avr attiny2313 failed programming

Thread Starter


Joined Jan 9, 2010
Hello, I just tried to program an avr attiny2313. I have programmed it successfully before, but now AVRSTUDIO will not identify the chip. I did the same thing as always. Is this a problem with avrstudio, or do I need to buy another chip? I appreciate any help.


Joined Jan 28, 2005
It is important that you not discard the attiny2313. There is a very strong possibility that it is perfectly fine and that all that has happened is that one of the configuration fuses has been accidentally mis-programmed.

What programming tool are you using as your AVR programmer?



Joined Jul 22, 2007
Yeah, I'm a similar problem with Tiny2313. Using AVRStudio and AVR Dragon. Tried both HV and ISP, 5 different chips, still can't get one to program. Even tried ISP with an external crystal, still fails. Can't change fuses, or program the chip. Really bugs me, since I already built and soldered the circuit, kind of useless now. Will order some new chips sometime, but will use a Tiny26 for this project, and breadboard it first.


Joined Jan 28, 2005
I have seen HV (high-voltage) programming work wonders with different AVRs including the ATTINY2313. The HV programmer that I used was the STK500 development board.



Joined Nov 28, 2008
Can't change fuses, or program the chip. Really bugs me,
Can you at least read the device ID off it?

I'm guessing you can't. Most likely you have enable the RSTDSBL fuse bit feature(so that you can use PORTA 2 as I/O pin) and then programmed the chip with internal oscillator with fast startup.

The moment you apply power, the chip is running and thus can never go into program mode. Sometimes even the Dragon HVP will not help. The STK500 HVP could be better but I don't own one.

I don't know why but my Dragon HVP fails to read the first ID byte correctly 50% of the time(getting 0xFF instead) while it reads the second and third byte perfectly. It happens on Mega48, Mega8 and Mega88 so I don't think its the chip problem. I have no problem with most other tools reading the ID bytes including one I built myself.

Some information here:

1. avoid changing the RESET pin into I/O pin

2. if one have to, make the reset pin an input other than output

3. if one has to make it an output, just wait about half a second after power up in code before doing so

4. place aside a known good MCU at hand to serve as "sanity check" of the tools hardware if something is not working as usual.