[ARC5] Solid State --- diodes...

[Brian Clarke]

Hello Ken,

The capacitor across the diode has another benefit. The reverse Voltage
capacitance is not specified and varies within and between batches. If you
put diodes in series, the Voltage distribution in reverse will be related
(proportional?) to capacitive reactance. By adding 10 nF in parallel with
each diode, you swamp the unspecified reverse capacitance and balance the
reverse Voltage drop across all diodes in the string. Using diodes in a
bridge rectifier, you are actually putting diodes in series.

The older way of balancing Voltage across diodes in a string was to use
parallel resistors - while that may reduce the turnoff spike as the applied
Voltage reverses, it actually increases power consumption, thus lowering the
overall efficiency. Parallel capacitors deal nicely with the turnoff spike
without reducing efficiency.

73 de Brian, VK2GCE.

On Friday, 9 February 2018 10:29 AM, you said:

I haven't seen the following issue mentioned yet, but I do remember that
when using Si diodes in a power supply in a receiver, there appeared
significant noise, i.e. RFI, from them.

A constant sort of hissing or static-like interference.

The way I cured this at the time was by including a 0.01 mfd bypass capactor
across each series-connected diode, and another from each end of a bridge to
ground.

At the time, I attributed this RFI to the abrupt cut-off of conduction by
the diode as compared to the slower one of a tube when the sine-wave
reversed, but that was most probably wrong.

Now this was a long time ago, and I haven't looked for such a thing lately,
but I can't see that bypassing each diode with a good disk-ceramic would
hurt anything.

Comments?

Ken W7EKB