Whatever happened to.....

[K9ACT]

Filter chokes?

All the old manuals and projects use filter chokes in power supplies and all the newer ones seem to make believe they do not exist.

I am building a 6AG7/807 exciter for my rig and could not find a single choke at the Sandwich Hamfest today which seems to confirm that they do not exist.

What is it about new technology that allows one to bypass the filter choke in a HV power supply.

Jack

[Ian VK3KRI]

What is it about new technology that allows one to bypass the filter choke in a HV power supply.

Jack

Bigger caps and silicon diodes? 

The 6ag7/1625 TX in the '63 handbook used tube recifiers and 2x16uF caps in series and a 7H choke,
The 12by7/1625 TX in the '67 handbook used Si diodes absd 2 x40uF caps in series and no choke.

Incidently the earlier cct used a 400-0-400 transformer and the later 270-0-270 - pretty close to 1.41 voltage ratio.

                                                   Ian VK3KRI

(I've obviously read those handbooks TOO many times!)

[WBear2GCR]

Chokes are still around...

But commercial supplies have tried to reduce size & weight, as well as cost.
Big Iron costs $$ compared to caps.

Thus, switching supplies.

You can make ur own choke, if you like.
It's easy, since you do not have to make the windings perfectly pretty.
You can take an existing bit of transformer iron, and even use the existing windings assuming they are of sufficient gauge, and the inductance you need is GREATER than what you are aiming for. Why? Because ur going to take the thing apart and flip the lams around so that you have a gap. Then the total mH or H will be lower than with the thing set up with alternating EI lams (no gap).

Alternately, you take the bobbin, or make a bobbin and wind up the requisite number of turns according to "the formula". Magnet wire is to be used.

If you are going HV or need/want to, you can find an old pressure cooker, fill with appropriate goop (usually an electrical varnish), suck the air down through the hole in the top of the lid - takes the trapped air out of the windings. Then after a bit, release the vacuum and pull the sucker out, bake or air dry... done. Custom choke to specs.

fwiw, some motor rewinding shops do vacuum varnish insulating, some just do a dip w/o the vacuum part. They would probably sell you a gallon or so of varnish for not much $$... They will also usually sell you some magnet wire, or even do a wind...

If the current draw is modest or constant, you may be able to get away with just the turns of a transformer used as a choke - test the operation in the circuit to see that the core is not saturating and that the smoothing is working sufficiently well.

 

                 _-_-WBear2GCR

[k4kyv]

Choke input filters have traditionally delivered better voltage regulation.  But I recall reading somewhere that capacitor input filters can give excellent regulation on the condition that the power transformer have low leakage reactance.  Most power transformers designed for tube type rectifiers and choke input filters have substantial leakage reactance, and would perform very poorly without a filter choke.  The voltage would sag substantially under load.  So you never know how well a hamfest find is going to perform in a supply without a filter choke, unless you have some specs on it.

Another factor is that many final amplifier power supplies are designed with SSB in mind. That means light duty cycle.  So they just put in a large filter cap, and hope for the best - that the voltage well stand up for the short instantaneous peak demand from a SSB leenyar.  Put a full 100% duty cycle on one of those amplifiers, and the power supply often goes belly-up.

Finally, there is a  difference between dynamic and static voltage regulation.  A power supply may have good static regulation.  You place a voltmeter at the output, and load the power supply down.  The voltage sags very little.  Let the supply run a linear, cw final or AM modulator, and the voltmeter reads almost the same with load and no load.  Under modulation, there is very little movement in the voltmeter.  You may think you have excellent voltage regulation, until you probe the power supply output with an oscilloscope.  You may be  dismayed at how much the instantaneous voltage kicks around under a varying load, but the variations are of such short duration that the mechanical meter movement cannot follow.

A case in point is my Gates broadcast transmitter that I converted to have CW capability.  I managed to get enough inductance in the filter choke by allowing the modulation reactor to double as a filter choke when the transmitter goes into the CW mode.  The voltmeter reads 2700 volts on the final, key up.  Key down with the final loaded to 500 m.a., the voltage only drops 100 volts, down to 2600. At first I though wow! what excellent regulation.  But the cw keyed waveform was atrocious.  So I set up the scope to monitor DC plate voltage.  I found that when I pressed the key, the voltage immediately took a nosedive to about 1500 volts, then rose back up to 2600.  When I unkeyed, the voltage soared up to 4000 volts for an instant, then dropped back down to 2700.

I was using the stock 8 mfd (or is it 6?) filter cap in the output.  I replaced it with  a 25 mfd cap, and changed the stock 100k bleeder resistor to a 50K.  That immediately improved the dynamic regulation to an acceptable level.  It still bounces around some, but the waveform looks fairly good, and the instantaneous (dynamic) variation is only about 350 volts, compared to 2500 volts with the smaller cap.  To get excellent regulation you need to go to about 100 mfd of filter.  But then you begin to run into inrush current problems, requiring step-starting the power supply.

[Steve - WB3HUZ]

More on static and dynamic regulation here.

http://www.amwindow.org/tech/pdf/geps1.pdf

http://www.amwindow.org/tech/pdf/geps2.pdf