[ARC5] Diode Mystery

[Kenneth G. Gordon]

On 30 Nov 2020 at 12:30, Bob Groh wrote:

> 
> Regarding RFI from diodes in a power supply, that is (in my 
> experience as a design engineer) due to current spike that occurs 
> every cycle when the diode turns ON and the capacitor in the power 
> supply is wacked with the input voltage. When doing a redesign on 
> HP-1144() power supply at Heathkit, we initially failed RFI tests so 
> we dug into it and there it was - on every 1/2 cycle when the diode 
> input swung up and the input voltage exceeded the filter capacitor 
> voltage by rectifier diode drop, there would be a jump as current 
> rushed into the filter cap to replace the energy lost during the off 
> time. That current 'bump' generated quite a batch of RFI!
> 
> Hope that makes sense. 

It sure does, and I thank you very much for this info! After all these years, finally, an answer. I 
really appreciate this.

As I said, my explanation for this was only a WAG on my part. I first noticed the RFI issue 
when I built a 2 meter repeater in Missoula, Montana a long time ago.

The transformers I used were very high-quality ones I had scavanged from an airport NDB, 
having unusually low internal resistance. I figured that since an NDB had to operate 24/7/365, 
such components would make a repeater equally reliable.

I killed the RFI by bypassing all diodes with something like 0.01 mfd 1 kV disk ceramic caps, 
one across each diode, and one from each end of the bridge to ground.

Although I am not completely certaini, it does seem to me that modern diodes don't have this 
problem. However, I still watch for it, the RFI, and use the above method to clean it up.

Ken W7EKB