Can we design the teensy alternative for keyboards?

pomk · 01 Dec 2016, 13:06

tentator wrote: what would be the issue if we keep the pullup?

or what would be the issue if we disable spi in the bootloader/flash as default/factory? (is this not possible because the chips arrive "empty"?)

Depending on the strength of the pullup required for stable operation, it could be used as an input directly. The internal pulldown on the chip should be maybe at least twice as strong for that to work reliably. If used as an output, there would not be a problem in its use.
As for pre flashing the chips, it will raise the price quite a bit, especially if they don't have the tools already in place for this specific chip. As for getting the chips pre flashed at the chip factory, I don't think that we are a large enough player for them.

If we want the boards flashed, a board edge connector could be used (think pci express) to which the board would be crammed into, and then a SWD debugger could load the first firmware at the factory in a very automated manner.

Wodan · 01 Dec 2016, 13:12

I think there should be SOME kind of bootloader on the controllers even if it means higher costs. Although this is an enthusiasts community, only a few people will have the means to flash a blank Atmega µC.

Unless someone proxies all the orders and offers manual pre-flashing as a service ...

matt3o · 01 Dec 2016, 13:28

pomk wrote: As for pre flashing the chips, it will raise the price quite a bit, especially if they don't have the tools already in place for this specific chip. As for getting the chips pre flashed at the chip factory, I don't think that we are a large enough player for them.

it would be feasible but we really can't do that for now, too expensive for us.

Wodan wrote: I think there should be SOME kind of bootloader on the controllers even if it means higher costs. Although this is an enthusiasts community, only a few people will have the means to flash a blank Atmega µC.

Unless someone proxies all the orders and offers manual pre-flashing as a service ...

we have the ROM bootloader for that (if we can make it work)

pomk · 01 Dec 2016, 13:50

Wodan wrote: I think there should be SOME kind of bootloader on the controllers even if it means higher costs. Although this is an enthusiasts community, only a few people will have the means to flash a blank Atmega µC.

Unless someone proxies all the orders and offers manual pre-flashing as a service ...

There is the rom bootloader always present. the problem is with that it tries to use SPI as well due to a resistor missing.

Wodan · 01 Dec 2016, 13:56

Okay got it now. In that case I believe it's a good idea to look for a HW solution for that floating PIN. Using a pull...up? resistor sounds like a great idea since that can also be chipped off with a sharp screw driver having the same effect as a cut trace should there be scenarios in which the resistor combination doesn't cut it.

pomk · 01 Dec 2016, 14:11

I'll try with a 30k resistor first, that should be weak enough so that the internall pulldown will cancel that efficiently if that pin would be used as a input for the switch matrix.

matt3o · 01 Dec 2016, 14:20

Hasu suggested to use a 10Kohm or lower.

pomk · 01 Dec 2016, 15:03

That will not be weak enough for the internal pull down to 'cancel', as that is about as strong as well.

matt3o · 01 Dec 2016, 18:13

IT WORKS!

: test.gif (378.38 KiB) Viewed 5436 times

ohmy I'm happy like a child in a candy store!

Thank you so much to Hasu and pomk! Now let's get this resistor fixed so we can go production!

Like I said the only way I have to access the bootloader is from a windows tablet not connected to the wall power. Every other PC I have fails to connect. But anyway at least we know the board is working!

Now I'll try to make a decent keyboard firmware out of it

Laser · 01 Dec 2016, 20:26

Maybe compatible to Hasu's TMK

?

PJE · 01 Dec 2016, 21:08

I have an option jumper which you can solder - it's two semi circles with a small gap been the two flats. If you add a connection between the two halves on the schematic you get a cut track which can be resoldered as needed.

Sent from my SM-G900V using Tapatalk

matt3o · 01 Dec 2016, 21:12

Laser wrote: Maybe compatible to Hasu's TMK ?

believe it or not, that blinky thing is already TMK firmware!

Laser · 01 Dec 2016, 21:27

Heh, cool then - firmware problem solved

pomk · 01 Dec 2016, 21:37

matt3o wrote: IT WORKS!

Great job! I should have time this weekend to modify the PCB layout accordingly.

tentator · 01 Dec 2016, 23:39

cool! but did not understand, did you flash it because that tablet somehow is not interfering too much to let the rom bootloader not go towards SPI or is it because you soldered a 10k resistor somewhere now?

matt3o · 02 Dec 2016, 08:31

tentator wrote: cool! but did not understand, did you flash it because that tablet somehow is not interfering too much to let the rom bootloader not go towards SPI or is it because you soldered a 10k resistor somewhere now?

no resistor was used. that tablet I have somehow doesn't interferes with the rom bootloader via USB. Don't ask me why.

ohaimark · 02 Dec 2016, 08:34

Did you try with a laptop as well? It almost sounds like a noisy ground connection causes the issue...

pandp · 02 Dec 2016, 08:36

matt3o wrote:
tentator wrote: cool! but did not understand, did you flash it because that tablet somehow is not interfering too much to let the rom bootloader not go towards SPI or is it because you soldered a 10k resistor somewhere now?
no resistor was used. that tablet I have somehow doesn't interferes with the rom bootloader via USB. Don't ask me why.

How nice! You have all needed production equipment to run GB

matt3o · 02 Dec 2016, 08:41

ohaimark wrote: Did you try with a laptop as well? It almost sounds like a noisy ground connection causes the issue...

doesn't work on a Mac laptop

I'll try with another laptop later.

pandp wrote: How nice! You have all needed production equipment to run GB

okay, let's add the resistor anyway

tentator · 02 Dec 2016, 19:24

That tablet for sure doesn't cost as much as a dedicated flash programmer ahah

You know maybe you have less issues in flashing there because you use a otc usb cable with the 5 pins instead of the traditional 4??? Could be.. then you just need a cable crafted for this purpose

))

pomk · 02 Dec 2016, 19:53

the fifth pin is not connected anywhere on the board though

Wodan · 02 Dec 2016, 20:01

I need 10

tentator · 02 Dec 2016, 20:04

But is somehow cleaning up the d+ pin with a resistor to nowhere.. anyway still a mistery..! You see what I mean..

pomk · 02 Dec 2016, 20:56

tentator wrote: But is somehow cleaning up the d+ pin with a resistor to nowhere.. anyway still a mistery..! You see what I mean..

It has probably more to do with the USB controller on host side. On my PC, some usb 3 ports work flawlessly, while usb 2 do not.

Scarpia · 02 Dec 2016, 21:23

Sorry for asking a silly question, but.... The hole pattern for the two switches isn't Alps compatible, is it? From the early posts I got the impression that the PCB design was aiming to be universal, but looking at the prototypes I can't see how it would work. Or am I missing something?

pomk · 02 Dec 2016, 21:40

It should be alps compatible. I dont have a switch to test it with however. The footprint complies with matias spec.

matt3o · 02 Dec 2016, 22:26

yes, it works with alps

Norman_ · 03 Dec 2016, 01:10

Have you considered making one as small as possible for converters specifically?

Give it the minimum amount of pins and keep it small, ideally it would be cheaper as well.

tentator · 03 Dec 2016, 01:24

well I suppose you cold do that without a pcb by just buying an atmega32u and then handwire a couple of wires to usb and to ps2, done

okok, maybe a couple of piggybacked resistors as well..

pomk · 03 Dec 2016, 01:58

Norman_ wrote: Have you considered making one as small as possible for converters specifically?

Give it the minimum amount of pins and keep it small, ideally it would be cheaper as well.

This one already has almost the minimum components. The reset switch could perhaps be removed, as well as the led. The price would stay almost the same however, as the pcb area is ~free when compared to the p'n'p and chip + usb socket.