New CPU Project

[antto]

contact him, use the emal from the .net domain

yes, this new cpu is all about MOAR code and thus MOAR features
polyphonic midi sequencing, drum sequencing, anyone? ;]
and moar blinky LED animations

[mario1089]

..and a powerful, flexible sequencing core engine? where one just writes down what one wants to happen? and then the event queue gets it right? and we have enough flags at a single step? that I can set certain modes per step? semitone toggling within a note? stepped slides? and a proper swing at the Dinsync out? and the entire ProTools (uargh!) groove library in it, so that simple 1/16 swing is just one of 100 groove options? and this again happens at the dinsync output also? well, could be fun (but then again, maybe this is all just not enough 303-ish for some here ).

[antto]

it's HERE \o/

[xlarge]

Textile came through. Awesome!

[antto]

so i got the new cpu plugged in, didn't damage the old one
i only had to desolder C105 cuz it was sticking out too much, i have lots of spare 1uF caps, soldered one with longer legs and it bends down now out of the way

the cpu runs, you can tell the tempo works differently now, much better
it uses the eeprom from the mainboard, which is kewl, no need to swap it
c0nb0x connects to it without a problem..

now, i tried to get it into bootload mode..
i expected it to act like the old cpu - restart in "BOOTLOAD" mode and it gets stuck, pressing buttons or turning switches doesn't do anything - well, not quite
if it's supposed to behave like the old cpu, then the symptoms are that it doesn't get into bootload mode at all

after i reboot it, for about a bunch of milliseconds i get that semi-random LEDs being lit up, and then it immediately clears them and lights up Bank1 LED and stays there
then, if i turn the FUNC knob to other modes - it *does* change, aka, it's the firmware

i downloaded the bootloader source code from guest's website, i see where he has modified it - specifically in main() he has added a peice of code to check the FUNC switch, so i assume it should work the same way as the old bootloader in that respect

i tried the avrdude command from guest's website, tho i didn't see how that would work given that i am pretty sure the thing doesn't really go into bootload mode

Code: Select all

avrdude -p atmega2561 -c stk500v2 -P COM3 -b 57600 -e -U flash:w:x0xb0x2.hex

and as expected, it doesn't work

avrdude: stk500_2_ReceiveMessage(): timeout
avrdude: stk500_2_ReceiveMessage(): timeout
avrdude: stk500_2_ReceiveMessage(): timeout
avrdude: stk500_2_ReceiveMessage(): timeout
avrdude: stk500_2_ReceiveMessage(): timeout
avrdude: stk500_2_ReceiveMessage(): timeout
avrdude: stk500_2_ReceiveMessage(): timeout
avrdude: stk500_2_ReceiveMessage(): timeout

and if that's not enough, i tried one more thing to see if whether it runs the firmware or not
restarted the x0x in "bootload" mode, again, and i tried to connect with c0nb0x
it connects, so it *is* the firmware

Altitude: you were involved in this, can you please give me a hint what to do next?
do i have to buy that programmer device? or am i doing something wrong..

[mario1089]

you can tell the tempo works differently now, much better

so what`s that?

[darffader]

you can tell the tempo works differently now, much better

so what`s that?

http://wiki.openmusiclabs.com/wiki/x0xb0x

The only major difference between this firmware and the stock v1.05 is that the tempo knob is sampled rather than run on an interrupt. This was changed due to its lines not being connected to external interrupts on the new CPU.

[Altitude]

what are you using to program it? I have one of these btw so we can probably work in parallel

[antto]

what are you using to program it? I have one of these btw so we can probably work in parallel

uhm.. i thought it's flashable over USB like the old cpu
or is it not?
i don't have the "programmer" device thing

[Altitude]

Where did you get it from? If you built it from scratch, the AVR will be blank but it sounds like you already have something there.. That mega has a JTAG programming interface so you need the JTAG programmer to flash the bootloader and I dont know if the USB code upload was ever done/tested when guest was working on it, i'll go back through my correspondance with him and see what I can find. I tried to just wire in the ISP pins and use my AtTiny that I had at the time but that thing did not support the avr that was in there so I shelved the project

[antto]

well, textile sent me the cpu ready-built, it's programmed too (it runs the stock firmware)

[Altitude]

You should check with him if there is a boot loader there, I dont know if he followed what guest did or just set out on his own. There are some AVR JTAG clones on ebay for $10 and I ordered one so I'll be set to go on this soon.. We should probably get Mario in the mix also

[antto]

i'm waiting for a reply still
from his last letter he said:

Have you looked at Guest's page for the bootloader?
http://wiki.openmusiclabs.com/wiki/x0xb0x

It uses AVR dude , you may have better luck with the method described there.
I used a direct program to load the firmware so you may have to re-boot the firmware with avr dude.

i don't quite understand what he meant there

[Altitude]

hmm, thats a little nebulous. I'll play around with it when my programmer gets here. You might want to think about ordering one: http://www.ebay.com/itm/AVR-USB-JTAG-em ... 2320949afa . It claims to be a clone compatible with AVR studio and considering the official one is $400, its worth trying for sure

[antto]

i spoke with another person, he linked me to a ISCP cable thing, i'm confused

i suppose this is about those 10 pins that stick out of the cpu mod PCB, right?