Predistortion to reduce IMD, etc.

[John K5PRO]

Pat,

Adding regulation and possibly ripple cancellation to the bottom side of a HV power supply is done in our big DC proton accelerator at workplace. The power supply is something like 690 kV, using a Cockroft-Walton n x voltage multiplier running around 8 KHz. On the bottom, there is a 60 kV bouncer power supply, that does just that. Feedback is taken from the top and drives the bouncer.

[K1JJ]

Interesting circuit, Pat.  I wonder if it could be done with MOSFETS too.


BTW, my old class E PDM rig had a feed-forward input on the PDM generator board. By sampling some HV power supply ripple, it reduced the hum to near zero. There is a pot that let me zero it in.  The HV supply was about 135 V.  It worked very well.


To adjust it I set the scope to high sensitivity and looked at the enlarged carrier edge. The small saw tooth could be nulled to nothing. I could hear the difference on the AM detector monitor.

I don't think many of the class E guys use this FF port now, but it does work.

Second thought, I'm using this same PDM gen board for my 4D32 tube PDM rig. I could sample some HV PS ripple and feed it into the FF port.


T

[K1JJ]

Rod, KQ6F requested a schematic of the feed forward hum reduction circuit mentioned above.

Here's WA1QIX's PDM generator schematic in case someone else needs it too:

http://www.classeradio.com/pwm_generator_rev_f.pdf


The FF input is at the bottom center.

It notes a 47K resistor between the 135 VDC PS and input, but for a higher voltage supply, take extreme HV sampling precautions to prevent problems.

T