The AM Forum

I had a nice discussion about constructing Mod reactors by taking two like filter chokes & removing the "I" iron & putting them together with a spacer to multiply the inductance 3-4 times what the chokes would be if they are just connected in series. The discussion came around to wondering if anyone has experimented with secondary windings on HV transformers? I seem to remember a lost thread on this last year and was wondering if anyone had tried this? We figured the inductance would collapse as soon as the DC started flowing unless the Iron was removed & rebuilt as a choke setup with an air gap. Anybody had any success with this? It seems that coming up with 40-50 H that can handle some current is a very rare animal indeed. If I build...... What do I mean "if", when I get the bug to build anither GlassFet rig this part seems to need as much attention as the mod iron.
Keith
WA1HZK

I wonder if it's possible to cancel the core saturation or mag field created by the HV by using the other winding with dc applied to it in the opposite direction?

Just flow the DC through the primary at opposite Phase? I think voltage breakdown would be a problem due to primary 220 VAC insulation. Possibly run a fixed low voltage DC to "tune" the choke. Anybody got a rig apart and want to try this theory. If the rig explodes contact W1RKW.
Hi

Don't jump off a cliff if I tell you to do it...  ;-)

Heat would probably be an issue with the core with DC on both sides rendering the XFMR totally useless at some point.

No reason why you cannot series a bunch of filter chokes.

Have been doing it for years and works great. Am using a 7200 volt 5 kva pole pig (use the 7200 side leave the 220 open NC) . It gives me about +/_ 90 henrys. W8tph has used other transformers with around 6000 volt secondaries with good results.The thing is use BIG iron and at any normal power levels saturation will not be noticed. That's what is in my avitar transmitter ( pair of 4-1000s modulated with a pair of triodes 5868s).I know I will get a lot of flack why this won't work. Save it guys it will just go in one ear and out the other. I'm very happy  :-P  :razz:  :mrgreen:  :cool:

Roy,The trick is very big iron running at low power to stay away from saturation. Sounds like a good idea to me.
Anytime you take iron out of a transformer the L dives. the more iron you take out the higher the current it takes to saturate it. Then there is the series R to worry about.  Not in Roy's case though....

Run Dc in the reverse direction and you get no inductance because you cancel the field.

It is a good way to get better bass response for expence of odditional power consumption. In such case it should have couple of windings.
Or, if it is a modulation transformer, polarity of connecting primary and secondary windings does matter.

Yes if you bias the core with a little DC, and the BH curve is linear through zero. The DC has to be very clean or crud will end up on the signal.

I recall reading in a 1920's-early 1930's (QST?) an article about a regulated HV power supply.  It used a saturable reactor with two windings.  One winding was in series with the mains voltage, and the other was connected between the power supply and load.  More DC through the winding to the load saturated the core, thus reducing the inductance of the winding in series  with the  mains voltage, which boosted the voltage to the HV transformer primary.  With careful adjustment, it was claimed that the voltage would stay constant regardless of load, and it could even be adjusted to produce an INCREASE in output voltage under load.  Just the thing for a class-B modulator plate supply.  I recall that somehow the winding in series with the HVDC was split arranged to cancel the a.c. voltage that would otherwise modulate the DC from the power supply and generate hum.

I suspect that the reason those things never caught on was that even though the mechanical DC meters indicated steady voltage under varying load, the instantaneous dynamic voltage was very poor.  The same situation exists for choke input filters using a swinging choke.  The only way to maintain good dynamic regulation is to use a large HV output filter capacitor - on the order of 25 to 100 mfd.

Back in those days, HV filter capacitance was VERY expensive.  Most HV supplies used no more than 1 or 2 mfd of capacitance.  Instruments that  responded instantly to changes in voltage, such as oscillographs, were also very expensive and hard to come by.  Hams who built transmitters using typcially available components and a few mechanical movement meters, wondered in vain why, although by all indications they should have a good solidly performing system, the audio sounded like crap.

In 1960'th sweep generators used RF coils inside of powerful AF electromagnet.
Also, so called "Magnet Amplifiers" were widely used in Soviet Union. There were also magnetic resonant voltage stabilizers.

Once when I was a student, we had a rock band, and come once to the villadge at the far North of the Tomsk region. They in the very deep Siberian forest did not have electricity, they had own diesel powered electric generator. When we switched on our appliences, our electronic organ produced deep frequency modulated sounds. My hair stood up when I realized that lamps around give stable light!!!

But very soon I understood that it was my 2 KW stabilizer guilty! The voltage produced by a diesel generator was stable! But frequency was not! So, before my stabilizer I had clean 220V AC, but after it I had floating voltage... :lol:

Interesting stuff!  I have been using a plate transformer as a push-pull heissing reactor in my 4-400 rig for years (see avatar icon to the left).  Steve (KA1SI) suggested it when I was scrounging around for a suitable mod transformer.  I did not mess with the core laminations at all, but just connected the B+ to the center tap of the secondary, with the plates of the 833's connected to the 2200 volt taps and then coupled one end of the secondary at the 3000 volt tap through a 4uf 3kv oil filled cap to the 4-400 final's plate while the other end of the secondary just floats.  The B+ to the 4-400 was fed through a 10 henry choke.  It produces tons of audio that has a solid low end and is reasonably flat to around 10k.  Using a variac on the plate supply and a variable regulated bias supply to set the resting current on the 833's the rig will run from 75 watts to just under a kilowatt. It has proven to be a very durable setup and is still running fine now after about 20 years of use.

Rob WB1AEX

Switching regulators with multiple outputs plus and minus use coupled inductors in their outputs to help regulation and reduce the core size
by providing bias to the core in both directions.

Tolly, Sweepers made in the 50s and 60 here had an AC electromagnet
to drive a variable cap to sweep the oscillator. It is so much easier today.
It took real talent to design a color TV with a hand full of tubes.

I remember waaaay back (1980s ?) hearing the local powerhouse AM (VK3AML) talking about applying a bias current to his modulation transformer to reduce DC saturation. I remeber though that he had to put a number of hefty chokes in series with the 'bias' winding , other wise it just acted like a shorted winding. - I presume he would have had a well filtered DC supply   for the bias  that looked low impedance to Audio.

  I have no idea if the thory actually translated into practice  to extend low freqency response.
                                                                      Ian VK3KRI

Ian, there are high voltage FET's now available, for example used in horizontal sweep of TV's. They are excellent current sources, if to connect a drain to the winding, and to apply regulated bias to it's gate, and insert a resistor between source and minus of the "compensating" voltage source.

Current source means extremely high AC impedance while the current corresponds to gate bias minus gate to source voltage obtained from cutves for appropriate current (usually 2-5 volts) divided by resistance of the resistor between source and ground. However, such compensation consumes power, but as the result you may get the best bass response available from the iron.

DC balance tricks - I think if you used three tubes in push-pull instead of two, you could unbalance the primary current correctly, and get more positive peak drive, with the appropriate primary and secondary connections.  Bias the two-tube side only for unusually high resting current if the audio tubes can take it, and it would help the DC balance.   You could bias the one-tube side to cutoff then, or beyond cutoff, effectively moving crossover away from center - there is no reason it has to be centered.  Lots of wasted power with that much resting current, though.

Or you could go all the way to a class A modulator on one side of a split winding, and the RF final on the other side.  With the right bias, that ought to give DC balance, but for high power you'll need a 3-1000 modulator or something, and it will run hot...  More turns on the RF PA side would offer more modulation percentage, but then full DC balance would require more modulator current.

But heck, for such low efficiency, just run AM linear.  Much simpler, and clean, plus you can run Swingle Swidebwand if you want, too!  Oooph Fhoophh  Phwap-fooof...

Hey, has anybody tried a microwave oven transformer as a filter inductor?  I think the magnetrons in these ovens are operated in half-wave mode.  If so, the power transformer must be designed for some DC unbalance.

We need to get more Big Rigs under construction. You know you want it. Use the force Luke, turn on that drill press!

Ha! Swingle Swidebwand. Reminds me of a recording I have of you in 1969 telling some sidebander to get back up above 3900 before you kick him. Great stuff.





Interesting idea. I know the power supplies in those things are but one step above raw AC. Ever listen to the the microwave signal from an oven using a receiver? It sounds like a buzzsaw  - lots of AC ripple and drifting all over the place.

Microwave transformers have a magnetic short-cut, so they are acting as a transformer with a series choke. Rectifiers are half - wave, so together with "choke" rectifying high voltage they generate high level of 60 Hz noise with a lot of harmonics. They usually have one side of a secondary high voltage winding grounded, so it's problematic to use them instead of just ordinary chokes.

Ha ha... yes, that was funny!  Of course, I grew out of that phase long ago.  heh heh

I remember when I first heard SSB, on a short wave receiver, way back when... I said to my father, "Listen to this ignoramus!  Can you believe he's talking that way?"  He showed me what a BFO could do to the signal. "Oh," I said.  But I learned how to imitate that kind of speech sound, and it has amused me greatly for all these years!  It's also funny to imitate VOX, and kids on Christmas with their first walkie-talkies, learning how to push the PTT button.

VOX: "M"  "Unning"  "Ox."  "Ows it"  "Ound?"  (Translation: "I'm running VOX.  How's it sound?")

Kids on CB: "Ey jo"  "Ud ya s?"  "Ud ya s?" "An't hear y!"  "Ud ya s?"  (Translation: "Hey Joey!"  "What'd you say?"  "What'd you say?" "Can't hear you!"  "What'd you say?")

One night back around 1970, I had Bill (W3DUQ) tune a few KHz down the band where some sidebanders were eeeebie-geebie-ing away... I started going "EEEEBIE EEEEBBEEE-GEEBEEE" into the mike, and Bill recorded it in sideband mode 2 KHz away... It sure sounded wierd!

I remember back in the late 60s, Tim 'HLR had a VM4 rig, and he used to turn off the PA plate supply and go "Huwwo fwee kweww, fwee kweww"

(edit: fwee kweww, not cwee kweww)

Su quee, su quee ...... hwullo su quee.....

Heh heh - that's inverted audio... you can decode that by listening on SSB LSB on the high side or USB on the low side.

Magnetics and magnetic coupling are interesting subjects.  I am guessing that magnetic shunts produce a current limiting effect in these transformers, and possibly a voltage limiting effect as well.  It certainly makes them tricky to use.

I'm not exactly sure about the physics of the shunts but if I were to take a guess they must have something to do with flux sharing between primary and secondary windings.  There's probably a flux cancellation effect going on in the core thus causing voltage and current limiting.

I helped a guy design a C.V. transformer once. The Shunt and resonant winding with cap had a big effect ot voltage regulation dynamic range.
Then we had to make it work at 50 Hz. Very interesting stuff.

Not exactly like a current limiting, just as a higher output impedance, like a transformer with a choke in series already connected. One full wave rectifier may be made from 2 microwave ovens. I suppose, they will supply with half of a kilowatt of a plate voltage. :lol:

Hmm, really high leakage inductance.  Or maybe a choke input?  It might give poor regulation with a varying load, but the idea of a cheap high voltage transformer is appealing.  I could use two, and put one primary winding on one side of 220VAC, and the other winding on the other side.  If I can't make the amplifier work, I could at least cook two TV dinners!

Hmm. Thats interesting. I have a 250Watt 50 V DC supply here (good for testing MOSFET amps!) that has some sort of regulation that is frequency sensitive. It certainy seems to regulate pretty well for no active components, but Ive always wondered if it should draw as much current as iit does with no load. Is that normal for these CV transformer thingies

                                    Ian VK3KRI

I have a big lorain 48 volt supply on  my test set up for the 75 meter final.
They all want to see a minimum load to stay in regulation. I'm sure your unit has an internal load resistor. A bigger cap will also help regulation. fc

Hey Frank. What's the model number of that Lorain supply ?

Lorain Products is now Emerson Network Power.
My employer !

http://www.angelfire.com/blues/wd8bil/WORK.html

I picked up a mambo-jambo regulated DC supply at Dayton last year 0-175V at 20 amps (!) - takes 220V input and uses a saturable reactor for regulation.  THing I don't like about this unit (other than its 100+ pound weight) is the talkback from the reactor.  I have no skizmatic and the voltage jumped all around... found that putting a load between the remote sense input and ground stabilized it somewhat.

Would make a superb class E supply, which is what I bought it for.  It's even fairly compact, only 6" high rackmount.

Anyone know why the reactor and/or xformer talks back so much?  There's a FB 2 KHz squeal that's pretty annoying.

73 John

Bud,
I will have to check at home but would like a schematic.
It was used in a phone system C.O. to charge batts. 48V 30 A.
Very well built! I use it as a test supply for the class E 75 meter final.
I've had the final making 1200 watts carrier so DC power was up around 1400 watts. The output is fairly stable in CW mode.  fc