SOLVED: ergodox: row 3 RH: no response (fried teensy?)

[AKmalamute]

EDIT: short version, too many pins. Just because you can connect it there, doesn't mean you should. F5 was one too many.

So, I've attached the teensy to the Ergodox board, and loaded firmware and tested it in aquakeys, and after looking at my layout, I realized the keys that didn't show up were layer-switching keys. No problem, except for row 3.

I've tried three different .hex files, one of which is cubuniac's year-old dropbox hosted TMK build of the qwerty layout. In none of them do I get a response in aquakeys for matrix position 3:8 through 3:13.

I've tested with a multimeter, and found (am I right?) that the pin labelled F4 handles row three ... and there's no impedance on whichever side you'd expect that. I can also find the columns, with impedance fairly low for, say B2 & B3. But shorting B3 to F4 while having aquakeys up does nothing. I can trigger some keys around it, in row 2 and 4 that way, but not row 3. Ever.

Now, I know I damaged it while attaching it. There's a black flea-sized thing that came off. Completely. Near F5 but my pinout sheet says it's AREF. Anyway I reattached it, I think. But the LED near the reset button only very dimly glows. You can't see it anymore in a normally lit room.

But I can, as I've said, load firmware successfully. And make every RH key that isn't in row 3 work ... haven't gotten the left hand attached so I can't test that, just yet.

suggestions? What should I even test for?

[AKmalamute]

Bumping because of hope which springs eternal until the well dries up ...?

[Thechemist]

Can you take a close up picture of the area in question on the teensy?

[AKmalamute]

Can you take a close up picture of the area in question on the teensy?

1st try ... I'll see if I can convince the camera to let me do better

here is the back of the board, and this is the part in question I know I screwed up -- lower left, that thing was unattached for a time



That's about the best I can get of the board as a whole. The blocks of color are post-editing as I told irfanview to "auto adjust" the colors so it wouldn't be so deep into shadows.

[Photoelectric]

It looks like you're putting your phone / camera too close to the part so it doesn't focus properly.  In your 1st try photgraph, the parts farther away are in much better focus.

Also, I'd clean all that flux residue off

In this photo, some of the joints look dull gray instead of shiny silver, and they might be cold joints.  If they are actually that dull gray after cleaning with alcohol and a cotton swab, I'd try to reflow them.

[Thechemist]

I have circled the pins in red I would reflow, make sure you touch the pad and the pin together.

The blue circled pin should be removed from teensy and pcb.

The 1uF cap on the teensy seems to be ok.

I would remove that one pin and reflow the other pins on the pcb and try again.



Edit:

On second thought why is there solder on the pads I circled in yellow?

[AKmalamute]

Alright, I'll try to do that tonight.

The yellow circles ... I don't remember, but I think there's pins there. In the second edited image, the right still-red pin is the one I removed and re-seated last attempt, actually. That's F4 that I'm pretty sure is responsible for RH-R3.

I may not be able to update tonight since I've got social engagements but I have some steps I can try now. Thank you for the input, guys.

[AKmalamute]

Alright, I'll try to do that tonight.

No updated pictures and the board isn't here, but here's what happened with that:

reflow'd pins as suggested, including removing the one pin that had been so much trouble initially, and tried again in aquakeys. Still didn't work, except pin 3:8 caused itself and all but the last (3:13) to trigger, and I think flicker every so often. I never saw them be "depressed" but they registered has having been pushed sometime in the past.

most of the time 3:10 caused itself, :11 & :12 to register in the same way. Thinking I was making progress, I reheated a few not-sufficiently shiny pins, and removed completely and replaced the pin I think handles row 3 (F4) and ... the first time in aquakeys, it did that for 3:8 (but not 3:10) and after that, did nothing. No response anymore.

I even unplugged the soldered USB cable so I could plug straight into the teensy just to prove it wasn't my attachment of the USB connector itself ... no change.

Unless this triggers a thought here, I'm going to remove the teensy and try my backup unit (boy am I glad I bought two of them) and report back with what that changes.

[Thechemist]

Alright, I'll try to do that tonight.

No updated pictures and the board isn't here, but here's what happened with that:

reflow'd pins as suggested, including removing the one pin that had been so much trouble initially, and tried again in aquakeys. Still didn't work, except pin 3:8 caused itself and all but the last (3:13) to trigger, and I think flicker every so often. I never saw them be "depressed" but they registered has having been pushed sometime in the past.

most of the time 3:10 caused itself, :11 & :12 to register in the same way. Thinking I was making progress, I reheated a few not-sufficiently shiny pins, and removed completely and replaced the pin I think handles row 3 (F4) and ... the first time in aquakeys, it did that for 3:8 (but not 3:10) and after that, did nothing. No response anymore.

I even unplugged the soldered USB cable so I could plug straight into the teensy just to prove it wasn't my attachment of the USB connector itself ... no change.

Unless this triggers a thought here, I'm going to remove the teensy and try my backup unit (boy am I glad I bought two of them) and report back with what that changes.

If you can desolder the teensy without damaging either it or the pcb, give it a try and see what happens with the other teensy.

Seeing that the problem is on the RH it points to the teensy since the entire row isn't working.

[Thechemist]

I just did a quick continuity test with my DMM and the pin with a red circle ( f4 ) corresponds to row3 in a teensy while the yellow circle is for the MCP23018 I/O expander. If you have both of the pins soldered to teensy and the pcb this may cause a problem, also if you have a cold joint on the f4 pin. Reflow the f4 on teensy side and pcb side, remove the yellow circled pin and try again.

[AKmalamute]

If you have both of the pins soldered to teensy and the pcb this may cause a problem

You're kidding, right?

As an update ... I just swapped to my "backup" Teensy ... absolutely no change.

My next thought was to beg borrow or steal a working .hex that someone is using on an ergodox built using whatever revision MK is selling, in case it was wired differently than the MD boards whose .hex file I'm using.

But I will remove the inner pin (that would also have saved the 0.1uF resistor, too) and report back.

[AKmalamute]

That was totally it. HOW WAS I SUPPOSED TO ....
*sigh*

A thousand thanks, The_Chemist, and to GH in general.

[/thread]

[Thechemist]

That was totally it. HOW WAS I SUPPOSED TO ....
*sigh*

A thousand thanks, The_Chemist, and to GH in general.

[/thread]

Glad to hear you got it working.