We had a camera on our robot burn up. this camera was connected to the main power source (through a spike not directly of course) and we had just connected it. It ran fine for several hours but the robot was suspended and couldn’t move. Later when we ran it on the ground the wire burned through the rubber and the camera emitted a foul smelling cloud of smoke. :eek: I’m almost positive there was a spike in current caused by our motors when the robot turned (we really need a better drive train and the motors literally fight each other). My only worry is how to protect the camera. I’m thinking a resistor or a DC regulator. Also I’m considering using a capacitor to provide power to the camera when the motors cause a power dip, and I’m wondering if this fits under the “no other power source rule”. Any help or advice would be great meanwhile ill shop for another camera. 
I know that the CMUCam provided by FIRST this year hooked up to power via a PWM connection on the RC. This seemed to work very well, except for when the backup battery died and you spent forever trying to figure out why the camera stopped working.
Well, if you really had your camera connected to a Spike, then that’s your problem. Spikes will output the battery voltage, which should be around 12 V, and the camera was supposed to be connected to a PWM header, which should output about 7 V. It’s actually quite impressive that a sensitive device like the camera lasted such a long time running at a voltage 70% bigger than its rated voltage.
no this was not the same camera it was a 12 volt DC camera not part of the kit
If it was a spike, is it possible that you reversed the polarity on it? Otherwise it sounds like something just shorted out.
There are two indications here that need to be checked. The first is obvious, are you sure that the input you were using was intended for 12 volts? The camera may be only able to take 9 volts on one input while taking 12 on another connector. If you changed batteries just before the camera burned, the higher voltage that is sometimes present on a newly charged battery may have exceeded the camera max input voltage.
The second is are you sure that there is no electrical connection to chassis? The camera frame is connected to power common in most cases. A motor input tied to the frame and set into reverse would force 12 volts into the ground of the camera and provide a return path through the camera and it’s power wire. It is possible you didn’t fry the camera but may have burned up the power connector and the wiring it is connected to both inside and outside the camera.
I think you can just put a capacitor on the power leads for the camera. That’s a strategy most circuit boards use to regulate power to sensitive chips. I don’t know what size you’d need - I’d imagine a rather large one… but this should accomplish what you want. If I am totally crazy (I’m an ME, not EE - anything is possible) and you can’t use a capacitor, I am fairly confident there are voltage regulators available on Digikey.com .
<edit>
This isn’t quite what I was talking about… but here is a voltage regulator on Digikey: http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=10626&Row=449095&Site=US
Hope this helps!
</edit>
A filter of just about any kind would do good things for sensitive chips.
If it was inductive kick, how long was the camera working? Inductive kick typically takes stuff out fast. If you’re protecting against that, you’d want an inductor.
Check the datasheets, so you know the limits of the camera, and check to make sure, as Al suggested, that the camera is really as dead as you think it is.
My advice? Just double and triple check everything, and if it all checks out, run tests with voltmeters and oscilloscopes. Excessive? maybe. If it keeps you from losing some expensive hardware, it’s worth it.
Sparks
I read in another post that this was a 12V camera and not the CMU2cam.
Since this is a sensor (albeit a sophisticated one) the camera should be connected to the fuse panel via a circuit breaker (20A) and not to a spike. The spike output can go from +12 to -12V and negative excursions would fry the camera.
The 12V battery should mitigate any inductive spikes and the only thing you have to worry about is undervoltage dips when large loads are present on the 12V bus. A UV dip should cause a solid state device to stop working temporarily but should not cause a catastrophic failure.
Pay very careful attention to the specs on your device. If a solid state device lists 12V as a maximum voltage, it will likely be destroyed by 12.1V!
Most 12V (nominal) devices will have a maximum voltage of 16 to 18V.
You may want to post the part number of the component and let one of our resident experts give it a look over.
At a very minimum, loose the Spike and wire directly to the CB on the fuse panel.
Mike
the camera is a http://www.supercircuits.com/index.asp?PageAction=VIEWPROD&ProdID=4213 12 volt color camera. It was ment to run at 12 volts. the next one will have a fuse so i wont have to worry. also if just the power connecter is fried it doesnt matter since its all integrated into the camera.
Ok,
Things are getting clearer now. What was the camera connected to on the robot? Did it go to a cable that feeds a video recorder or did it feed a transmitter? As I stated before, sneak paths exist in many of these devices and the shield of the output video is tied to power common on the camera. Back feed power through the video cable and most everything on the board is going to fry. The smell you mentioned in your first post may have been the main filter capacitor blowing up. The smell of burning is nothing to the release of boiling electrolyte steam when power is applied to an electrolytic capacitor in reverse. It is possible that you somehow got the Spike polarity reversed so as stated before using the Spike is/was a bad idea. Use a breaker on the panel with nothing active in between for power source. When connecting the video output to another device be sure it is run from the same power or is insulated from the robot frame.
no the camera itself is fine we had seperat power running to the transmitter and camera. the smell and smoke came from the power wire melting yet im hesitant to say it was a short becuase we were getting transmission throughout it. the only damage to the camera is the power port but thats enogh becuase its integrated into the camera.