Pro 2 Controller auto actuating number 2 with modified pin

[CMMS-Freather]

Hi Kanru, thank you for sharing your source code for the controller.

I was following your code and zmk design guide and prototyped my own hhkb pro 2 controller with e73-2g4m08s1c module.
Since e73 did not have p0.11 out, I moved the hys pin to p1.02
while the firmware was able to compile, and able to flash, the board will auto actuate key number 2 itself, causing it unable to use.
Do you perhaps have any idea on how to fix the issue?

The HHKB is confirmed working on yang's controller, and the controller itself will not actuate any keys when it was plugged in alone (main board not plugged in)

I have attatched my schematic for the pinout portion so that perhaps the situation is better understood.

The schematic is designed based on https://github.com/ebastler/zmk-designguide
while the pinout followed your shared schematic.

Any suggestions will be greatly appreciated!

[kanru]

Very cool! Does other keys work?
Is number 2 actuated randomly or stuck in pressed state?

Number 2 is RC(0,0) in the matrix. If we detect key down on every beginning of the scan loop, I guess it may be related to

1. timing
2. residue voltage
3. hys not working?

When I debug this kind of issue I change the scan code to only scan a single key

https://github.com/kanru/hhkb-zmk/blob/main/config/zephyr/kscan/kscan_hhkb_pro2.c#L90-L93

from the matrix

https://github.com/kanru/hhkb-zmk/blob/main/config/boards/shields/hhkb_pro2/hhkb_pro2.overlay#L25-L37

[CMMS-Freather]

So all other keys doesnt work. since it will keep actuating number 2.

i will try debug thru single key. thank you for the advice

[CMMS-Freather]

so i found something interesting.
right now I am using nrfmicro's bootloader
If I make definition in build.yaml board:nice_nano_v2 to nrfmicro_11
it does not input anything on any keys.
if I change to nice_nano_v2, it will random type.
Will this be relating to bootloader/board definition issue?
Since E73 runs in DCCH HV input mode and that will make some changes in the firmware to be adapted

this is before debugging the issue by scan a single key like the suggestion you mentioned above

[kanru]

Sounds like it might be related to the pinout definition?

After changing the code to scan a single key you can measure each pin to see if the voltage are expected.

[CMMS-Freather]

What would be the expected voltage on pins? As long >1?
Will without the connection of the main board and only measure on the controller be enough?

Btw, since the chip os operating in dcdc hv input mode, the vdd is measured at 1.8v. maybe that will be the issue?

[kanru]

What would be the expected voltage on pins? As long >1?
Will without the connection of the main board and only measure on the controller be enough?

yes, that would be enough to verify the pinout are set correctly.

Btw, since the chip os operating in dcdc hv input mode, the vdd is measured at 1.8v. maybe that will be the issue?

https://github.com/joric/nrfmicro/wiki/ZMK/424d80933d500326847a85a039d26b71d55e5a09#hhkb-pro-2

joric wrote that HHKB consumes 0.8W

Hasu also shared some measurement

https://github.com/tmk/tmk_keyboard/blob/master/keyboard/hhkb/doc/Power.txt

In my experience the High Drive mode in 3.3V is enough to power the keyswitch (15 mA)

Nordic's doc https://docs.nordicsemi.com/bundle/ps_nrf52840/page/gpio.html#ariaid-title46

says VDD ≥ 1.7 V can only provide 3 mA in High Drive. So I think it's likely not enough. It looks like you can configure the REGOUT0 register to change the output voltage though.

https://docs.nordicsemi.com/bundle/ps_nrf52840/page/uicr.html#register.REGOUT0

[CMMS-Freather]

i managed to up the vdd to 3.3v
now it reads at 3v
and after modifing the code to actuate number 4 only, I was able to get number 4 key while pressing number 2 key. However, the typing experience was normal, one press one actuation
any suggestions perhaps on this kind of situation?

[kanru]

That is some progress!

Number 4 is RC(1,1) - does your ROW/COL bits output 001001?

Does other keys also actuate?

It sounds like even if you set ROW/COL bits to 001001 the out is still 000000 so number 2 still triggers other keys.

[CMMS-Freather]

Hi Kanru,
thank you so much for the help. I think i made some mistake on initial scan function. after changing it back to your code for the scan for loop. I was able to compile and make it work like normal.

I think main issue was on the power delivery. 1.8v was not enough to drive it but after bump it to 3v3 it was good to go

really appreaciate for your help!