Problem with IBM Model F AT - all LED lights light up, no key output

[kmnov2017]

Just got a pristine condition IBM Model F AT Italian layout (super rare). I plugged in an active USB converter, all LED lights light up but I don't get any output.

I tried with a Promicro with Soarers's firmware - still no luck.

I measured the Voltage output between VCC and GND and also between RAW and GND on the pro micro, I am recieving 3.4V.

I remeber reading in the past that the F AT needs at least 5V to work, is that correct?

Both my XTs and all my M122 work fine with the same Promicro.

[Muirium]

What’s the voltage coming out the keyboard side of the active converter? Still under 5v?

[kmnov2017]

Active converter - 5V directly on the active converter
4.36 V on the PCB leads when the keyboard is connected.

All three keyboard LEDs are continuously lit...

[Muirium]

I’m still thinking power draw. The AT is no ordinary AT/PS2 keyboard. It’s a beast.

If you’ve got a machine with a PS/2 port handy, you could always try a simple DIN to PS/2 adapter. Those are drop dead simple pass-through. The AT is kinda sorta still supported by relatively modern hardware, after all. Even just for testing’s sake.

[jim121049]

Make sure that J1 on the Pro Micro (next to the USB connector) is bridged.

[kmnov2017]

I did bridge J1, still the same result - No output...

[Rauha]

I have external Model F AT converter with female DIN5 connector. It had exactly same symptons (all lights on, no output) when it had wires accidentally touching each other. This was caused by that female connector's plastic cover being toot tight and squeezing the wires together inside it. I solved it by just adding non-conductive tape on those connections. Took me a while to locate the source of the problem there. Thought I'll mention it, just in case you have similar type of connector.

[kmnov2017]

Still struggling to get this to work.

This is what I get on HID listen by disconnecting and reconnecting. Only on occasion it identified the correct protocol and then output some gibberish.

Waiting for device:
Listening:

Device disconnected.
Waiting for new device:..............................................
Listening:
rFC wEE R100wF2

remaining: 0012
12@0008 remaps
layerdefs: 00
max_layer: 00
total_macros: 00
alloc ok.


Keyboard ID: 0000
Code Set: 2 (extended)
Mode: AT/PS2

wED W185wED W185wED W185

Device disconnected.
Waiting for new device:........
Listening:
wEE

remaining: 0012
12@0008 remaps
layerdefs: 00
max_layer: 00
total_macros: 00
alloc ok.


Keyboard ID: 0000
Code Set: 1
Mode: PC/XT



Device disconnected.
Waiting for new device:......
Listening:
wEE

remaining: 0012
12@0008 remaps
layerdefs: 00
max_layer: 00
total_macros: 00
alloc ok.


Keyboard ID: 0000
Code Set: 1
Mode: PC/XT



Device disconnected.
Waiting for new device:.....
Listening:

Device disconnected.
Waiting for new device:....
Listening:
wEE

remaining: 0012
12@0008 remaps
layerdefs: 00
max_layer: 00
total_macros: 00
alloc ok.


Keyboard ID: 0000
Code Set: 1
Mode: PC/XT

[Muirium]

Still struggling to get this to work.

This is what I get on HID listen by disconnecting and reconnecting…

This still your Pro Micro running Soarer's converter? Looks like it's resetting over and over. Which still points to hardware, not anything software can do.

[kmnov2017]

I though I'd just mention - every time you see "Device Disconnected", I am physically disconnecting the device and re-plugging it back. I wanted to show the different output I receive each time I re-plug the keyboard. And yes, its still the Pro Micro running Soarer's, except this time the J1 is bridged.

[Muirium]

That info would have helped!

Note that the poor converter is saying XT on subsequent tries. It’s not identifying the keyboard correctly after the first, abortive attempt. The kind of thing you might expect if there’s just not enough juice reaching the thirsty AT original controller. It may be powering up for just a moment each time.

In my experience, comparing Pro Micros to hot garbage is unfair… on the garbage. Swap it out!

[kmnov2017]

Ordered a teensy...

[XMIT]

kmnov2017 sent a PM requesting help.

When building a Soarer's converter from a Pro Micro, some temperamental boards require a roughly 10K pullup resistor on the converter to handle reset correctly. I recall a couple of keyboards benefiting from this addition but cannot recall if the Model F AT was among them. There should be some mention of this on the various Soarer's Converter threads.

I don't have a Model F AT on hand at the moment so I cannot test further.

[kmnov2017]

10k or 1k? Soarers guide says 1k...

[Muirium]

Try both. You’ve nothing to lose. At a guess, I’d trust Xmit. Soarer’s instructions are specifically for Teensies, not Pro Micro.

[kmnov2017]

I tried both set of resistors (1k and 10k), the keyboard still doesn't work. If it's of any help, HID listen shows that at least the current scan set is identified (AT extended).

[Nuum]

I had a problem like that with my Model F AT and for me it was a power problem, as Mµ suggested. I used one of those Blue Cube PS/2 to USB adapters. I can't find one for sale on the quick, but just google "Blue Cube PS/2 to USB adapter" and you should find one.
I see that you're already using an adapter, but it might be worth a try.

I just got the Blue Cube and the keyboard works, so it was indeed a power problem!
For future reference, I'm using a Asrock H77 Pro4/MVP motherboard and it apparently doesn't deliver enough power over the PS/2 connector for an AT Model F, a Blue Cube does work with this keyboard.

Sorry for the trouble I caused and many thanks for all the help.