Objective Development Forums

You're activating pull-ups on D+ and D-. These bits of PORTD must be initialized to 0.

What do you mean with "appears and instantly disappears"?

The GUI tool seems to be compiled for a customized device name, maybe even a different vendor name. You can try to change the name of the device accordingly.

It's possible that the code exceeds 2k due to the longer addressing instructions. You van check the size with avr-size main.hex.

I don't think that it is technically possible to flash through USB to RS232 converters reliably. USB support in PonyProg may refer to USB based programmers.

I would recommend that you build a simple programmer such as USBasp. You can flash it with a computer which has RS232 or a parallel port.

Please try enabling debug mode in a 4k boot section. If it's a problem with descriptor addressing, you should see that wrong data is sent to the host for getDescriptor requests. If it's a problem with interrupt vectors, the program should crash pretty soon.

That should be correct, but I have never tried so large devices myself, so I don't know for sure.

If the checkboxes with hook indicate a flashed (=0) bit, then the cksel values (=1111) should be OK. These are the most important settings. The bodlevel of 2.7 V should also be OK...

The compiler uses absolute addressing on bigger devices. Have you told the linker to set the base address accordingly?

If you provide your own host software, you don't really need a precise description in the report descriptor. You don't want to explain the device to a third party's driver, after all, as it is the case for mice, keyboards and joysticks. Please have a look at the Automator example. We transfer data i...

Crystals never have deviations of more than several 100 ppm. That's certainly within the tolerance (of much less than 1% for crystal clocks, BTW). If it works on one computer but not on another, you probably have voltage level problems on D+ and D-. Please have a look at the zener diodes and voltage...

Regarding slow programming with PonyProg: USB to serial converters are very slow when handshake lines should be toggled and most of them don't even work for this purpose because they optimize away short high or low pulses. Have you been able to verify that the flash content is correct? Does windows ...

So it is normal that my device might dsconnect and then connect again without getting into trouble with the client program sending commands to the device? No, this is not normal. But IF you get these resets, you can at least fight the symptom and write the firmware, driver and application so that a...

See the project's Makefile: # Fuse high byte: # 0xdd = 1 1 0 1 1 1 0 1 # ^ ^ ^ ^ ^ \-+-/ # | | | | | +------ BODLEVEL 2..0 (brownout trigger level -> 2.7V) # | | | | +---------- EESAVE (preserve EEPROM on Chip Erase -> not preserved) # | | | +-------------- WDTON (watchdog timer ...

I've tried your code just now (sorry, did not have the time to do it earlier). First one comment: You have set F_CPU and DEBUG_LEVEL in main.c. This is not the place where they should be set since these are global settings for all modules. They should be added to the compiler command line with -DF_C...

I think you must have another error such as the swapped supply lines.

Rebuilding the device from scratch will probably get this resolved, since you rarely make the same mistake again. The wire type does not matter, though.