The AM Forum
March 29, 2024, 01:56:14 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1] 2   Go Down
  Print  
Author Topic: effects of using a non-swinging choke in a choke input design  (Read 24064 times)
0 Members and 1 Guest are viewing this topic.
N3DRB The Derb
Guest
« on: August 16, 2009, 05:03:17 AM »

I'm wondering what the practical effects are of using a non swinging choke in the place of a swinger in a choke input supply. the handbook doesnt really address this too well, it simply assumes you'll be able to procure whatever you need, which in 1941 was true, now less so.  Tongue

Assume I have 1 choke and only one choke. On a load with no big dynamic current swings, say a rf deck,
where it's more or less on or off, would it be better design to use the choke as a swinger, or go cap input?

What about a load of wide current variations like a mod deck?

assume there is more than enough beef on the transformer and choke to handle any reasonable load and the choke in question is >10 henrys. lotsa scrote.
Logged
W3SLK
Member

Offline Offline

Posts: 2651

Just another member member.


« Reply #1 on: August 16, 2009, 08:47:40 AM »

I thought the idea behind the swinging choke was to improve voltage regulation. I'll be needing to visit this subject in the future but I'm pretty sure that is why.
Logged

Mike(y)/W3SLK
Invisible airwaves crackle with life, bright antenna bristle with the energy. Emotional feedback, on timeless wavelength, bearing a gift beyond lights, almost free.... Spirit of Radio/Rush
WQ9E
Member

Offline Offline

Posts: 3282



« Reply #2 on: August 16, 2009, 10:11:37 AM »

Part of the idea behind the swinging choke was that the inductance increases sufficiently at low current draw so that the inductance current product is still high enough (generally referred to as critical inductance) that it acts as a choke input filter.  With a fixed inductance choke, if your resting current draw is too low based upon the choke value then the filter will act as a capacitor instead of a choke input filter during low current draw.  Under that condition you will get wild voltage swings because if the current draw is too low then the capacitor will charge up to as high as 1.41 RMS under low/no load conditions and drop to around .9 RMS once the current draw is sufficient to meet the requirement of "critical inductance".

If you use the fixed choke AND the choke and transformer have the current capability then you could use enough bleed coupled with the bias to draw enough resting current to keep the combination operating in choke input mode.  Otherwise, you could design it as a capacitor input filter but realize the secondary current rating of most vintage transformers is based on a choke input filter since there is significantly more heat generated in capacitor input filtering due to the repetitive peak charging current.

The older handbooks should have the formula for calculating the minimum inductance, basically critical inductance is equal to supply output voltage/current (in mils)  So for a typical transmitter modulator supply if your choke is 15 H then if you are shooting for a 1500 volt output you will need a resting current draw of 100 mils (bias plus bleed) to keep your supply running in "choke input mode"
Logged

Rodger WQ9E
KM1H
Contributing
Member
*
Offline Offline

Posts: 3519



« Reply #3 on: August 16, 2009, 10:38:43 AM »

All chokes vary inductance with current but not at the typical 4X or 5X ratio of the swinger.

The swinger was introduced simply as a weight and cost savings measure and a smoothing choke will work fine as long as the critical inductance requirement is met. The smoothing choke can also be tuned as Henry Radio did in many of their amps.

The Bill Orr and earlier series of Engineers and Editors Handbooks give a much better explanation than the ARRL version.

Carl
KM1H
Logged
KE6DF
Member

Offline Offline

Posts: 784


WWW
« Reply #4 on: August 16, 2009, 11:42:05 AM »

Another way people used to cut down on bleeder current was to put a capacitor across the filter choke designed to resonate with the choke at 120 hz.

That way you could get by with a fairly small choke and low bleeder current and still have reasonable regulation.

See attached pdf from the Orr 1959 handbook -- the PS on the bottom 1/2 of the page.

Only a 5Hy choke for a 2500 volt supply.

Some issues with this, however.

The resonating cap has to be non-polarized and withstand HV spikes and high current. Not easy to find such caps and using too wimpy a cap will result in failure.


* resps.pdf (117.54 KB - downloaded 495 times.)
Logged

KM1H
Contributing
Member
*
Offline Offline

Posts: 3519



« Reply #5 on: August 16, 2009, 02:11:24 PM »

Quote
The smoothing choke can also be tuned as Henry Radio did in many of their amps.

As I said earlier above.

There have been many resonant choke discussions on RF amplifier forums and it is not as easy as it sounds. The frequency has to be a bit above 120 Hz otherwise there are fireworks and finding the correct LC with fleamarket components is close to impossible. Add to that and Henry chokes do have a history of failing as they age.

Carl
KM1H
Logged
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #6 on: August 16, 2009, 02:45:23 PM »

If you use the fixed choke AND the choke and transformer have the current capability then you could use enough bleed coupled with the bias to draw enough resting current to keep the combination operating in choke input mode. 

The way most broadcast transmitters do it is to use a common power supply for modulator and final.  That way, the rf final acts as a bleeder for the modulator supply.  But I believe you were planning to use separate supplies.
 
Quote
Otherwise, you could design it as a capacitor input filter but realize the secondary current rating of most vintage transformers is based on a choke input filter since there is significantly more heat generated in capacitor input filtering due to the repetitive peak charging current.

The biggest problem using an older transformer designed for choke input is that the voltage regulation will probably be very poor.  A transformer with extremely low leakage reactance is necessary for good voltage regulation when using capacitor input.

One possibility, if you have a smoothing choke, say 10 hy @ 500 mills, that can easily be disassembled, would be to take it apart and remove the spacing material that forms the gap in the core, and replace it with thinner material so that the new gap is much smaller.  That will convert it to a swinging choke.  The UTC series chokes all come in pairs, one smoothing and the other swinging.  The core and windings are identical; the only difference is in the spacing of the gap.  It might take some experimenting to get the gap just right.

With a swinging choke, I would advise at least 25 mfd of filtering.  With less filter capacitance, the regulation may look good on the meter, and the voltage may stay within 5% or 10% from no-load to full load.  But hook up a scope or LED meter to measure plate voltage, and you might be surprised how much the voltage bounces around, but the mechanical meter movement cannot follow it so it averages out the voltage.  With my gates, the original capacitor was 8 mfd.  When I tried running it on CW, even with more than 60 henries of filter inductance, the voltage would vary from about 1600 to over 4000 volts, even though the meter read 2600 volts +/- 100v.  I replaced the the 8 mfd capacitor with 25 mfd, and it now bounces only about 300 volts.  I had similar experience with my HF-300 rig when I tried using only 6 mfd in the modulator power supply.  Now, both modulator and final supplies use 28 mfd each.  I try to use as much capacitance as I can get away with without having to resort to step-start or blowing fuses when I key on the transmitter.

Perhaps one of those solid state relays that turn on just as the instant a.c. voltage is crossing the zero base line would help alleviate the blown fuse problem.

Another consideration with swinging chokes is to make sure you have at least the optimum inductance at full load.  This is about twice the critical inductance.  Less than optimum inductance results in excessively high peak rectifier current at maximum load.  This may not be so much of a problem with solid state rectifiers, but will drastically shorten the life of tube type rectifiers and tends to make MV rectifiers act cantankerous and puffy and sporadically flash over.

Here is a good article on power supply filters, originally published in Audiocraft, Feb 1956.

http://www.audioxpress.com/resource/audioclass/ga499ac.pdf

Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
N3DRB The Derb
Guest
« Reply #7 on: August 16, 2009, 04:24:23 PM »

Quote
But I believe you were planning to use separate supplies.


actually, I was planning on using a common supply for both the modulators and finals. I dont have any DREADNOUGHT BATTLESHIP class swinging chokes, and my plate transformers are both too big to actually mount on decks.  Cool  Both in the "75~120#" class of transformer. I cant lift them myself. Massive props go to K7Yuh-oh and Frank AHE for getting them here.  Cheesy

Plans are to mount each plate transformer to the bottom of each rack, with 3B28's and choke and bleeder resting in a deck mounted as normal just above the plate transformer. I'll have to be careful to tie my grounds together real well, as well as making sure the HV CT is referenced to the same point.

It sounds like with a decent value of henries, one can do quite well with using 1 choke and using a common supply for both final and modulator. Don, using the final load as a bleeder of sorts is a pretty groovy way of explaining it.

I'm going to use the Therma-Whore with the 4-400's and the Stancor with the 2 hole HK257 rig. They'll look pretty mean when they get clean. They're big units.



* P1020120.jpg (135.92 KB, 1280x960 - viewed 688 times.)
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #8 on: August 17, 2009, 10:50:35 AM »



Ok,

Let me see if I have this right?

The swinging choke is no gap or very very small gap, so the choke has high H at low currents and then drops H at high currents?

The regular or "smoothing choke" is gapped core and so for the same windings and core has less total H, but remains more constant H vs. current drawn?

Do I have this correct or confused?

            _-_-
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #9 on: August 17, 2009, 12:37:53 PM »

BEAR, that's pretty much it.

And Derb, I use a  single 10hy 1 amp smoothing choke and 50 mfd filter cap in the little 8005/805 rig.  The same power supply @ 1350 volts runs the rf finals, the modulators and the 845 audio drivers, and the voltage regulation is very good.

When I run the rig on CW, the 845 drivers stay energised, so their class-A plate current gives me enough bleeder effect, along with the bleeder resistor, to hold the voltage to within 5% key up vs key down. I disconnect the plate voltage from the modulator when I switch to CW.  The keyed waveform on that transmitter is better than it is from either of my other transmitters.
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
WQ9E
Member

Offline Offline

Posts: 3282



« Reply #10 on: August 17, 2009, 12:55:48 PM »

Don,

Johnson does the same thing with the Desk KW.  The 810 modulators are biased to draw sufficient current in CW to provide better regulation.
Logged

Rodger WQ9E
N3DRB The Derb
Guest
« Reply #11 on: August 17, 2009, 03:07:17 PM »

well, that will be the way I go. Why build 2 supplies when 1 will do? I'll just make sure it has a decent bleeder rx on it, maybe slightly more than normal. I have a 50K 120 watt job waiting. half a amp ccs which is what the ThermaWhore is rated at should handle it ok fine.  Cheesy

I have a Thor choke from the 30's to use with the P-8030, but the damned thing has not got a model number on it anywhere I can see. I have no idea what it is except tht its a choke and it damned heavy and big. I bet that will mean it will work.  Roll Eyes

Logged
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #12 on: August 17, 2009, 03:32:45 PM »

Thordarson was notorious for changing all the component part numbers every year or so.  If you don't have the catalogue for precisely the year the transformer was manufactured, you are probably SOL.  Very often, the same exact part is listed in catalogues for adjacent years but they have different part #'s, that even use the same format.

The problem with the common power supply is that maximum modulation capability is strictly a function of modulation transformer turns ratio.  If you are stuck with a mod xfmr with a ratio that allows a maximum of only 85% modulation, you are stuck with that.  With separate supplies, you can raise the modulator plate voltage or lower the PA plate voltage to increase  modulation percentage.  The only alternative is to use the old Heising modulation scheme of a dropping resistor in the PA plate lead, bypassed with a capacitor.
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
Tim WA1HnyLR
Member

Offline Offline

Posts: 159


WWW
« Reply #13 on: August 18, 2009, 02:03:11 PM »

Well ,Well ,Well,
There is another approach . Derb. Try resonating the philter reactor with a parallel condensher. You may have to play around with values. In theory resonating things @ 120Hz is the goal. I have used the parallel resonating cap method along with Art Collins in a number of transmitter and receiver power supplies with great results . The largest one was the Collins transmitter used for WBCQs 5110 service . The plate supply was a 7500 volt 167Kva pole pig with a bridge around it and a 1 henry 11.5 amp filter reactor from an old RCA tv transmitter. As the transmitter was comming on line the RF drive was delayed for a moment . The plate voltage would soar to better than 7Kv operating voltage. The power supply voltage was set up for  5800v under load. This problem led to some scary arcs and bangs at sign on some times. I took 3 -4mfd 10Kv oil condenshers in series and place in parallel with the  1Hy reactor. The plate voltage came up to 5900 volts with the PA pulling 500Ma of idle . The plate voltage settled in @5800 under full load. Try it out Derb,
DeTim WA1HnyLR
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #14 on: August 18, 2009, 03:44:24 PM »

Yes, I have done the little cap across the choke also, a good use for all those little value oil filled caps!

I have also just used a choke input, a lot of capacitance (40 or more UF), and a good bleeder resistor and had no problem with it.

When I started out, I went overkill, those nice UTC commercial chokes, 12 henrys at 500ma ccs were $10.00 at fests, so I put 2 chokes on the deck, and three big oil filled caps, step start, bleeder, two 14kv 1 amp diode bricks in each leg, and ran the thing at 400ma...

I had one swinging version and put it in the mod deck power supply, but they were rare, I had 6 of those chokes and only one was a swinger.

The 3X 4D32 rig does need a seperate modulator power supply though, its slightly overloaded at 340 ma (rf deck) and the 811a modulators on peaks (300ma?) at 1200 volts.
Thats over 600 ma on peaks on a 500 ma choke, and the voltage does sag a bit.
But it does not seem to matter much, it does 300 watts carrier and over 1200 watts PEP.
I think I have about 60 UF in the power supply.

The power supply uses one of those fair radio specials, solder sealed chassis mount, Collins transformer, 2250 at 560 ma I think it is, cost something like $30.00 years ago...
Wish I had another, should have bought a lod of them...

Brett




Logged
KM1H
Contributing
Member
*
Offline Offline

Posts: 3519



« Reply #15 on: August 19, 2009, 02:57:07 PM »

Quote
I took 3 -4mfd 10Kv oil condenshers in series and place in parallel with the  1Hy reactor

Was that pure blind luck or was it calculated?

With many others using tuned chokes it has been found that resonating a FW PS at 120 Hz is a formula for disaster. Going about 15% high appears to work very well and at a calculated (ideal tolerance) your circuit came in at 138 Hz.

Being able to come up with the LC values necessary, for a HB supply, without custom ordering, could be an impossible situation.

Carl
KM1H
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #16 on: August 19, 2009, 03:59:48 PM »

I just guessed at mine.
I used a .5 or 1 uf, because that was what Collins used.
Never had a problem in 30 years...

Brett
Logged
KM1H
Contributing
Member
*
Offline Offline

Posts: 3519



« Reply #17 on: August 19, 2009, 08:00:45 PM »

You may not experience problems at the low voltages used. The failure mode is choke breakdown from excessive HV overcoming the rated insulation. The failures I know about are with normal, expected voltages in the 4-7 kv range.

Henry Radio used tuned chokes in many of their products and there have been many reports of choke failures there also in amps with decades of use. Those could be insulation deterioation and a less than ideal resonant frequency.

Carl
KM1H
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #18 on: August 19, 2009, 08:10:50 PM »

Yes, the cg108/109 are rated at 7000 volts I think, and I run them at 2000.

But that is the way to build things, over rated....

Brett
Logged
w1vtp
Member

Offline Offline

Posts: 2638



« Reply #19 on: August 19, 2009, 09:38:32 PM »

Gosh

If that's true that the gapping is the only thing that makes the difference between a swinging and a smoothing choke, I have a new in box Triad 300 MA swinging that I could take apart and change over to a smoothing.  Sweeet!

Al
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #20 on: August 19, 2009, 10:52:41 PM »



Al,

...you can, but keep in mind you will lose H as you increase the gap... so there is some calculating and dorking in practice (test rig) to find the right gap for the H you want/need vs. the max I drawn...

               _-_-bear
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #21 on: August 20, 2009, 07:40:06 AM »



On the subject of capacitors to resonate a smoothing choke, I have an idea. Others have discussed that this is tough territory for a capacitor since the capacitor current and voltage put the device into a hostile environment.
I once used a .47 uf 450v Black Beauty to resonant the choke in a 300v power supply. The circuit worked wonderfully until one day the cap shorted out.

With all this in mind, I wonder if a series resistor with the cap would be of any benefit? Sure the resistance of the choke does this, but adding more might lower the 'Q' such that the capacitor environment is lass harsh. Maybe a few hundred ohms would be of benefit without degrading the whole effect too much.

Also for low B+ power supplies, say 500V or less, most of us have many good resonating capacitor candidates in our junk boxes without realizing it. These days switch mode power supplies often have a nice capacitor across the AC line voltage. These are usually rated at 275 VAC, are polypropelene, and are often valued at 0.15, .25,  .5 and 1uf. I looked up the data sheet on some I had here, and these dudes have to withstand over 2000 vdc. The polypropelene design has low ESR, so these caps should tolerate a fair amount of current. The variety of values there just for the taking on the pile of SMPS's many of us have in the junk box should allow us to tune a choke with a series parallel combination of these little dudes.

Maybe someone will try these caps, or series resistor idea and report back to the group.

73,
Jim
WD5JKO
Logged
w1vtp
Member

Offline Offline

Posts: 2638



« Reply #22 on: August 21, 2009, 01:22:44 PM »



Al,

...you can, but keep in mind you will lose H as you increase the gap... so there is some calculating and dorking in practice (test rig) to find the right gap for the H you want/need vs. the max I drawn...

               _-_-bear

Bear

My swinging is a Triad C33A. 

DC current range = 30 / 300
Ind Range = 25 / 5
Res ohms = 105
Weight = 7.6 lbs
Test voltage (kinda wimpy, in my opinion) = 3 KV

Wonder if I could get the gapping so that I'd have ~ 10 Hys @ 300 ma

Al
Logged
KE6DF
Member

Offline Offline

Posts: 784


WWW
« Reply #23 on: August 21, 2009, 05:51:07 PM »


Wonder if I could get the gapping so that I'd have ~ 10 Hys @ 300 ma

Al

Seems like you have a nice swinger there. They are generally harder to find than smoothing chokes.

Why not look for a 10hy smoothing choke on ebay rather than going to all the trouble to change the one you have?

Or perhaps just trade it.
Logged

N3DRB The Derb
Guest
« Reply #24 on: August 21, 2009, 07:31:06 PM »

yeah good swingers are much harder to find. keep the swinger and look for a smoothie.

actually, i think I have a Triad smoothie in my iron stash. let me me look an see.

yep, its a C22A 10 h @ 500ma. Al, if you want it it's yours. this is also test voltage 3000.

I'm not going to use it so lemme send it up there.
Logged
Pages: [1] 2   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.073 seconds with 18 queries.