I agree with @TGB1718. That corrosion looks horrible. I'd be willing to bet, especially since you said you said pressing down on POKEY got the keys to work, that if you clean that corrosion or replace that socket, that will probably take care of your issues.

 

[slacker]

New POKEY socket installed. Seems like key presses are consistent now.

 

Still no shift, ctrl, break though.

 

 

[+mytek]

9 hours ago, slacker said:

New POKEY socket installed. Seems like key presses are consistent now.

 

Still no shift, ctrl, break though.

 

Since you could get those keys to work previously by pushing down on the Pokey chip, that suggests that you might have a broken trace close to where the chip socket resides.

 

Try running a jumper from Pokey pin-16 which is the Key Response input for SHIFT, CTRL, BREAK over to R124 to see if that restores operation (see images below).

 

Edit: put the jumper on the underside of the PCB.

Edit2: Also heat up and re-flow the solder on R159 next to Pokey pin-16 (that's a pull-up that establishes the key-up state). Perhaps just bridging with solder between that and Pokey pin-16 will make things right without doing the jumper, because I'm suspecting that the corrosion ate away the trace around pin-16.

 

 

[slacker]

Thanks!

Solder reflowed on R124 and R159.

When I had a jumper installed from R124 to POKEY pin 16 from the bottom, the keyboard actually stopped responding. I'm not sure if I used too thick of a wire or it was a bad solder job (probably). I used a thinner jumper and the keyboard worked but CTRL, SHIFT, etc still didn't work.

 

Removing the jumper did as expected, all keys worked but those few.

 

Sorry I wasn't more clear earlier. My shift keys etc worked originally but stopped working randomly. I also had issues where other keys on the keyboard would drop in and out working. Pushing down on the POKEY fixed those keys but not the shift, etc keys.

 

Is there a spot on that KB pin 9 route that I can jump to simulate the shift key being always on?

[+mytek]

On 8/25/2022 at 11:41 AM, slacker said:

Is there a spot on that KB pin 9 route that I can jump to simulate the shift key being always on?

Surprised no one answered this yet. Well if I'm interpreting the keyboard matrix properly from this webpage, it looks like pin-8 to pin-9 should be the equivalent to pressing the SHIFT key.

 

And pin-5 to pin-9 looks to be the CTRL key, with pin-1 to pin-9 being BREAK.

 

[slacker]

Thanks!!! Yes, that seemed to work!

Setting those pins and then jumping the character pins displayed their shift counterparts.

 

I guess that means my keyboard is messed up possibly on PIN 9 as that seems to be the common pin between all three keys.

 

Is there anyway to repair these keyboards or is it one of those things of "buy a new one"? They seem like they're getting hard to find these days.

 

[bfollowell]

I ordered one from Brad at Best Electronics several months ago. It seems like I only paid about $25 for it. His stock must be getting low. He still has them listed, but they're now $45.95 each, plus shipping. These are NOS and mine was beautifully pristine if that helps any.

 

Atari 400 US Mono Panel Keyboard CA017999 down on page 83.

 

[ClausB]

3 hours ago, slacker said:

Setting those pins and then jumping the character pins displayed their shift counterparts.

The problem could be in that keyboard socket. Your jumper pins are likely thicker than the keyboard pins. If the socket has lost its springiness, it might make poor contact.

 

Or it could be the flexible cable. Put an ohmmeter between pins 8 and 9 and press Shift. Do this both at the bottom and at the top of the flex cable.

 

Some keyboards have circuit traces on the underside. Follow those and look for cracks or scratches.

[slacker]

On 8/27/2022 at 4:10 AM, bfollowell said:

I ordered one from Brad at Best Electronics several months ago. It seems like I only paid about $25 for it. His stock must be getting low. He still has them listed, but they're now $45.95 each, plus shipping. These are NOS and mine was beautifully pristine if that helps any.

 

Atari 400 US Mono Panel Keyboard CA017999 down on page 83.

 

Thanks. Looks like Best might be the only option right now if I go that route. I set up an email alert on eBay but seems pretty empty there for just a KB. Even completed listings only show one sold back in June.

On 8/27/2022 at 5:42 AM, ClausB said:

The problem could be in that keyboard socket. Your jumper pins are likely thicker than the keyboard pins. If the socket has lost its springiness, it might make poor contact.

 

Or it could be the flexible cable. Put an ohmmeter between pins 8 and 9 and press Shift. Do this both at the bottom and at the top of the flex cable.

 

Some keyboards have circuit traces on the underside. Follow those and look for cracks or scratches.

Thanks, I tried this out. I put my ohm meter on pins 8 & 9. Without anything pressed it showed open (expected). Then, pressing the shift key I noticed it was actually reading a value! I moved the keyboard around and it still showed a value when shift was pressed. I want to say the value was ~60k ohm when pressed but not sure.

 

I plugged the keyboard back in and started the computer. The shift key was working again without any issues.

 

I'm tempted to put it back together with it working now but I'd hate to have to take it apart again for the 1000th time if it randomly stops working again.

 

I guess best course of action for now is to keep it apart, use it for a bit and if the keyboard stops working again:

• check the resistance on keyboard PIN 8 & 9 again.
• If it's not showing a value, keyboard ribbon most likely has a bad trace
• If it's showing a resistance value when shift is pressed, try jumping the keyboard port on the main board.
• If that works, keyboard port might be too worn for KB pins to make a good connection,
• if still not working, possibly bad trace somewhere on the main board that's making an iffy connection that's breaking if there's any flexing (ex: cartridge insertion or something).

I'll keep this thread posted on how it goes after some more testing.

[bfollowell]

3 hours ago, slacker said:

Thanks. Looks like Best might be the only option right now if I go that route. I set up an email alert on eBay but seems pretty empty there for just a KB. Even completed listings only show one sold back in June.

Thanks, I tried this out. I put my ohm meter on pins 8 & 9. Without anything pressed it showed open (expected). Then, pressing the shift key I noticed it was actually reading a value! I moved the keyboard around and it still showed a value when shift was pressed. I want to say the value was ~60k ohm when pressed but not sure.

 

 

I've seen several over the past 4-6 months, a few of them described as NOS, but nothing at the moment. They still pop up every so often though. You could always look into  the B-Key or one of the other after-market keyboards for the 400 too, but if you're like me, the old membrane, retro futuristic style keyboard was a big part of the appeal.

 

Good luck in your search!

 

[skaboytd]

Good ay all,

 

So I have the same problem on my Atari 400 the Shift, Cntrl, and Break appear not to work. This thread has been a wealth of information. I have checked all the connections, reflowed the solder and checked continuity, all seems to be okay. When I look at the signal on the Pokey pin 16 (KR2) it changes state from 5V to around 4.5V when the shift key is pressed. Is this enough of a drop to trigger the Pokey chip?  

[+mytek]

2 hours ago, skaboytd said:

Good ay all,

 

So I have the same problem on my Atari 400 the Shift, Cntrl, and Break appear not to work. This thread has been a wealth of information. I have checked all the connections, reflowed the solder and checked continuity, all seems to be okay. When I look at the signal on the Pokey pin 16 (KR2) it changes state from 5V to around 4.5V when the shift key is pressed. Is this enough of a drop to trigger the Pokey chip?  

No that isn't going to work. I don't recall the NMOS chip requirement for acknowledging a logic 'LOW', but that's certainly not going to be enough.