I have decided it was nessasary to move this topic out of the BAP bug thread due to the ongoing work involved.
Here are the pictures again:
The latest board:
img85.imageshack.us/img85/3991/revadsjboardwf6.th.jpg
The board that released the magic smoke:
img411.imageshack.us/img411/7319/rscusmalluj1.th.jpg
REV AC Overview:
That is a nice looking board! Is this going to be a new Lynxmotion product? I am sure interested in this. Can it be used with I2C or just standard TTL serial?
8-Dale
Thanks!
No, it is not going to by a new Lynx product, just a personal project I have been toying with for the past year. I was going to make one for Jim to play with and thought it would be nice to add the Lynx name on it.
It uses standard TTL, it communicates with the BAP via serial pins, the same pins that you hook up the servos to.
Mike, what power supply were you using when you reversed the polarity? If it was a 9vdc battery it may not have damaged everything. Maybe just a cap on the output of the regulator. However if you were using something with more kick, the damage could be severe. If the reverse voltage made it through the regulator, which it sounds like it did, any and everything connected to the 5vdc line is suspect. You can actually start removing semiconductors one at a time monitoring the 5vdc line for a short to ground. When the short goes away and the removed components replaced it should be functional. Sorry to hear about the smoke dude… 
Well, maybe Jim should consider something like this as a new product then.) I think it would be a great addition to the Lynxmotion product like to have a board like this, perhaps even with a dedicated MCU which could provide alternative interfacing mechanisms such as I2C, SPI, USB, etc. An 18F2550 or 18F4550 is what I would use, or a larger 18F87J50 for more I/O (including two I2C and SPI buses and USB). I’d use the QFP packaged versions of these PICs, and the 18F87J50 is only available in that package.
8-Dale
There is always a possibility that something like this could become a product but Jim is curently swamped converting a lot of the kits over to aluminum as well as adding new items here and there.
As for I2C, USB interface, onboard MCUs etc… there is just not much room on a board 3" x 2.3" if I remember the size correctly, and if you look at the photo, you can see just how dense the board is.
First step is to get the BAP taking to the SpeakJet. 
I was using a 9v wall transformer, not sure of the amps. It was an adapter that went to some toy or electronic device, I don’t remember. I usually save the adapters when throwing away the device it came with and then cut the plug off and strip the wires to use for other things. Not a very good idea, but it worked fine, except that I forgot that the white stripe on one of the wires was positive and not negative. 
After I replaced the 5v regulator, I used the ABB boards 5v regulated output, VL? I can’t remember. But the regulator on my board still gets extremely hot very quick even with the 5v supplied from the ABB. I will remove the ICs one by one tonight to see what happens.
I would go the other way, remove all the ICs, test to make sure it isn’t the regulator or some passive like a capacitor, then put them back in one at a time until you find the dead bugs.
I know Jim is very busy with the new body and kit parts. I am just suggesting as something he should look at for the future. It would be great to have a board like this for a robot. 
The MCU would provide interfacing to the speakjet, and perhaps the rest of the board. It could process sensor readings and make them available in an easy format. I could see an sensor routines for many different sensors programmed into a PIC. This would make it usuable with slower microcontrollers by doing processing on board. You might even be able to get smoother speech.
An onboard MCU could also be used with an EEPROM to store often used phrases which could be recalled and said by sending a simple code to it, different profiles that could be easily called up, etc.
Think of this as an SSC-32 for speech.
If you look at the QFP package details, they are not very big and there may be just enough room.
Of course!
8-Dale
Yeah, that’s a much better idea. I will do just that.
Either way you just need to do a process of elimination. You already replaced the regulator, so active components are the first to go in a reverse polarity problem. Also look for tantalum caps, they like to go short city with reverse polarity. They even blow up and catch fire. Fun stuff…
When I was 11, I would purposely pop caps placing them backwards on a 9v battery. Oooowwww, it was so much FUN! 
Well I removed all the ICs and the regulator still gets hot enough to light a cigarette. Well not quite that hot… 
Anyway, what should I check next?
Put an ohm meter across the 5vdc output of the regulator to groud and measure the resistance. It’s likely to be 0.1ohm! then start pulling caps off the board till it opens up. You may want to remove the regulator too, because you don’t know if it is still functional anymore.
We used to put 5vdc at ~20 amps on the 5dvc line to ahem, clear-out the short, but sometimes you can lose a trace. 
lol Ah the good ol days…
Oh and wear safty googles. 
Seriously this is a last ditch effort. Don’t do it unless you have basically given up.
Yep, .2 - .4 ohms! its a short alright!
I will start pullen caps starting with the ones closest to the regualator.
Here is the schematic for the power supply circuit that I am using.
http://img46.imageshack.us/img46/5496/psqq2.gif
I think it might be C10 since a diode check shows .7 volts when the other caps do not. I need to remove and check this out.
That sounds…er…“exciting”.
<KA-POW!>
“Short’s gone!” “Ummm…so’s about half of the PCB.”
I replaced C10 and it did not fix the problem. I am taking this board in to work with me.
No Mike the regulator is getting hot because something on the output is shorted. Your most likely cantidate is C11.
Ooooh yeeah! Well I have the board with me today at work so I will be able to get my team of experts on it.
