4-65A Transmitter: Plate Supply Bleeder

[W9ZSL]

First it was a 350 watt rig built around a 4-125A.  Next it was a parallel 807 rig.  Now it's a 200 watter using a 4-65A.  Make up my mind already!!!

Bob, W4RFM a regular here, asked about my Hammond 2220 modulator iron I offered for sale.  One thing led to another.  In going back in the tech files I found that he has been planning to build a similar rig and wondered if a swap could be arranged.  Since I decided not to build the 807 low powered rig or the 4-125A rig, we traded mod iron.

Bob sent me a UTC CVM-3 and a 4-65A in exchange for the Hammond.  So far so good. I like the idea of building a 200 watt rig around that 4-65A bottle because of the wide range of plate voltages it likes which brings up my question.

Modulation is no problem because I can build a unit with 807's, 809's or 811A's.  I have a UTC S-45 plate iron so 700 to 750 VDC is a done deal.

However, what I've collected for power supplies doesn't stop there.  While the Packers were being slaughtered, I went down in the bazeboo and started messing with my plate transformers.  Yeah, my intent was using that 4-125A for "George" but the 4-65A is more practical and much cuter!

Here is what I came up with.  I have 2 Thordarson plate transformers.  Both are 115 V primary and good for 300 ma.  I did a big breadboard on the work bench with 120 on their primaries.  I set up a choke input filter with a 12 Hy followed by a 12 mfd cap. For the test I used two bleeders.  One was 100K and the other 60K; high wattage.

The lower power iron has a dual secondary.  With 120 on the primary and 100K bleeder I get 1575 and 1250 VDC.  With a 60K bleeder it's 1425 and 1150.  The second transformer has dual primaries; high and low.  Again with 120 I got 2250 or 1840 with the 100K bleeder using the high-side of the primary and 2050 or 1650 with the 60K bleeder using the low side of the primary.

Of course, the additional load of the plate would change that.  It's obvious that I have a lot of choices to work with.

My question.  I would like to get close to 220 watts input with that 4-65A.  Given the wide range of plate voltages I can obtain and using the recommended voltages for the screen and grid of that tube, how can I nail down the bleeder value?  I have plenty of caps and chokes so I guess computing how to get to get around 1800 volts is the question.  I need to know how to compute the best bleeder value when the tube is added to the load.

Is there anyone out there who also can't make up their mind with a project?  Oh, never mind.  I guess that's why we can never see the work bench only a slew of parts! 73!

Mike - W9ZSL

[KA2DZT]

Mike,

You need to go back and read your descriptions of your voltage test.  I can't follow it.  The hi voltage or lower voltage is a result of what??   Same bleeder different taps on the primary or same tap and different bleeders?

That Thordarson xfmr with the tapped primary is best run on the lower voltage tap.  Some of those early Thordarson tapped primary xfmrs were designed to run on 110 vac  later ones may have been 115 vac.  Putting 120-125 vac on that tap may over stress the primary.  The lower voltage tap primary has more turns than normally needed for even 120-125 vac.  So the xfmr will run better using the entire primary winding.

As for the bleeder current, anything in the 30 ma range should be good.  You are only trying to keep the HV from soaring when there is no other load but the bleeders.  If your supply is going to be keyed on and off (xmit/receive) then not much bleeder current is needed as the rig is keeping a much heavier current drain on the supply.  If the supply is going to remain on with receive then you may need a bit more bleeder current to keep the HV from rising up too much.

Try to clear up the description.  You have too many variables at once.  I never did well on those equations that had more than one unknown.

Fred

[W9ZSL]

Both transformers are capable of two voltages.  One has a single 115V primary and two secondaries.  It is 1600 and 1280 CT.  The second has a tapped "high / Low" 115V primary.  The resulting voltages are 4600 or 3750 AC, CT depending on which primary is used.  

The DC voltage readings were taken using the 100K and then the 60K for testing.  Alternating bleeders among the 4 secondary voltages, the result was 8 voltages DC; 2250, 2050, 1840, and 1650 for the dual primary iron and 1575, 1425, 1250 and 1150 with the single primary / dual secondary.

As far as the bleeder values, those were just some I had on hand.  I mainly wanted to find out if the transformers were good which they are.  I know how to calculate bleeder current draw, but am clueless about how to calculate the load of the 4-65A, how much voltage drop to expect as a result of the additional loading and finding the best bleeder value to come up with around 200 watts input.

Obviously the 100K is too high.  The 60K might work.  My pick for the iron would be the dual primary, low-voltage primary tap which supplies 1840 VDC with the 100K bleeder, but that isn't even close to drawing 30 ma. 

I could put a vaiac on the rig, no problem.  It was rescued from a Gates BC-1T so it will handle anything.  Question is, how much additional voltage drop will I get when the 4-65A is added?  The rig will be for AM on at least 75 and 40 meters.  Twenty and above, probably not.

  The ideal would be 1700 volts @120 ma for 204 watts.  MK

[N2DTS]

I would say, finalize the design first.
The hot setup would be to run the 4-65 at as high a voltage it will take, and modulate it with a pair of 811a's.

I do not remember what a 4-65 likes, never used one, but 811a's like 1200 to 1500 volts.

If the 4-65 wants 2000 volts, build a 2000 volt supply, and a 1200 volt supply for the 811's.

Or use the taps on the iron to give two voltages, one high for the rf, one lower for the 811's.

807's and other small tubes want 700 volts, which is far from what the 4-65 likes I think.
And they are only good for about 120 watts, and its best to have at least as much audio power as rf input power, things run clean and easy that way.

I would tend to go light on the bleeder, why waste a lot of power in that?

[DMOD]

If you could make a table or listing of your results it would easier to follow.

The second has a tapped "high / Low" 115V primary.  The resulting voltages are 4600 or 3750 AC, CT depending on which primary is used. 

If I understand you correctly, I would use this one with the CT secondary on your "low" primary tap with a FW SS rectifier system and you would get about (3750/2)X1.3 = 2,438 volts and use the 100k resistor as a bleeder. Bleeder should be a 100 Watt unit.

According to http://frank.pocnet.net/sheets/140/4/4-65A.pdf you can run the tube plate at 2,500 volts in Plate Modulated CCS.

I would recommend 12uFd on the first cap after the rectifier and 40 uFd to the right of the choke, assuming you want to continue to follow the choke filtering system.

What is the choke resistance?

Phil -AC0OB

[W9ZSL]

I'd have to check the resistance of the chokes.  As for the modulator, my UTC  VM-3 transformer is 125 watts.  I have 2 x 4mfd, and 2 x 12mfd, an 8 and a 10.  Chokes include 1 Triad swinging, Thordarson 12Hy @ 300 ma and another 12 @ 200.  With luck I'll nail another swinging choke on eBay tonight.

The two Thordarson plate transformers are rated as follows: both are 115 primary, 300 ma.  Their DC rating is for a two-section filter.

Type T-19P59, 550VA, single primary.  Secondary 1:  1560-0-1560 = 1250VDC; secondary 2:  1250-0-1250 = 1000VDC.

Type T-16P03, 865VA, tapped high/low primary.  Secondary High:  2140-0-2140 = 1800VDC (my first choice); Secondary Low:  1740-0-1740 = 1450VDC.

I was thinking of the following modified with a pi-net.  The 4-65A will work with any plate voltage between 600 and 2000 in AM service with 120ma max.

I also have a UTC S-43 plate transformer good for 1800 or 1500CT for the modulator.  This diagram has 807's but I have several 5933's that would work.  Fortunately the mod iron will work with 811A's also in which case I could use the T-19P59 low-voltage secondary for 1,000 VDC if a bit more audio is needed although p/p 811A's would be overkill at that voltage.  A P/P pair of 811A's can actually do 175 watts audio with zero bias and only 750 on the plates.

Thanks!

[N2DTS]

811's would last a long time at that power.
Unless you go high in voltage on the 4-65, you are not going to get much power out of it.

[DMOD]

Type T-19P59, 550VA, single primary.  Secondary 1:  1560-0-1560 = 1250VDC; secondary 2:  1250-0-1250 = 1000VDC.

Type T-16P03, 865VA, tapped high/low primary.  Secondary High:  2140-0-2140 = 1800VDC (my first choice); Secondary Low:  1740-0-1740 = 1450VDC.

I am pretty sure those values were for tube rectifiers that have a lot of cathode to plate resistance.

[W9ZSL]

Like I said, the modulation iron is rated at 125 watts.  I originally wanted to do something closer to 350 watts input using a 4-125A modulated with a pair of 811A's but decided to scale back and keep things really simple.  I also considered building a 120 watt rig.  200 watts input is plenty and considering I have not built any ham gear in years, this is enough project for me!

I did score an NOS Thordarson swinging choke tonight so I have two plus another pair of smoothing chokes to go with them.  I just dug up a 75K at 100 watt adjustable resistor so I should be able to ding around with more bleeder values.  

The 866 solid-state rectifier substitutes would explain why I'm getting higher values than the transformers are rated.  Working a 4-65A at between 1600 and 1900 volts will do fine.  According to spec sheets, any of those voltages with 120 ma and the tube will be happy.  MK

[KA2DZT]

The 16P xfmr is a bit older than the 19P series.  The older one probably has more iron than the 19P.  I use a Thordarson P-6411 in my HB xmtr.  All these xfmrs will make more voltage than they're rated mainly do to higher line voltage and SS rectifiers.  My 19P59 xfmr has much less weight than the P-6411 even though they have about the same ratings.  Early (30s-early 40s) Thordarson xfmrs had 4 digit part numbers.  Thordarson then began with the part numbers like the 19P series.  I think the 16P series were re-labeled earlier xfmrs.

The only way you can test your PS is to load it down to the expected full current and see what voltage you end up with.  Remember you will lose a little more plate voltage from the resistance of the mod iron.

With the power level you're running I would use 811s for the modulators  My HB 6146/813 rig uses 811s to mod the 813, 6550s to mod the 6146.  It will be hard for 807s to fully modulate the 4-65   I use 1500 volts (one supply) for the 813 and 811s.  At 1500 volts the 811s use 4.5 volts bias.  I use a zener in the filament xfmr CT to ground.

The 19P59 has the dual secondary which will work for both the 4-65 and 811s.  I would not use the more than about 1800 volts on the VM-3.

If you use the 19P59 xfmr you'll need two sets of rectifiers, one for each secondary.  You will then need separate filters.  The final should have a two section filter LCLC.  The modulator can be only LC.

The 19P59 doesn't make enough plate voltage.  There is an easy way to bring up those two voltages.  It will require another xfmr.  A power xfmr that can make 300 - 400 volts, like 600-800 volts CT can be used to bring up both voltages at once.  Use a SS rectifier circuit for the 300-400 xfmr.  Connect the diode rectifier output (no chokes, no caps) directly to the CT of the 19P59.  This will bring up the output of the two voltages by whatever you raise the voltage on the CT.  You would then continue with the same filters as before.  This boost xfmr will require the same current rating as the plate xfmr (300 ma).  You can change power levels by simple shutting off the boost xfmr without disconnecting anything.  Just simply switch on or off the boost xfmr.  The CT of the boost xfmr gets grounded as the CT of the plate xfmr is NOT grounded but is connected to the diode output of the boost xfmr.

This two xfmr set up reduces the VA power load on the plate xfmr as some is being supply by the boost xfmr.  The total VA power load the xmtr requires would be supplied by the two xfmr.

Fred

[N2DTS]

This is the fun part!

It makes things simple (and smaller) to run everything off one power supply, and if you have a variac, you can just pick the voltage you want.
I also built a rig that had relays to select the primary taps, it had a nice big RCA power transformer that had primary taps for 105, 110, 125 and 130 volts, and a front panel rotary switch selected the relays to the primary.

But, it often works best to run seperate supplies for modulator and RF, and pick voltages the tubes like to run at.
And its always a good idea to go big with modulatoer tubes. Mod iron also, but you are stuck with what you have.

Since the cvm 3 is good for 120 watts, its likely best to run it at 100 watts and keep the voltage (and current) down, so 811's might be a bad idea with that transformer.
Not good to have 340 watts into a 120 watt transformer...

Not sure about the cvm 3, but the cvm 5's I had were very poor in fidelity.

[w4bfs]

have you calculated your modulated impedance ?  .. a single 4-65 at 200W is going to require 2k hv and low plate current ... this is difficult because even with the multi-match transformers a high operating impedance brings in the undesired transformer parameters into full play ...this wrecks transmitter audio frequency response ... best to hold modulated impedance to around 8kohm max and sweep yer iron to see iffin you can live with it

ADDED: transformer leakage reactance is one of the worst of the undesirables

[W9ZSL]

My original plan was to use the lower-voltage 19P59 for the 811A modulator and the 16P03 for the 4-125A plate.  2000 volts minimum is recommended for that tube and the transformer won't put out that much voltage whereas the 4-65A can handle anything from 600 to 3KV.  Even with 1600 VDC on the plate at 120 ma, that's 192 watts and the 5933 modulator should be able to handle that nicely.  With the UTC S-43 Plate transformer I'll have a solid 750 volts and can use 807's, 809's or 811A's for an audio power of at least 120 watts...which should be plenty.  The 19P59 could end up as surplus.

[N2DTS]

Yes, but the 4- tubes do not seem to play well at all at low voltage.
You might get 100 watts out of it....

[W2PFY]

I thought about this layout many times as noted in Eimacs data on the tube. You could use a total of three 4-65 tubes, one in the final and two in the modulator. You don't have enough 4-65? contact me, I have some NOS that i would part with. Well anyhow, it would be interesting to see if 4-65 could be triode connected and if they would work as zero bias tubes?  Since I am lazy, I would try them as tridoes in the modulator and drive them with a reversed standard audio transformer.

Another nice thing about this PDF is Eimac notes all the parts needed.

[W9ZSL]

The nice thing about the schematic diagram and configuration (from the '49 East Coast Handbook) is the modulator iron uses a CVM-3.  I have a VM-3 which is the same UTC guts in a different housing.  Using the recommended voltages, the pin-outs are already listed so I won't have to do any guessing.  Anything other than what I have now means more parts hunting and expense and as I noted at the beginning of this thread, this is the third transmitter which has gone fleeting through my nogema in the past 12 years so there comes a time when the work bench must be cleared and construction started.  I'd even go as far as building this bugger with that link-coupled output with a home-brew coil and get totally primitive.  Yes, I have a low-pass TVI filter.  I know some hams still experiment with the old-style coupling as in the diagram and rejoice when their output cap doesn't arc, but even though I'm only planning on two bands and have what I need, modification of the 4-65A amp in question with a pi-net sounds more civilized!  See my Don Martin comparison. 

[KA2DZT]

With the line up you're suggesting, 4-65 at 2000 volts modded by 807s at 750 volts, the mod xfmr will have to be a big step-up in the turns ratio to make 100% modulation.  Not too sure about running the VM-3 at that high a voltage.  You may need a 2/1 step-up in turns ratio that becomes 4/1 in impedance ratio.

Fred

[W9ZSL]

Fred, I can't get 2KV with the plate iron I have that's why I gave up on building a 4-125A rig.  I can get between 1600 and 1800 max...1700 is realistic.  The VM-3 can handle that just fine.  The amp/modulator in the schematic is rated at 1600 VDC on the 4-65A.  I don't see a problem as long as I keep the plate voltage within that range.