exxos's Atari, Amiga & retro forum

Afraid not no, though I may well have another go at some point.

It won't be soon though, as if everything goes to plan, we will be moving house in the next 12 weeks, so a bit otherwise occupied right now.

Accidentally, I've bumped into this thread while searching for something else but in the end, perhaps it is the same cause (missing a delay and connection to /RESET).

There are at least two HID (USB mouse/joystick/joypad) adapters to the DSUB9 connector which seem to have trouble with the MegaSTE/TT keyboards. And in my case, also in my Falcon.

Basically, in the Falcon case, the adapter (powered by DSUB's +5V) would confuse the IKBD controller upon boot. The confusion demonstrates as key presses not doing anything or printing different character each time (!). Interestingly, this behaviour is fixed when I press reset. Which sounds awfully similar to the known problems with the NetUSBee and CT60 power lines (where a delay waiting for +5V to stabilise is put in the works).

However in the MegaSTE/TT case, this doesn't help at all. And this thread has shown me why (at least that's my guess) -- the keyboard doesn't have a RESET signal connected. So unless I miss something, it seems that the keyboard is in the same state regardless if how many soft and hard resets you do, therefore unable to reset one more time (when +5V is OK).

Which brings me to question whether you @rubber_jonnie (or anyone else) had any luck making this adapter / delay circuit work?

EDIT: It seems that a year before @Steve managed to get it working even without the /RESET connection: https://www.atari-forum.com/viewtopic.p ... 04#p387304.

mikro wrote: 04 Jul 2024 10:50
Which brings me to question whether you @rubber_jonnie (or anyone else) had any luck making this adapter / delay circuit work?

I didn't, no and ended up getting Eiffel adapters, and the idea slowly slid off my radar.

The reset signal is there to keep the IKBD processor in reset until the system is up and running. The Mega/MegaSTE/TT keyboards have a timer circuit of their own to hold them in reset until the power is stable so doesn't need the one from the system.

I'm not sure what you mean by delay, the IKDB protocol is merely a serial connection.

Maybe my wording wasn't correct, I was referring to the delay discussed in this thread.

Thanks for the information about the timer! Apparently, it's not enough for the USB adaptor. And the lack of RESET means that If the timer screw up, there's no point of return, unlike ST/Falcon.

mikro wrote: 04 Jul 2024 10:50 However in the MegaSTE/TT case, this doesn't help at all. And this thread has shown me why (at least that's my guess) -- the keyboard doesn't have a RESET signal connected. So unless I miss something, it seems that the keyboard is in the same state regardless if how many soft and hard resets you do, therefore unable to reset one more time (when +5V is OK).

Even without a hardware reset signal, resetting the computer should still reset IKBD because TOS sends a reset command to the IKBD. Of course, if the IKBD processor is locked for some strange reason, it might not be able to receive any command. I would need to double check, but IIRC IKBD might lockup if a mouse button is pressed during start up.

mikro wrote: 04 Jul 2024 14:31 Maybe my wording wasn't correct, I was referring to the delay discussed in this thread.

Thanks for the information about the timer! Apparently, it's not enough for the USB adaptor. And the lack of RESET means that If the timer screw up, there's no point of return, unlike ST/Falcon.

You can unplug and replug the keyboard in to reset the TT/MSTE/MST keyboards, that does a power cycle on them.

@stephen_usher that's exactly what I do to work around the problem. The annoying part is that it is the keyboard's voltage problem, not PSU's so I have to either boot with the adaptor detached (easier) or to disconnect the keyboard, then the adaptor, then reconnect the keyboard and then the adaptor (annoying).