K1JJ Maul PS Choke

[W7SOE]

Schematic here:

http://www.amwindow.org/tech/htm/813/813ps.htm

What is the required voltage rating on the .5A 10H filter choke?

Rich

[Steve - WB3HUZ]

Preferably the supply voltage plus some safety factor like 2x. But if you put the choke in the negative lead and isolate it from ground, you can usually get away less.

[W7SOE]

Yea, that sounds reasonable.  Might be another tough part to find, 10 Hy @ .5A and ~ 2500V.   

Rich

[KA2DZT]

Yea, that sounds reasonable.  Might be another tough part to find, 10 Hy @ .5A and ~ 2500V.   

Rich

All chokes should have a voltage rating on them.  Usually you will see a RMS TEST voltage rating.  What does it mean?  Example;  if the RMS TEST voltage is 5KV, the choke is good to a maximum of 2KV rms.  You subtract 1KV from the 5KV RMS TEST voltage and then divide by two.  Sometimes a choke may be rated with a DC rating, say 2.5KVDC.  That choke should be good to 2.5KVDC  Some are rated with a PEAK voltage.  Don't exceed that peak rating.

Don't make the common mistake of seeing a RMS TEST voltage on a choke and using the choke at or near that voltage and then wondering why the choke shorted to ground.

If you find a choke that doesn't have a voltage rating on it,  post the make and model number and I'll check my transformer books for the ratings.

Placing the choke in the negative lead of the power supply like JJ has it, removes the need for high voltage chokes.  For some reason, which I forgot, I don't like using chokes in the negative lead.

Fred, KA2DZT

[K1JJ]

Remember that even though a choke in the negative lead may have less voltage to ground potential (even though I still mount it on Plexiglas or wood) there is still a peak voltage induced across it over a complete cycle from winding terminal to terminal. The inductive reactance di/dt? IIRC, based on current drawn will produce voltages that require the choke to have a good internal insulation/isolation in respect to the windings themselves.

T

[The Slab Bacon]

Placing the choke in the negative lead of the power supply like JJ has it, removes the need for high voltage chokes.  For some reason, which I forgot, I don't like using chokes in the negative lead.
Fred, KA2DZT

I seriously advise against this. I learned the hard way. It places the center tap of the transformer at the peak of the ripple voltage. This can easily cause an insulation breakdown between the center tap and ground crapping out your plate iron. Also do not use FWB rectification unless the the transformer is rated for it!! It will also break down the centertap to ground insulation. Most plate iron is designed to run with the CT directly grounded, and that is the way you should run it unless you like playing russian roulette with your plate iron.
If it is rated for it you are ok, but if not, dont chance it!! I learned the hard way!

If your choke is not rated for the voltage, you can set it up on insulators, but beware of a possible electrocution hazard if the insulation breaks down. (If you set it up on insulators, put a plexiglass box around it or something so you cant come into contact with the case)

[K1JJ]

I can see your point there, Slab.  Next time we modify the power supply schematic, I'll have the fullwave CT circuit choke put in the positive lead. No sense pushing our luck.

The other alternate schematic of the HV bridge rectifier showing the choke in the negative return of the diode stack is OK, however. This assumes the transformer is rated for bridge operation.

BTW, in real practice, I've never had a smaller plate transformer short on me despite using a choke in the negative CT lead or  when using CT transformers in fullwave bridge config.    Maybe I've been using transformers that were rated for it or just lucky...

OTOH, my main supply of 25 years uses a pair of pole transformers which I put the secondaries in parallel and uses a full wave bridge. (primaries in series = ~3-4KV out) There is no CT to worry about in these pigs, so it is rated OK.

T

[k4kyv]

Some pole pigs do have a CT.  Usually the winding is wound in two identical sections and split symmetrically, with several taps near mid-winding on each section. This is used for voltage adjustment, with a switch wired to select which taps are strapped together, to adjust the exact turns ratio.  With this configuration, the CT has to be rated to be at least half-voltage above ground, since one end of the HV winding is grounded in pole service.  But since the secondary (primary in pole service) is usually wound symmetrically, both ends of the HV winding are normally fully insulated for HV.  With such a transformer, it would be safe to either use it in a FWB circuit or running the choke in the negative lead.

I suppose it is possible you might run across a pole pig that was wound and insulated like a microwave over transformer, with HV insulation at only one end of the winding and the other end grounded or intended to be grounded, but I have used many pole pigs over the years and have never seen one wound that way.

[KE6DF]

I did a little playing around with the Duncan Amps power supply simulator.

It doesn't support negative lead filtering, but it does let you get the voltage across the choke and from that you can calculate the voltage in the negative lead configuration with a FW (not bridge) circuit.

Using a choke input filter in the negative lead, the voltage across the choke would vary from +0.55*V to -0.86*V where V is the RMS voltage between the center tap and one of the hot secondary terminals (in other words, half the total RMS voltage of the transformer).

This is the steady state voltage swing across the choke.

So as Slab Bacon says, there is a pretty large ripple voltage swing on the center tap and a transformer that "expects" the center tap to be grounded may have a problem.

But this isn't the worst of it.


During the first few cycles of startup, the voltage across the choke in the negative lead configure varies from +1.4*V to -1.4*V. So there is a very good chance that a negative lead design will put a voltage transient on the center tap that the transformer can't handle during startup.

The bottom line is that negative lead filtering isn't the panacea one might think..

[KE6DF]

All chokes should have a voltage rating on them.  Usually you will see a RMS TEST voltage rating.  What does it mean?  Example;  if the RMS TEST voltage is 5KV, the choke is good to a maximum of 2KV rms.  You subtract 1KV from the 5KV RMS TEST voltage and then divide by two.  

Fred, KA2DZT

In the UTC catalog it says that the working voltage for the S series transformers is just what Fred said above. You take the test voltage, subract 1000 and then divide by two.

In the old Kenyon catalog I found on the Web there is no working voltage specified -- just a peak voltage.

But for the heck of it I looked at some of the schematics in their transmitter catalog.

And in specifying which chokes to select, they used the above formula as well.

For example in the power supply for a 1KW transmitter, in a power supply delivering 2KV,  they specified a choke with a 5KV test rating.

The same formula held in other designs I checked.

I think chokes tend to get beat on pretty hard, and trying to make do with under rated chokes is probably the cause of a lot of failures.

[K1JJ]

"During the first few cycles of startup, the voltage across the choke in the negative lead configure varies from +1.4*V to -1.4*V. So there is a very good chance that a negative lead design will put a voltage transient on the center tap that the transformer can't handle during startup."


That's probably the reason many of these transformer shorts occur when FIRST turned on. Not only is there a current surge in the diodes and capacitors to worry about, but this negative lead problem too, if used.

It says a lot for using a soft start on the HV primary. For years I've used the simpest soft start possible for my big HV supply. I have a 5 ohm 200 watt resistor in series with one 240VAC line going to the transformer. I turn on the main breaker, wait a few seconds for the 140 ufd caps to charge, then I close a second breaker across the 5 ohm resistor. Manual, but effective.

Another thing - After having many, many HV dead shorts over the years in transmitters that were being tested or operated, I've never blown my solid state diode bridge stack. I use a fine wire from RG-213 coax shield between two ceramic posts in series with the diode stack negative and positive leads. These "fuses" always flame out and open, saving the diode stack.  The pillars are spaced about 2" apart. The breaker is not as effective as the fine wire in the HV lead. I also have a permanent 10 ohms in series with the HV lead to the final as a current limiting failsafe.

T

[KM1H]

Ive made a habit of mounting the chokes up on insulators even tho they are running within ratings. They seem to be the most unreliable part in many supplies, even commercial. Ask anyone who has had a tuned Henry choke let loose also.

I dont go into a PS with the voltage on but Id still suggest an alarm indicator if the case should become hot

The only time I'll mess around with a CT is in low voltage apps such as a dual voltage supply such as 250/500V, 350/700, etc aka McCoys Economy Supply. Just dont use 6X5's as I found out back then

Carl

[Opcom]

PSUD says 1800V peak to GND steady state, and +/-2800V to GND peak at turn-on.

I forgot to mention that there is 5600V across the choke winding there for a moment.

[KE6DF]

Right. Similar to the case I simulated.

The real killer seems to be startup.

Although there might also be a large transient when power is turned off and the mag field in the choke collapses. I don't think there is a way to simulate that with the Duncan program.

Some BC-610s had a spark gap on the choke according to the manual.

And those chokes were mil spec rated at 10KV if I remember right. That in a 2.5KV supply.

[Steve - WB3HUZ]

One supply I have has a cap and resistor to ground right at the choke input. I guess these shunt the pulse(s) to ground. Could someone simulate this - 1uF/5kV and 5k/100W in series. The choke is in the positive lead, not the negative lead. TNX

[k4kyv]

All chokes should have a voltage rating on them.  Usually you will see a RMS TEST voltage rating.  What does it mean?  Example;  if the RMS TEST voltage is 5KV, the choke is good to a maximum of 2KV rms.  You subtract 1KV from the 5KV RMS TEST voltage and then divide by two.  Sometimes a choke may be rated with a DC rating, say 2.5KVDC.  That choke should be good to 2.5KVDC  Some are rated with a PEAK voltage.  Don't exceed that peak rating.

Don't make the common mistake of seeing a RMS TEST voltage on a choke and using the choke at or near that voltage and then wondering why the choke shorted to ground.

It sounds like a safe operating margin, but does anyone know how that figure is derived?

In the example above, if the RMS test voltage is 5 KV, then the peak test voltage is 7070 volts (Vp = 1.414Vrms

The 2KV RMS voltage equals 2828 volts peak.

Maybe the test voltage was run using a hi-potter, which would allow a higher peak test voltage because the current from a hi-potter is usually too low to puncture the insulation, and the component would fail long before reaching that peak test voltage if a full current source like the output of a plate transformer were used for the test.

Or is that formula merely to assure a wide margin of safety?

[KA2DZT]

The RMS TEST voltage comes from the manufacturer's specifications.  It seems to be somewhat standard since I've found most manufacturers describe it the same way, (subtract 1KV then divide by two).  If the operating RMS voltage is 2KV then the peak is 1.414 above that.  That is why it's called the RMS TEST voltage and not the PEAK TEST voltage.

I'm not exactly sure how they run the test, but I would guess that they put the test voltage on one winding terminal and the case and measure any leakage current.  I may be wrong on this, but that's not important.  It's only important that one uses the choke within its ratings.

It's only the input choke that sees the peak voltage,  a second smoothing choke doesn't see much more than the DC voltage.

Fred, KA2DZT

[K5UJ]

I have a friend who has a plate transformer with CT on the secondary but because it is a newer Dahl (made after around 1970) and came out of a bc rig with a FWB rectifier we think it was made to be used with either fwb or a pair of diode strings and grounded CT, but the older ones from the 50s like the Thordarson I have were made when the rectifiers were a pair of 872s so the FWB is not an option.

I have been wondering if resting the iron on ceramic floor tiles is a good way to insulate the windings.  I have a few stacks of plain ceramic tiles left over from some remodeling here and last week I was pondering setting down a mod reactor I have on a couple of them.  Cheap and widely available from home despot etc. in a variety of sizes.

[The Slab Bacon]

FWIW, I have a transformer out of a 1 KW BC rig rated for FWB rectification. (3700v out) It is a modern (1970s) open frame transformer. The double bobbin (looks like a pair of binoclears) bobbin / coil design is velly interesting! The primarys (with a million adjusting taps) are wound right on the core, but the secondary is where it gets interesting! The secondarys are wound on plastic rod spacers, holding it somewhere between 5/16 amd 3/8" off of the primary. The secondary connections are screw terminals on the sides of the secondary windings and, of corse both secondaries are wired in series! It's a big mother and I think it's grossly under rated.

But I'm sure the winding spacers are to prevent primary / secondary insulation breakdown in FWB service.


                                                                  the Slab Bacon