Help me bring the Beige Whale/Moby Click back to life

[Melvang]

Alright guys as some of you have seen I have aquired this beautiful beast of a keyboard.  My end goal is to get it working over USB but I want to keep the look 100% stock down to the cable and connector at the computer end.  If this means putting the teensy inside the PC on a custom breakout board from the plug thats fine.  Or I could put it inside the keyboard and just use that big stock cable to handle the USB duties.  I just don't want to do a hack job.  I don't want to get into cutting PCB's, factory wires, and such.  If I have to solder to factory pads and remove something I could handle that as it would be easily reversible provided I don't lift any pads.

Now herein comes the problem.  To my knowledge this is the first time this has been attempted with this keyboard as pictures of the system it shipped with are pretty scarce let alone shots of the keyboard.  This is not your run of the mill 5291 bigfoot.  The best I can tell this is the keyboard that shipped with the 5324 system.  While, this does appear to have the same case, it is totally different inside.  I have included pictures of what I believe to be all needed to figure out this beast.  Dfj and I have been speaking in irc and it appears to be very possible to get this working as it has its own cpu, cap sense chip and power regulation (though input power is +12 volt and dfj is a little nervous of feeding that much power for fear of frying something.

This is the orignal thread for this.

I do have the pinouts for both the internal cable that runs from the back panel of the computer to the cpu planar board and from that plug to the "adapter card" or daughter board.

Pinouts are as follows.



This image was pulled from this manual in chapter 3 around page 87 of the pdf file.

This is where it gets interesting for my understanding.  I would like to use soarer converter/controller.  By my impressions stuff with its own controller would use the converter firmware and few wires or in the case the the 5291 Bigfoot without its own controller you would use many wires and the controller firmware.  How am I going to get this working here??

Please help me get this thing running.  I have not had the plates apart and I don't think I want to until I have proof of concept that this thing will run.

Thoughs comments and suggestions more than welcome.

More pics can be provided on request.

[dorkvader]

This is where it gets interesting for my understanding.  I would like to use soarer converter/controller.  By my impressions stuff with its own controller would use the converter firmware and few wires or in the case the the 5291 Bigfoot without its own controller you would use many wires and the controller firmware.  How am I going to get this working here??

Please help me get this thing running.  I have not had the plates apart and I don't think I want to until I have proof of concept that this thing will run.

Thoughs comments and suggestions more than welcome.

More pics can be provided on request.

Soarer controller (not converter) is required. The KB has sensing and basic chips but no real protocol, fortunately soarer's controller can read it.

There is even a sample configuration file for it in his download.

The only issue now remains wiring it up. Other than power / GND, the traces can go to any pin on the teensy, though if you choose different ones that soarer, you'll have to change the config file.

You can just wire it up to the cable keeping the KB 100% stock.

[Melvang]

This is where it gets interesting for my understanding.  I would like to use soarer converter/controller.  By my impressions stuff with its own controller would use the converter firmware and few wires or in the case the the 5291 Bigfoot without its own controller you would use many wires and the controller firmware.  How am I going to get this working here??

Please help me get this thing running.  I have not had the plates apart and I don't think I want to until I have proof of concept that this thing will run.

Thoughs comments and suggestions more than welcome.

More pics can be provided on request.

Soarer controller (not converter) is required. The KB has sensing and basic chips but no real protocol, fortunately soarer's controller can read it.

There is even a sample configuration file for it in his download.

The only issue now remains wiring it up. Other than power / GND, the traces can go to any pin on the teensy, though if you choose different ones that soarer, you'll have to change the config file.

You can just wire it up to the cable keeping the KB 100% stock.

So would I wire it up through the adapter card or directly to the pins on the actual PCB?

[dorkvader]

the large heatsinked part indicates that there is likely some voltage conversion and regulation going on. I would advise against it. See if you can find the part numbers for it and the 2 chips on the card to make sure.

I think DFJ or kbdbabel has a schematic somewhere, look for it.

[Melvang]

the large heatsinked part indicates that there is likely some voltage conversion and regulation going on. I would advise against it. See if you can find the part numbers for it and the 2 chips on the card to make sure.

I think DFJ or kbdbabel has a schematic somewhere, look for it.

According to DFJ those pairs of chips with 2 and 4 leads are caps and resistors.  The three pin chip handles the voltage regulation duties.  With a 12V, 5V, and GND in the middle

[Melvang]

[dorkvader]

(Attachment Link)

this is why I love geekhack.

What's the verdict? letting the smoke out already?

[Melvang]

(Attachment Link)

this is why I love geekhack.

What's the verdict? letting the smoke out already?

Please ignore the horrible potato shot in the previous post.  I have found a 12 volt power source that is easy enough for me to work with.  The battery in one of my cordless tools measured 11.52v, I figured that was close enough.  I started by feeding 12v into the plug that comes from the computer.  This resulted in 5v leaving the card where it plugs into the actual keyboard pcb which is what dfj and I were suspecting the one chip with three pins was doing.  Next I fed 5volts from usb power across the pins that showed 5v from the battery and I still got 5 volts where it leads into the keyboard pcb.  This is where I am at currently

[dorkvader]

Forgot to link this thread of interest. This way anyone reading this thread will also read the thread I link here and info will be in teh same place.

https://geekhack.org/index.php?topic=48950.0

from that topic, here's what soarer has to say:

^from kbdbabel. Pin names (D5, D6) are general names for data pins not teensy pin assignments)

Here's xavierblak's controlelr photo. Note his is missing the adapter card board with the vreg on it


aha! it matches the mant library pic of it:

so now we know what all those chips are.

Here's the pinout from the hardware manual:

Code: [Select]

           D1  1      20  Ground
(missing pin)             -Strobe
           D2             Ground
           D3             D0
     (unused)             (unused)
     (unused)             (unused)
     (unused)             Frame Ground
           D4             +5V
           D5             Ground
           D6  10     11  Output
Here's a better description of the connector. Note how this compares with the kbdbabel image of the cable pinout.

I don't see specifically what pins go where but 5v goes to 5v or Vcc. GND goes to the GND pin on the teensy. All the other pins go to data pins on the teensy. You can actually see which ones by looking at his picture:

^soarer's 5291 wired up and ready to go. You can see 5v (red) and GND (black) plugged in at the proper locations 13 and 12e GND to co to any of the GND pins which are 12 20 and 18.

Also if you get random capacitance errors you might need to make sure the frame is connected to "frame ground. If you can tap into the USB cable's shield (usually it's tied to ground at the cable end unless you have made your own) then this should be connected to frame ground as well (Pretty sure).

Looking at the controller config file also tells a similar story:

Code: [Select]

scanrate 1
debounce 1
blocking 0
muxstrobe_port PB6:0
sense_delay 1
muxstrobe_gate -PD1
sense_polarity 1

sense PD0
If I were you, I'd look hard at the image of him wiring it up to a teensy, and follow those pins and write a table of pin # on connector, pin name on teensy and pin assignment from the chart in the maint. manual and pin assignment from kbdbabel link.

Then see if you understand what's going on with the controller.

(basically it looks to me like a 1*83 matrix that exists as the output of several muxstrobe chips The location in the list of the muxstrobes is related to the key's position in the matrix. This is the sort of thing that I still don't understand *too* well.)

On teh plus side since there is no controller for the KB it's a LOT easier to convert to USB with soarer's code there. It also becomes easy to "overclock" the KB. Soarer said it supported fully a scanrate of 1000Hz, about what a cherry KB does now, and IIRC much faster than most IBM KB controllers (which were something like 14 us scan time).

[Melvang]

dfj, here is the output I was getting from the JD40 when holding a wire from the LA to the one pin that was always high.

[dfj]

A couple of important distinctions here: 1) is that melvang's beast has a controller, and 2) *outputs* the scancode it just sampled, as opposed to the 5291 which accepts the code to scan, decoding it onto the matrix.
 Still - it should readily convert... assuming it is in working order.

until,
dfj