Another Class E project by KA2QFX

[KA2QFX]

Although only the power supply is well under way as yet, I thought it time to post something to the board for those who might be interested in such things.  You'll see I tend to do things a little different.

http://home.comcast.net/~msed01/eps.html

[WA3VJB]

Please think about taking photos and outlining a text that could become an article in QST and possibly some future edition of the ARRL Handbook.

Several years ago when QST's editor became receptive, I had asked QIX to write one.

Don't know whatever happened, but their offer is still open.

[K1JJ]

Mark,

I like the regulation circuit for the HV. Eliminates the pri and sec step start.  Cool.

T

[WA1GFZ]

Might consider a couple big beads on that SCR lead. They make a fair amount of HF noise switching off zero cross, just like a lamp dimmer.

my 160 meter modulator is closed loop current mode regulater that I shut down  in RX.

[KA2QFX]

Yes, Frank, I agree. There will likely be a liberal dose of ferrite and bypass caps sprinkled into this by the EMI Fairy before it's done.  The final schematic is a work in progress.  I always have the option of keying off the SCR enable during receive. Although, with low bleed-off, it rarely fires but one cycle every few minutes. 
Initially I've been testing this in the wide open gain mode where it pretty much turns on quietly as voltage rises across the SCR.  If I add feedback to the voltage sense amp I can get it to fire at later angles for better resolution, but that, as you know, can get noisy. 
I have a lot of full load testing of this full-scale application before I likely settle on some modicum of feedback in the regulation.  I'll post the results promptly. Any suggestions are welcome.

[WA1GFZ]

Mark,
I have worked on some large SCR switches running at line frequency and have chased the noise monster a number of times. The large inductor will help but you might consider a HF inductor in series if there is a lot of noise. Also we found SCRs susceptable to RF so a common mode choke we used a type 77 bead cathode - gate helps.  Your feedback circuit should work fine. One thing to consider. Set up your feedback to always fire on the zero cross and skip beats to regulate. This will increase the ripple a bit but will be a lot quieter but more complicated.

[KA2QFX]

Frank,
   I initially started with a zero crossing trigger circuit I employed with triac systems in a past life. But they were on the AC side of the load. 
   I found here, when I did my initial testing, that once a DC level was established at the cathode the turn-on pulse was lost since the SCR was reverse biased at 0 crossing.  So I went to the simpler anode-gate resistor method which turns on very quietly, just like a diode.
   The next step, will be to apply some gate C, and more feedback in the loop amp to delay turn-on even later, in inverse proportion to the voltage droop at output.  This WILL be noisier and may gain me very little additional resolution, but it’s worth the exercise to gain the data. IMHO
    Type 77 huh? I would have used some of the ferrite more typically sold for EMI applications which I thought was higher u.  I have both and will see what I can determine.  I have a bunch of those little tiny FBs for component leads but have no idea what their u might be. Guess I’ll have to put a few turns through one and do some testing. 
   See you at the BBQ?

Mark

[WA1GFZ]

Sorry can't come up before Friday.
It will be interesting to see what you come up with. SCRs are a pita at 400Hz but easier at lower frequencies. The gate delay causes a step at turn on.
I looked at rectified DC as 120 Hz AC referenced to ground.