Atmega8 Unable to Read Signature

[silent069]

Hi all, just having some trouble here with the 50 dollar robot. Ive built the circuit, and have tested throughly  to make sure nothing has been corssed by mistake and everything seems to be in order. Im currently having trouble trying to send the code to my Atmega8. Ponyprog was giving my trouble so i decided to use AVR studio 4. The problem im having is that the program is unable to read signature. And when i push "read signature" a warning message is displayed below it.  "warning: signature does not match selected device."   Ive turned the isp Frequency down to 125 as i read in another forum. I ran a test of my battery and its reading 6.5v. Both lights are on my Programmer (more expensive one)  Is there anything else I can do to rule out what may or may not be right. 
Absolutely any suggestion's would be awesome!

Thanks, Bobby

[bens]

A few standard questions:

1) Are you really using an ATmega8 (as opposed to, say, an ATmega168)?
2) Do you have your device configured as an ATmega8 in AVR Studio's AVRISP dialog (under the main tab)?
3) Are you powering your ATmega8 while you're trying to read the signature?
4) What signature are you getting when you try to read it?  Do you consistently get this signature or is the returned signature seemingly random?

- Ben

[silent069]

ok to answer ur questions.

 1.  on the chip itself it says..  atmega8-16pu
 2. I do have the device configured to the atmega8 in avr studio. (following 50$ robot tutorial)
 3. Yes i am powering the board while trying to program it. and i have checked the battery with a multimeter and its reading 6.5v
 4. as for the signature, there isnt one that is coming up. unless of course 0.0 0.0 0.0 0.0  is a signature. The message says the signature does not match selected device.

Im beginning to wonder if maybe i have a faulty chip.

[krich]

Probably the most common mistake once you've got the double green lights on the AVR ISP is inadvertently mixing up the MISO and MOSI pins.  Triple check those.  You could also check the voltage on the ISP connector between VCC and GND, but I'm guessing that's okay because of the double green lights.  The double green lights imply that your reset connection is good too (not always).

I've personally been battling some really gnarly flux residue and have had some of the same issues because of that.  The flux, in sufficient goopy quantities, can form a bridge between connections.  Take something pointy, like a thumbtack, eyeglasses screwdriver, or dental pick and clear out any goop that may be betwen your solder joints and traces.  Get a magnifying glass to see better.  If the flux is still liquefied and is not effectively being removed by this method, use the scrub method below for better results.

Instead of scraping, you can grab a toothbrush and some alcohol and scrub the solder side of your board.  Scrub gently and thoroughly.  You should probably unsocket your atMega8 before scrubbing and then let the board dry for a while before using it again.  Most household alchohol has water in it, so it won't evaporate as quickly as 100% alcohol.  The alcohol will dull the shine on your solder connections, but that's okay.  If you like them shiny, just temporarily reheat them with your soldering iron and they'll shine again.  And, just as a disclaimer for the folks with a death wish, don't use that toothbrush to brush your teeth.  In fact, mark it as containing lead with a little skull and crossbones.  K, thanks.

And finally, inspect your board for "bloated" solder connections.  The connections should look like a mountain peak, not like a ball.  If they look like balls, you've got too much solder on the connection.  Balls tend to find ways to connect to neighboring solder connections that are hard to see with the naked eye.

Don't feel bad.  My first $50 robot board didn't work at all.  Heck, even now, half the things I put together don't work at all.  The scrape and scrub method has saved a few boards and now I'm at about 90%.  I'm sure you'll have better luck.  My soldering skills really stink.   That's why I know so much about fixing it. 

[silent069]

awesome, thanks for the quick reply krich. I did notice the flux situation before i did the "smoke test" on my board. So at that point in time i went through most of the joints with a razor blade thinking maybe they could form a bridge. Ill go back through everything again tonight. Ill double check the miso and mosi connections as well, but im assuming they are correct as ive checked those connections 3 and maybe 4 times. Ill post back if im still having problems, and if not ill make sure to make a post to help others if they have problems with the $50 robot.   

Yea the cleaning part did not work either. so im really stuck on as to what is wrong.  If anyone has built this 50$ robot with the NEW 6 pin programmer avrmk2, could u post a pic of the bottom of your board so i can compare and try to rule out whats wrong. i dont wanna scrap this bot, nor do i want to start another one until this one is working.

update. Ive ordered a 5 pack of ATMEGA8 off ebay from hong kong. (hopefully they show up) was very cheap so no loss if they dont. Gonna use them to check and see if maybe i had a faulty chip or maybe i just need to start from scratch with my board.

Also the voltage from VCC to GROUND is 4.96 V so that looks to be correct....i really think i may have fried the chip. I believe i made the mistake of fre soldering a wire while the chip was still inserted.

[krich]

Could be.  I find it hard to believe that, with my lack of experience and carelessness, I haven't managed to kill any of my AVRs.  Yet somehow, with one minor soldering infraction, it's game over for your AVR.  Don't beat yourself up.  I'd put my money on a DOA rather than you killing it, if that's what it turns out to be.

I'd not even classify that as a soldering infraction, honestly, unless you took a phone call or something while leaving the soldering iron on the pin.  Those chips are soldered directly to boards all the time with no problem.

[silent069]

Ive even done a firmware upgrade on my avrispmkII which had no effect on my outcome at all. I tried making my own parallel port programmer that was listed on here in another post but even that wouldnt allow me to program my atmega8. Im really at a loss of what to do. I was gonna order a dongle serial port programmer but i dont know if its worth it to spend the extra $20 to have it if the other 2 don't work. Im gonna take a break for now on the programming and build the rest of my chassis today. Ill take some pics and upload them as well.  As i also said in a previous post. Ive ordered a 5 pack of atmega8 chips off ebay so when they come in the next week or so i can rule out a faulty chip.
If anyone has any other input that would be much appreciated.   

[silent069]

for future help to anyone with this situation im replying to this post. I managed to fix my problem! In this case it was just a faulty chip.

[Novice1]

I have the same problem, but mine was not a faulty chip. I downloaded hte latest version of AVR studio, and that did not work. I am doing the $50 robot. Please help!!!!!

[Conscripted]

Novice1

Have you followed all the advice given in this post so far? I had this problem too. For me it turned out to be the chip not seating fully in the dip holder.

Good luck.
Conscripted

[Novice1]

I know, i figured it out, thank you for your help, though, i really appreciaate it