Programming the ICs

[japi]

I've read some people had problems with palm software/etc that capture the serial port, once de-installed everything went ok. Do you have another pc to give it a try?

[gastric]

I have a clean WinXP SP2 machine that's a bare OS only. So there's no risk of interoperability issues with other software. It's likely something to do with my build though I'm unsure what at this point. I'm getting 5V as I should, there's no shorts or breaks that I'm aware of and I tested all traces prior to soldering, and all solder points after soldering components. I'll have to spend some time pouring over Dmitri's suggestions for Estregan in the (true)MIDI USB Board thread where he lists of various values to compare, though those are for a completely different build so I'm guessing they are not going be directly apply to a V2.5 board build.

For the PIC when using WinPIC800 it can erase the PIC and can read the PIC without error. Its the programming that I have errors with. WinPIC800 errors quickly with "ERROR -> Writing address 0x000000 Written: 0xEF43 Read: 0x0000" So I apparently have some serious issues. I did manage to get WinPIC to seemingly program the PIC correctly once, at least it didn't generate errors, but the board was never recognized successfully by any PCs I connected it to. It would sometimes get recognized but reported as not working properly.

[japi]

Maybe is a dumb question but is the ICSP connector plugged the right way? Is pin1 where it should? What about the programming cables, they are triple checked?

[gastric]

Cables are correct, triple checked, continuity is perfect. I even removed the connectors from the cables and soldered the cables ends directly to the USB and ISCP headers to eliminate some issue with the connectors. I've double-checked my component orientation for those that require specific positive/negative orientation (1N4148, 47uF, 1uF, trim pot) and it appears correct. I'm assuming the 1N4148 has the positive where the black band is???? See them installed on the board in the picture below.

PIC SERIAL END - note I wired pin 4 back to pin 8 instead of soldering it mid-wire onto the lead coming from pin 8 though that should make no difference I think


Picture of board, including JACKS board USB wired to USB header on main board. Note pin 1 on JACKS is on left as pictured, while pin 1 on main board is on right and I've correctly wired everything correctly to that regard.


On the PIC ISCP header pin 3 is ground, and ISCP header pin 2 is correctly getting 5V as is pin 18 on the PIC. ISCP pin 4 connects to PIC pin 2, ISCP pin 5 connects to PIC pin 4, ISCP pin 6 connects to PIC pin 6. They're all passing continuity. Same with the USB header to the appropriate pins on the PIC. And I'm testing the actual leads on the PIC, not the socket, so my traces, solder points, and PIC socket are all connecting properly to the PIC itself.

Thoughts???

[bobnickelson]

I used Winpic, and it recognized the pic18f2550, and programmed it aswell. However i never got it to be recognized on any computer, plugged the usb in and got nothing. i tried Winpic800, and something else and still can program the pic but its never recognized via usb.

dmitri told me to check pin 14 on the pic18 with the usb plugged in, and see if it reads 3.3v, which it didnt for me on any pc. Since we both ordered from mouser, and used the 2.5 design, we might have the same problem.

I noticed that you soldered your icsp directly to the board, and you mentioned that you tried programing atmega, but did you use the atmega icsp cable? (doesnt use the pic18's, has its own that connects to the parallel port) i managed to program the atmega fine, i get verfication (midi led) from hiting a test sensor that i hooked up.

Also does the 47uF and the 1uF orienatation really matter? because i got them backwards, and the atmega works fine and sensors work fine. I just cant get the pic to work...

[gastric]

As I noted I get 5V on the power pins being fed by the USB jack (by following the traces on the board). There's no reduction from 5V occurring on my board at all that I can see.

Yes, I tried to program the Atmega with it's own cable, which has a modular connector on it, I simply removed the same type of connector from the USB and PIC cables to ensure a clean connection as the modular ends tend to let the cables come loose.

As far as the 47uF and 1uF I do believe they need to be orientated specifically. The board indicates a positive (square) and negative (no square), and both components have corresponding positive/negative sides, as do the diodes. The data sheet for my surface mount 47uF indicated the positive was on the side with the irregular shape. My 1uF didn't indicate positive, but had a stripe down the side that indicated negative. The diodes I just assumed the end with the stripe was the positive, which appears to jive with pictures of other peoples builds. But it's entirely possible I've installed them all ass backwards.

Overall I've set this project off to the side until Dmitri returns (this upcoming weekend I believe) as apparently he's the only one with enough knowledge to assist with this issue. I was actually hoping Synthex would pipe in with some comments. I'll PM him to do so in case he can help.

[Synthex]

Yes, 47uF and the 1uF orienatation really matter. The negative pin must be connected to the ground.

[gastric]

Synthex, what about the voltage issue? I'm seeing 5V, which is what the USB provides. Supposedly Dmitri indicated it should be 3.3V to the PIC. Programming the PIC should be easy from what I can see. The tools are simple, there's not a whole lot of connectivity from the ICSP and USB to the PIC. So I'm assuming it should be the easiest to troubleshoot. Can you specify how to properly test the V2.5 board to ensure it's constructed and feeding the PIC the proper voltages, etc.? There has to be something obvious and critical wrong with my board, though I can see nothing.

Just to clarify, when I'm counting pins on the ICs I'm calling pin1 the pin with the square as indicated on the board layout, and every pin in that same row is the odds, and every pin in the opposite row is the evens. Thus, looking at my board, when I say pin 18 on the PIC I'm referring to the the 9th pin from the bottom in the left row. Pin 2 would be the first pin at the bottom of the left row. My PIC label specifically says PIC18F2550-I/SP then 0827SDC in case that makes any difference.

[japi]

Just to clarify, when I'm counting pins on the ICs I'm calling pin1 the pin with the square as indicated on the board layout, and every pin in that same row is the odds, and every pin in the opposite row is the evens.

I think thats not right, as per the specs sheet of the pic, the pins are like this:

[Synthex]

Did you test the PIC without the Atmega ?