Transformer as Heising Choke

[W1RKW]

Has anyone ever used a transformer as a Heising choke?  I've never actually sat down a measure the inductance of the transformer for this purpose but was wondering if it would work?

[k4kyv]

The greatest problem is that the transformer core is almost certain to have no gap, so the inductance measured with no current will be many times what it will be with the full DC through the winding.  With the ungapped core, it wouldn't take much DC on the winding to saturate the iron.

[KL7OF]

What about using the primary or secondary of a modulation xfmr as a heising choke?.......mod xfmrs are wound on a core with a gap....maybe a good use for mod iron with a bad winding..use the remaining good winding as a heising choke..

[k4kyv]

What about using the primary or secondary of a modulation xfmr as a heising choke?.......mod xfmrs are wound on a core with a gap....maybe a good use for mod iron with a bad winding..use the remaining good winding as a heising choke..

That's fine if the winding is open, and not shorted.

Another possibility is to series up the primary and secondary windings in a good transformer (in the proper phase so that the windings are additive) so that the whole thing serves as a choke.  This usually makes a high-inductance reactor, but it may not take a lot of current to saturate the core.  My first modulation reactor was a modulation autotransformer (one continuous winding, with taps for +HV, each modulator tube plate, and two RF final impedances).  I got more inductance when I used the +HV tap and one of the RF final taps, than when I used the entire winding from end to end.  With the whole winding, the ampere-turns increased to the point that the inductance dropped, exactly like what happens with a swinging choke in a power supply.  The whole winding probably would have worked well with a high-voltage low current final running the same power.

When I lived in Houston, John WA5BXO and I made up a modulation reactor out of two identical 12 Hy 500 ma power supply chokes.  We took them apart and removed the "I" laminations from each core.  We made a bracket and clamped the "E" laminations together (with windings intact) facing each other, separated by the same paper gap material that was used with the original chokes. You could visualise the resulting core as looking like a figure "8" with the winding on the middle part divided into two sections.  We connected the two windings in series, in proper phase to be additive, and experimentally determined the thickness of the core gap material that yielded the most inductance at the rated final amplifier plate current.  The best we could tell was that the resulting choke measured about 40 Henries, compared to the 24H that would have resulted from merely wiring the two unmodified chokes is series, and the whole thing was smaller and lighter in weight, with the two "I" sections removed and the remaining parts of the two chokes assembled into one unit.

I think John is still using  that reactor in his rig, even after it had been submerged in the muddy bottom of a creek for weeks following a major flood.

[Bacon, WA3WDR]

Yes - with a given winding, a given split core, and a given DC current, there will be an optimum gap width for maximum inductance at that amount of current.  Less gap, and the inductance drops because saturation begins to set in; more gap, and the inductance drops because the increased magnetic disconnection reduces the effective permeability of the core.

When you series-connect windings in a transformer with the right polarity, you will see more inductance - but a given amount of DC will produce more amp-turns and therefore more DC flux in the core, causing saturation at a lower amount of DC current.  You could widen the gap to get away from saturation, but that would reduce the inductance.  I wonder how to optimize this.  Would more turns and a wider gap produce more usable inductance at some given current, or would fewer turns and less gap be better?

And what if we somehow introduced a permanent magnet into the magnetic path, to cancel the magnetization caused by the DC?  The magnet could be laminated like the core.  Could this reduce the optimum gap?

[W2PFY]

K8VWX uses a pole pig as a reactor. Think he said he gets 66 henries out of it.

I use 25 on my Westinghouse and it will go down to 20 cps. I sound great!!!

de W2PFY

[k4kyv]

Yes - with a given winding, a given split core, and a given DC current, there will be an optimum gap width for maximum inductance at that amount of current.  Less gap, and the inductance drops because saturation begins to set in; more gap, and the inductance drops because the increased magnetic disconnection reduces the effective permeability of the core.

That reminds me of a device described in some early 1930's QST and RADIO articles.  It is a saturable reactor designed to imrove the voltage regulation of a power supply.  It  looks like a transformer, with E I laminations, with a separate winding on each of the three legs of the core.  One of the windings is in series with the primary of the plate transformer.  The power supply DC flows through another winding.  As the direct  current flowing through the load and the winding increases, it tends to saturate the core and reduce the inductance of the winding in series with the primary, reducing the voltage drop through the inductive reactance, thus boosting the primary voltage.  Adjusted properly, supposedly the current could vary quite widely with little effect on the output voltage.  The third winding was connected as a hum-bucking winding, since the winding in series with the primary of the transformer also acted as a primary winding in the reactor, and the winding in series acted as a secondary, the device would induce hum into the DC. 

Somehow, the hum bucking winding cancelled out the a.c. hum that would be induced to modulate the DC output of the power supply, but I cannot recall exactly how it was configuced that the ac would cancel out, but not the DC saturation current.

Like using a swinging choke, I suspect the dynamic voltage regulation using this thing was very poor, even though the DC meters may have shown a very stable output voltage.  This would have created problems when using it to power a keyed cw rf final, or a class-B modulator power supply.

Has anyone else ever heard of this circuit?

[Bacon, WA3WDR]

Don, that sounds like UTC 'variactors' at work. (Not to be confused with 'varactors' - a whole different animal.)  I saw variactors mentioned on page 20 of the 1936 UTC "Transmitter and Public Address Manual" in the Antique Transformer section on KD5OEI Patrick's b u n k e r o f d o o m   d o t  c o m   site.  They claimed that you could actually get about a 1% increase in supply voltage at full load with the variactor.

There are good details about the variactor in the description of its use in a controlled carrier system on page 13 of the manual.  The way they cancelled the AC flux in the DC coil was to have two AC coils in series on the ends of a figure-8 core, with the DC winding in the middle.  The AC windings were polarized so that the AC loops both saw DC, but the AC loops were out of phase, so it was a one-way magnetic street!   No AC in the DC winding, but the AC cores were both affected by the DC.  When DC flowed in the middle winding, it saturated both loops of the figure 8 and greatly reduced the inductance of the AC windings.

The manual also talks about hum-bucker filter units.  There are circuits shown for these.  The manual is one huge 46 megabyte pdf that takes a good long while to download, but it's worth it.

This stuff was freaking brilliant!  I have to wonder what these designers would have done with today's technology.

[k4kyv]

Don, that sounds like UTC 'variactors' at work. (Not to be confused with 'varactors' - a whole different animal.)

Bacon, that's it exactly.  The earlier QST/RADIO article was a build-one-yourself project.  Thanks for the info on the website.  I downloaded the entire pdf and saved it to disc.  That's one I didn't have.  Looks like he has loads of other hard-to-find catalogues and data sheets.  I may eventually try to print myself a hard copy.  Now I remember how the variactor works.  Not the same thing as a variac, either.

[WA2IXP]

Interesting reading . I have used microwave oven transformers as chokes. First I cut the small welds and remove the primary winding replacing it with another hv winding from another transformer. The winding is installed so that it aids and the original ground connection is connected to the others grd.This is the lead connected to the core. This is done so the original insulation scheme stays the same. The choke has to be mounted on standoffs or plastic . The I laminations are replaced with a paper air gap. Some I bolted. Later ones I tack welded. Have used 2 of these in series for hiesing on an 813. Have used trans. also coverted to choke with air gap leaving all windings (series pri and sec)  as original . Worked great in neg. lead of p.s. One hv winding measured over 30 H. but that was with no D.C. loading. The microwave diodes and caps(some caps have diode inside case) have been used for P.S. as well as coupling caps and neg peak 3 diode stuff.

[Bacon, WA3WDR]

That's cool, Jay!  I heard that the power transformers in microwave ovens have some kind of magnetic shunt and are unsuitable for chokes and power transformers and such, but your results indicate that there are possibilities. 

[WA2IXP]

yes, I think they are saturable reactors as manufactured. I forgot to mention to remove the shunts which are small laminations in the sides for the e-core. Knock them out. Wiht these gone they seem to make good power transformers. The last one I modified needed to be about 1200 volts so I removed hv winding and unwound about half from the inside which also left plenty of room for insulation so the secondary was easily insulated from core and could be used with bridge rectifier. The best setup is to use two in full wave with laminations grounded or floating by a series choke only. you'll get about 2000 volts.