A diode question

[Capt Slog]

These are glass, and I think they must be germanium, as they have a forward voltage of typically 0.23volts.  I conected one to a power supply and wound up the current untill it failed, this happened at around 2amps.

I was wondering if I could make a rectifier from them by using several diodes in parallel for each element of the usual layout for a three phase setup.  Ignoring issues such as space or cooling it seems feasible and would mean that I lose less volts in the rectifier compared with the stud diodes (silicon) that I'm using now.  Or am I missing something important?

[Flux]

Yes you could. It would be crazy and a complete waste of time and effort.

These are almost certainly signal diodes.

The volt drop thing is not as big an issue as often claimed unless you have some decent way of matching the load. If you are really worried then use schottkys. At one time I had lots of Power germanium diodes capable of 50A. They were fine but made no significant difference to output. Those little things will not share current and will give endless trouble.

It's always tempting to use things that are free, but it usually turns out rather expensive in the end with wasted time.

Flux

[Capt Slog]

You say they won't share current?

I naively thought that if I had two in parallel then it would handle 4A, and 3=6A etc.  Is this not the case?

[Flux]

If they are all from the same batch then 2 or 3 in parallel may share fairly well.

The sharing gets worse as the number in parallel increases unless you include sharing resistors.

If these blow up at 2A then they are probably about 500mA devices, you will need dozens of them in parallel to carry any useful current. Far better to build crystal sets with them.

Flux

[finnsawyer]

The forward current in a diode increases exponentially with the forward diode voltage.  What this means is that a slight mismatch in the forward current - voltage relationship between diodes can result in a large difference in the forward current of the diodes when connected in parallel.  For the scheme to work all the diodes must have identical I-V characteristics.  Not a likely occurrence.  One can force the deal to work by placing resistors in series with each diode before connecting them in parallel.  That would be a waste of power.