46 replies to this topic

[tiger99]

I wonder how long it lasted after that? The diodes would have been horribly degraded at best. It may be that the winding resistance is somewhat higher, and I guess it was not a Lucas, so the diodes may have been more robust. There are two factors at work, the actual short term peak current rating of the diode (unlikely to be reached) and the long term temperature rise, The junction generally has a thermal time constant of a few milliseconds, probably a few tens of milliseconds in an alternator, which relates to how quickly it will reach self-destruction at about 230 deg C, the temperature at which the aluminium bond wires will start to dissolve in the silicon. That often leads to short circuit. The bond wires will themselves fuse at some value of current, which usually will happen quicker, making the diode go open circuit. But then there is the effect of localsied hot spots in the junction due to non-uniform current distribution, which can form a localsied leak path. And, there are a number of other more obscure failure modes too....

 

At work we occasionally have a component failure in a metal can hybrid, due to someone doing something daft. We always get the lid removed and do a full examination to see what failed and why. So I have seen some failed diodes, transistors and bond wires, too many actually, and invariably caused by current of about 10 times the designed value. The good news is that an alternator will never generate excess current itself, due to leakage inductance, so even if you get a serious battery short, a major fire hazard of course, often due to the main battery cable touching the hot exhaust, the alternator will be perfectly ok, even if the engine was running at high revs. The only vulnerability is reverse polarity due to reversed battery or allegedly welding, but I dispute that welding is a cause of blown alternators, as it would be a very rusty shell indeed that caused enough voltage drop in the right place when welding to get an abnormally high or reversed voltage to the alternator. If using TIG with HF start, it might be different....

[Captain Mainwaring]

It lasted 4 years until it was stolen by gentlemen who like piebald horses.

I'd say that the current limiting factor in most cases would be the battery --> alternator cable.

I'd agree that it's unlikely that welding would cook an alternator, but more likely that a bad earth return or HF welding would make an ECU think it was a pop up toaster. What may kill an alternator when welding wouldn't necessarily be current - but the high starting volts, especially with an AC set. With a DC set (not that many would be using a DC stick set for carbody work) would be the fact that the stick is negative and the body positive in most cases.

 

From the work that I do, most "victims" of welding tend to be load cells.