Firmware upgrade disaster!!!

[kurtus420]

thanks dmitri, I guess if I have no luck programming new pic via icsp cable then I'll try this because I'm certian I have no shorts anyware on the pcb board and I've tried 3 computers that can't all have bad com ports. Is there somthing I can test on the pic to see if there is some kind of a schematic error or shorts?

[dmitri]

thanks dmitri, I guess if I have no luck programming new pic via icsp cable then I'll try this because I'm certian I have no shorts anyware on the pcb board and I've tried 3 computers that can't all have bad com ports. Is there somthing I can test on the pic to see if there is some kind of a schematic error or shorts?

I know only one way - with a meter test for continuity and shorts all the relevant tracks/pins using the schematic as a reference.

[kurtus420]

All right, thats what I sort of figured. I'll check it all out as I guess it's possible there is some really small bridge that my eyes can't see. As always, thanks for your help

[el-dr]

GLORY AND GRATITUDE!

Today I bought 3 Atmega644 20PUs and 1 PIC18F2550. I went back to the same old WinME machine with good old COM and LPT ports that I used for reprogramming the PIC two days ago, installed PonyProg, plugged in the ISP cable I made yesterday and furiously programmed all 4 Atmegas in about 10 minutes (including the old one that had a different bootloader to Dmitri's). It was a kind of revenge for the nightmare I went though in the last 5 days. Then I programmed the new PIC, just to show off my excellence, changed the 20MHz crystal to another one, loaded the 20090617 fw into all 4 MCUs and I had a fully functional MD with USB, MIDI and external power. Now I only need to replace the broken 4851 when it arrives and I'll be done, but in the meantime I'll execute some thorough tests with 24 inputs. I'll use MD live, so I don't want to take chances any more... I hope my next post will be the next video (after the recently posted http://www.youtube.com/watch?v=6WdMSnWhppk) and not another cry for help

[Ken Forgettable]

The latest version of PonyProg should be able to program Atmega644 just fine.

Your advice to use AVRdude still lingers in the 1st post of the 'Latest MegaDrum firmware versions' download thread.
It should be removed now?

[kurtus420]

Well still no luck with this issue, I installed a new pic18f2550 and still cannot program it so clearly there is an issue with the cable I think. I used 18 guage 4 conductor solid cable with 1/8watt 10k ohm resistors. I have checked for continuity on the cable as well as ohmed it out to ensure everything is correct. The db9 connector that I used is labled so I know the pins are correct, this cable is about 3 feet long. I have check the schematics to the board and there are no shorts at the pic or the icsp connect (I checked from the icsp header to the pins for continuity and found no shorts). Checked continuity to the pins while the cable was plugged into the icsp header and all correct readings from my dm. I have tried xp and windows me with winpic and no success here either. I'm seriously stumped here, is solid cable bad to use? is 1/8watt resistors going to cause a problem? I guess I will have to build another cable with 1/4 watt resistors and standed wire to see if this is the issue.

thanks dmitri, I guess if I have no luck programming new pic via icsp cable then I'll try this because I'm certian I have no shorts anyware on the pcb board and I've tried 3 computers that can't all have bad com ports. Is there somthing I can test on the pic to see if there is some kind of a schematic error or shorts?

I know only one way - with a meter test for continuity and shorts all the relevant tracks/pins using the schematic as a reference.

[dmitri]

The wattage of the resistors doesn't matter and neither type of the cables as long as they reasonable short. Anything shorter than 1 meter is reasonably short.

[kurtus420]

I just double checked the cable and it's ohming out at the lower side of 9k ohms instead of 10k at the end of the cable leads, could this be the problem?

[dmitri]

I just double checked the cable and it's ohming out at the lower side of 9k ohms instead of 10k at the end of the cable leads, could this be the problem?

No.

[kurtus420]

looks like I have a buggy os or version of winpic because the third time I tried winpic went alot slower and it still said that it error out. I just went ahead and installed the pulldown resistor on a whim and plug it into my laptop which recognized my megadrum as megadrum usb, ran midiox did the firmware update and it didn't miss a beat so I'm finally working correctly. Thanks so much for all the advise you have given me Dmitri . DAMN COMPUTERS ANYWAYS
I assume that this pic is successfully configured right? It seems to be working fine and this was a brand new pic so it definitly had no firmware on it. I can't even get a read from the pic either which is strange.