Defective 3.3v regulator...

QUARKSPIN

A while back, I started having a mysterious problem with my hydra board. To make a long story short, It turned out to be the 3.3v power regulator, and I went through an incredible ordeal to replace it. (I ended up learning how to repair scratched PCB traces). I managed to fix it however, and lived happily ever after for about 2 years. However, today I noticed a strange communication problem with my PPDB (Propeller Professional Development Board). I reached over and touched the 3.3v power regulator just to make sure... And sure enough, I burnt my finger. My wall wart was also quite hot to the touch, so I hurriedly unplugged it. Alas, I was too late, and the power regulator was hopelessly fried. I know from experience that the prop will be OK, and the only problem is a failed regulator. However, I dread the thought of having to do the gymnastics reqired to replace it, does anybody have any insight as how to fix it without doing any desodering? (The heatsink on the power regulator makes this next to impossible).

Duane Degn

I had a similar problem on a Proto board. I ended up pulling some of the traces away trying to get the old regulator off.

My guess is something like this would benefit from a hot air rework station but I'd also like to hear other's suggestions.

Duane

Leon

ChipQuik should enable you to remove the device without any damage:

http://www.chipquikinc.com/

william chan

From my experience with regulators over many years, they don't become faulty unless the input voltage is too high. ( more than 35v )
When they get hot, it is usually due to a short somewhere on the regulated line.
Moreover, a regulator's thermal shutdown circuitry will protect it from being fried.

Regulators are really very robust.

LoopyByteloose

Actually, running a voltage backwards through the regulator will often kill one very quickly. So if you were to run more than 3.3 volts on that rail, it might fail.

I admit I am not sure how one might do this unintentionally.

Maybe a crossover with the +5 volts, then shutting down. Maybe using a battery supply in conjunction with the normal supply. But the standard way for causing failure is to use rather large and unnecessary filter caps on the regulated side and then shut down. The shut down causes the cap to create a backflow through the regulator. The solution is a bypass diode in the reverse of normal current flow to safely discharge the cap around the regulator.

I do see a rather large cap that is supposed to filter servo motors on the Propeller Proto Board and have wondered if I should insert a protective diode when using hobby servos. I does seem wise.

QUARKSPIN

I'm pretty sure this one was defective, I was using a 9v wall wart and didn't have anything other than the prop conected to power.



EDIT: Wow, I feel so stupid now. I just looked at my supply and it says "Ouput: AC"!

Mike Green

It depends on the regulator. Some, for automotive use, are designed to survive being connected backwards across the power source (and protect the connected equipment). Others will instantly "give up their magic smoke" when connected to an AC source or connected in reverse.

The one used in the Hydra is not one of these automotive types.

QUARKSPIN

Wow... I decided to hook it up to a DC power supply, (To see if it hat really failed) and one of the caps exploded! No really! It glowed bright orange, and emitted tons of white smoke.


On the bright side, I just came up with an idea for how I can save it. I happen to have quite a few spares, do you think I could supply the power from an external supply? I'm thinking about wiring up the spare 3.3v supply on my breadboard, and having it output to the 3.3v rail on the breadboard. However, after the recent fireworks display, I thought I'd ask someone first. (Oh, BTW, I'm using the Propeller Professional Development Board revA.)

Oh, and I have some pics if anyone wants to see them.

Leon

Sounds like a tantalum has objected to the mistreatment; they often explode with a loud bang! Just replace the regulator.

QUARKSPIN

Ok, I have a possible fix set up. The extra 3.3v supply is set up to draw power from the 5v rail, and output power to the 3.3v rail. Should I power it up or not?


EDIT: I'm tired of waiting... Trying it now.


EDIT: Didn't work. Blew another cap. Trying something else...


EDIT: Didn't work ether... Finding a cheap aurdino...

EDIT: Lol, JK. You know I wouldn't go and do something dumb like that. I guess I have to buy another...

smbaker

QUARKSPIN wrote: »

EDIT: Wow, I feel so stupid now. I just looked at my supply and it says "Ouput: AC"!

Make sure to double-check the voltage even on DC supplies. I used a 10V rated tantalum capacitor as an input filter to a regulator on one of my projects, and connected it to a 7.5 VDC unregulated wall wart. The cap instantly became very hot, no voltage output from my regulator. Upon checking, I found that my 7.5 VDC wall wart was in fact putting out 11.5 VDC. Even after switching to an appropriate supply, the damaged 10V tantalum capacitor was now acting as a direct short (must be part of the failure mode of tantalums).

Even some 6 VDC unregulated wall warts that I have on hand put out > 10 VDC.

QUARKSPIN wrote:

EDIT: Didn't work. Blew another cap. Trying something else...

A current-limiting benchtop power supply is good for things like this. I know several people who usually test new circuits by always starting out with a low current and slowly ramping it up, to check for shorts.

Also beware that not all three-terminal regulators have the same pinout. I seem to recall one time long ago getting bit by an oddball regulator.

As far as removing components, a cheap hot air iron works great for such things.