Objective Development Forums

If you can, use gcc 3. It compiles smaller code.

Other than that, you can omit some functionality (e.g. EEPROM access) or include a custom vector table without all the unused slots. The latter is a bit tricky, though.

HID allows a wide spectrum of devices. If you implement a keyboard, it is as you say: You point the cursor at a window and the device writes the data. Most applications implement a generic HID device. This allows (almost) arbitrary communication in fixed size blocks between the host and the device. ...

This is a problem with your device. Probably something with the wiring, with the zener diodes or with the fuses. Did you compile for the Mega8? I think the hex files which ship with the examples are compiled for the Mega168.

I'm running AVR-Doper with Mac OS X in both modes. The message you mentioned is harmless. There is no need to detach any operating system drivers.

Your development environment seems to lack some of the USB driver libs. Did you try MinGW?

Did you try to run V-USB in debug mode and connect a terminal to the serial debug output (19200 bps)? This way you can check whether the same data arrives at V-USB which was sent by the OS and vice versa. If there is a discrepancy, something is broken in the low level stuff (e.g. data loss due to in...

The 16.5 MHz module is more stable than the 12.8 MHz module because everything is within the specs of the RC oscillator. The big issue is calibration and V-USB contains only suggestions how this can be solved. If you have a good calibration method, the RC version should be similarly stable as a crys...

There should be no problems mixing SPI and USB. I think there are even examples which convert SPI to USB. I do know about TWI to USB examples.

If your main() routine begins with a usbDisconnect() which lasts for some time (all our examples do that), then I don't see a difference to initiating the disconnect somewhere during runtime. If your main() does not do that, the host won't know that you had a reset. The USB driver's usbDeviceAddr va...

If the problem is only with the 328p, please look at the following thread: viewtopic.php?f=8&t=2663

Why should this prevent the disconnect/connect sound? A watchdog reset also triggers a disconnect. If it does not, there's a discrepancy in the host's notion of the USB ID and the device's notion.

You should be able to prevent EMI problems with a (very) good board layout.

For a serial port, yes. Except that you must divide by 10 instead of 8 due to start and stop bits.

For USB, look at the figures people report, or make your own experiments. It's hard to predict how fast V-USB will be in a particular application.

3.9 should be OK for most hosts. The minimum voltage is defined by the USB spec (but you must take into account that the last couple of millivolts are reached slowly). The maximum is defined by the common mode voltage range of the host's input amplifier. This may be +5V (especially with older design...

2.5 V on D- is not enough. You need something around 3 V.

8 MHz is out of question. But you can use 12.8 MHz with the internal oscillator (it can be tuned up to 12. OR you can use the PLL clock doubler and run on 16.5 MHz.