PDA

View Full Version : IBM XT keyboard error 301



Pet Rescue
September 24th, 2013, 05:24 AM
Hi all can anyone help me get this old girl back up and running. I've had lots of problems, firstly the power supply packed in which turned out to be
diodes gone short. Then the Rodime hard drive went bang which was a found to be bad tantalum that blew it's top off and melted the ribbon cable slightly.
This I replaced and tested on another board and amazingly still works. The full height tandon floppy doesn't work but I'll get round to fixing that later.
What is really driving me mad is the 301 error, I have replaced the plug on the keyboard and even tried a clone keyboard with at/xt switch.
I have even destroyed a perfectly good 8255a Pia chip which was a nightmare to remove and socket hoping it was this that was causing the problem.

So new PIA and still no go, I am running out of ideas if anyone can offer somewhere else to look at!:confused:

Chuck(G)
September 24th, 2013, 08:49 AM
Did you replace the 74LS322 shift register? That's really the core of the keyboard controller.

Pet Rescue
September 24th, 2013, 09:15 AM
Did you replace the 74LS322 shift register? That's really the core of the keyboard controller.

Thanks for the reply Chuck(G) I haven't replaced the 74ls322 but I certainly will, I should have posted here first before destroying the PIA!

What could have caused it to go faulty? Unplugging the keyboard when switched on or the power supply blowing?

Stone
September 24th, 2013, 09:30 AM
Have you tried a known good XT keyboard that is *not* a switchable AT/XT model?

Chuck(G)
September 24th, 2013, 10:01 AM
You could probably damage the '322 with a simple ESD accident. This is a real XT, right? Let me stare at the schematic a bit before you haul out the soldering iron.

Okay, the 322 (U27) is a likely suspect; less so, the 175 (U21). Amazing that there's no buffering between the connector and the '322. You'd think that IBM would have inserted a schmitt trigger at least on the receiving end.

Do check out Stone's suggestion that you find a real XT keyboard first, however. Absent that, try the output of the PS/2-XT converter kit.

Pet Rescue
September 24th, 2013, 10:17 AM
Have you tried a known good XT keyboard that is *not* a switchable AT/XT model?

Stone the switchable XT/AT works on my 286 clone ok in AT mode but I cant say it works in XT mode.
The keyboard that came with it is an 83 key type without LED's I cannot check as I don't have another XT I'm afraid

When using the clone keyboard the LED's light up, shift lock stays on but caps lock comes on and off when pressed and num lock does the same!
(when in XT mode on the XT).


You could probably damage the '322 with a simple ESD accident. This is a real XT, right? Let me stare at the schematic a bit before you haul out the soldering iron.

Yes Chuck(G) it's a real XT with a 64k-256k board. I've taken all of the ISA cards out to test except for the VGA.

Pet Rescue
September 24th, 2013, 10:44 AM
Chuck(G) on the schematics I see a 74LS322(U27), 74LS175(U21) and 7407(U68.) is that right?

Chuck(G)
September 24th, 2013, 11:22 AM
Yup...

Pet Rescue
September 24th, 2013, 11:48 AM
What is the purpose of the flip flop, is it to do with the clock?

When I put a scope on the clock pin of the keyboard socket I could not get a clock pulse.

I am very green on the electronics of computers involving shift registers and flip flops etc.

Could you give me a layman's description of what it's doing or should be doing.

Chuck(G)
September 24th, 2013, 01:07 PM
I will, but I'm headed out for the rest of the day. Was the keyboard plugged in when you used your scope? The keyboard generates the clock pulses.

Pet Rescue
September 24th, 2013, 01:22 PM
I will, but I'm headed out for the rest of the day. Was the keyboard plugged in when you used your scope? The keyboard generates the clock pulses.

I think I did with the keyboard unplugged and then with it plugged in with the cover off the keyboard.

With the clone keyboard plugged in the appropriate LED's light (num lock and caps lock) except the scroll lock stays on no matter if it's pressed or not.

Thanks for your help by the way!

Chuck(G)
September 25th, 2013, 09:38 AM
If I recall correctly, the reset pin on the keyboard connector isn't used by many XT keyboards--it's there, but most keyboard firmware tends to depend on the XT to pull the clock low for more than 10 msec. They keyboard then resets and sends an 0FFH response out when the XT releases the clock signal. So, at the connector, you should see this at power up--low for a time, then high as the XT waits for the response. At this point, both clock and data lines should be high.

If either one is low, you now know where to look.

If the keyboard doesn't send the 0ffh sequence at reset, you get a 301.

Pet Rescue
September 25th, 2013, 11:38 AM
If I recall correctly, the reset pin on the keyboard connector isn't used by many XT keyboards--it's there, but most keyboard firmware tends to depend on the XT to pull the clock low for more than 10 msec. They keyboard then resets and sends an 0FFH response out when the XT releases the clock signal. So, at the connector, you should see this at power up--low for a time, then high as the XT waits for the response. At this point, both clock and data lines should be high.

If either one is low, you now know where to look.

If the keyboard doesn't send the 0ffh sequence at reset, you get a 301.

Now I'm sure when I replaced the plug the original had a wire going from pin 2(data) to pin 1(clock) but that didn't make sense so I missed out the link.
Would some XT's need the reset wire, so a clone keyboard would not work in that model?

modem7
September 27th, 2013, 06:35 PM
Diagram [here (http://www.minuszerodegrees.net/5160/keyboard/5160_keyboard_startup.jpg)].

Chuck(G)
September 27th, 2013, 08:08 PM
Sorry and thanks for the collection. Successful self-test response is 0AA as noted, not 0FF.