Diode Failure

What happens when a diode partially fails.

Failing short circuit is also common, then the diode becomes a resistor.

I don't think I have ever seen a partial failure before but I think leakage current can increase outside component specifications in some circumstances if it is operated outside its limits (over temperature or over voltage possibly).

What type and model of diode are we talking about and what is the application?

Jack, You're right about diode usage outside of specs, but this is in spec. The diode is part of a bridge rectifier in an alternator. The voltage is within spec and ditto for the temperature.

The strange thing is when the battery is connected and the ignition is off, we get a 1.4A drain through the alternator. When the alternator lead is pulled, the drain goes away, but when it is reconnected, the drain is also gone. I'm thinking that a partially failed diode would consistently leak current.

The only other option is that the computer is somehow allowing current to leak through the alternator, but we can't figure out how this is happening. Any ideas?

I advise you perform a diode check of the bridge rectifier to see if you have got a diode that has failed short circuit.

They can leak, they can short, they can open circuit...

http://www.youtube.com/watch?v=xgikeXt91vM

They all "leak", there is no such thing as a perfect solid state device, hence the term "semiconductor". The definition of a shorted diode is that the leakage is higher than specified.

I have never seen a diode go open circuit, it would basically have to be vaporized for that to happen. Probably really small ones used on PCBs that can happen though, I have just never seen it.

Side note: if you are testing "hockey puck" style diodes, they must be fully clamped into the heat sinks before taking measurements across them, the clamping pressure is part of the design specs.

A leak of 1.4 A is quite a bit - not in the mA range for sure. I don't know if it's possible for a diode to leak that much without burning out.

In high school, we burned out some diodes to the point where they popped or cracked in half. We didn't know much about them, as they came in kits we bought and assembled in class.

This is possible, but would it draw 1.4A? If so, then the voltage regulator is malfunctioning, right?

Good answer! The field is in series with an alternator lamp on the dash. If the ignition is turned on, but the car is not started, the light should be lit. This should be in the MA range, not as high as 1.4A. Could there be a short in the dash wiring?

Most of my experience tells me there are only two failure modes for a diode -

Dead short - zero resistance and usually a blown fuse.

Open - no current, circuit has no voltage.

Leakage is part of a diode, agree semi-conductor, but failure always seems to be all or nothing.

Thanks Sid, I think you're correct, but I remember way back in physics class - I recall hearing of a diode being partially dead, but not completely. Of course this is physics where the improbable is analyzed.

Defective voltage regulator.

(edit: I see this was already suggested in a duplicate thread.)

I think you're right. The problem is the alt is over $600, because it needs to be a MBZ part, not a reman. I have a test set up and I'll let you guys know how it works out.

Thanks for your help.

Sounds like SCR latchup

Why are you having two blogs open for the same problem? You are just wasting our time.

Decide which one you don't need.

Moderator Notice:

This thread has been closed because it is a duplicate discussion. Please continue this discussion under Autobroker's post Strange Alternator Problem.