0022 works but not v1.0

I have many boards exactly the same. Somehow I must have damaged one. They are Duemilanove boards. 1 in particular only Uploads from 0022, not v1.0 IDE. The 2 green lights never blink. All the other boards are fine. How is this possible? What did I do to it?

It doesn't matter to me as it's still perfectly functional. I'm just curious how it can be.

Strange as both use Avrdude.exe for uploading, although there was recently a thread mentioning a difference between the versions of the avrdude. (numbers that pop up are 5.4 and 5.11 ??) You might need to upgrade it? (backup-copy-try...)

(you probably allready did) check your settings under the tool menu?

Both 22 and v1 are the 2nd board that I can't spell starts with D ;) I know the serial port is correct because it does blink briefly, only once. Why would I have to upgrade Avrdude for 1 board, but not the others? I did just install AVR studio 4. Unrelated?

I did just install AVR studio 4. Unrelated?

No experience with AVR studio , do not recall threads discussing this, sorry

Maybe the LED's are bad! http://www.ledlightingwholesale.com

[Moderator - this looks like spam - but I'll let it go. I personally had an LED fail in a matrix recently, which then caused weird behavior that had us chasing what appeared to be a short in the wiring. Replaced the LED as it wouldn't light up - and the short went away.]

I had the same occur recently with a standalone chip. How to fix it:

Need to bootload the board with Optiboot bootloader.

When downloading sketches: select Uno as board type.

That will ensure the chip's bootloader and the PC's IDE are both using the same baudrate.

I don't understand the problem you are fixing. Keep in mind it works with 0022, so yes, the LED's light. You're suggesting a different bootloader?

sbright33: I don't understand the problem you are fixing. Keep in mind it works with 0022, so yes, the LED's light. You're suggesting a different bootloader?

The board that fails only fails for one release of the IDE, and most of the work takes place on the board via the bootloader.

So, it follows that is is likely something to do with that. You can try bumping the verbosity of the upload process via the prefs and comparing the output from the board that works and the one that does not.

The usual reason for failure is mismatch between IDE serial speed and bootloader serial speed.

Downloading Optiboot and selecting Uno ensures they will agree.

Even though the hardware is designed to match that big D word that means 2009 bootloader? Why didn’t they change loaders in the first place? There must be some advantage to that bootloader.

A big advantage I can see comparing the hardware to Uno, is power usage.

Later bootloaders are smaller (fewer bytes, little more room for sketches) and I believe spend less time waiting for a start up message from the IDE (so faster sketch startup after reset is released). The uC itself doesn't care what is running.

Duemilanove vs Uno power usage - probably pretty close: Duem. has '328, 5V regulator, FT232 (supplying 3.3V also) for USB/Serial, and power LED. Uno has '328, 5V regulator, 3.3V regulator, ATMega8U2 (or 16U2) for USB/Serial, and power LED. In both cases, only the '328 can go into sleep mode.