I’m pretty sure the answer is no, but theoretically, if I wanted to replace the ATMega8 on one of the first-run Orangutans with an ATMega168, would I need to (or should I) make any other modifications besides swapping the chips themselves? I’ve already added the pull-down resistor to that LCD pin.
The devices should be interchangeable without any other modifications.
Well, I finally remembered to throw in an ATMega168 with a Digikey order, and I tried the swap today (mostly as an exercise in surface-mount soldering, I realized very quickly that the results probably wouldn’t be worth an hour of connection-checking). It was a little iffy, since I can check for shorts, but not really for good contact.
At first I was able to verify the device, and program and read back hex files form the chip, but I couldn’t get any buzzer, led, or lcd response. I tried touching up all the connections again, and got some really weird behavior. The first line of the LCD would flicker, and the red LED would come on slightly when I pressed on the chip. The weirdest part is that the power switch setting didn’t seem to have any affect on the power state of the board!
I tried touching up the contacts one last time and now I get no response out of the chip whatsoever. I think I may have toasted it with hot air to begin with (I was coming on petty heavy from the top). I might try again with just a soldering iron next time.
That does sound strange. Were you able to look at the joints under a microscope?
If the power switch wasn’t having any effect, I’d suspect that the VBAT line got shorted to some microcontroller line. However, that line doesn’t really get too close to anything else; the closest thing is the motor driver chip. Did that get moved? If VBAT did get shorted to some other line, it probably wouldn’t be good for things in general, so I don’t know how recoverable the board would be. If you would like, you can send it back for us to examine and maybe repair.
Thanks, I think I will let you guys take a look at it. If it was for sure just the AVR, I would try again, but I’m worried about how on earth VBAT could be shorted to something on the chip, and I don’t want to just keep frying them!
I (finally!) looked at the unit, and it looked like the soldering wasn’t that good (the chip was slightly rotated, and all of the pins on one side looked like they might be soldered together), so I tried reworking the chip but still couldn’t connect to it with the programmer. I put on a new mega168, and that seems to have fixed things. I left it with the 8 MHz internal oscillator since our standard test uses that clock speed, and everything seems to be working! We’ll be sending it back today.
As I was testing the second microcontroller, I realized that my programmer clock speed was set high enough that the 8 MHz was too slow. So, it’s possible that the original unit was okay and not running at 20 MHz the way I assumed. It sounds like it’s been through a lot of abuse, anyway.
Sweet! Frankenstein’s Orangutan Lives!
And yes, abuse is probably a kind word for what I put that chip through. No more MLF soldering for me, I think I’m going to stick to packages where the pins actually stick out from the sides for a while.