The AM Forum

We've been here before regarding using a single 6V6 and a UTC S-8 to drive a P/P pair of 811A's.  There was a lot of discussion as to whether the S-8 had a high enough rating to do the job.  Meanwhile, I've come across a Stancor A-4416 which is 3-10K CT primary, 40 MA / Side 5:1 Pri to 1/2 Sec ratio.  There is a speech amp driver in the 1949 ARRL handbook where this iron is used with a 6SN7 as a driver.  The modulator described there is a 6L6 job.

My modulator will be a pair of 811A's.  Given the wide primary impedance of the 4416, what if I were to use a pair of 6SN7's paralleling the triodes in each tube externally on the sockets, and then putting 2 + 2 triodes in P/P?  It would be like using a pair of 6B4G's.  I could then drive those with a 6C4 through the UTC S-8.  If you checked the 1957 handbook page 256 you would see what I mean.

Of course I could use a pair of 6B4G's, but hey are expensive as hull and I have several 6SN7's.  The driver in the handbook with the 6B4G's would have more than enough poo to drive the 811A's, but if a pair of 6SN7's could do the job operating as 4 triodes, why not try it?

there is merit in your premise ... airc a pair of 811 will need 5 or 6 W of drive as a minimum with 10W being a better number .... calculate the min impedance from peak swing and drive power ... these numbers are generally just a few hundred ohms .... reflect this impedance back to the primary by the square of the turns ratio .... this is generally a few k Ohm .... whatever you choose for the driver tube will need to be able to deliver req'd power at that load Z ... I would breadboard your parallel push-pull 6sn7 driver into a substitute resistive load and measure performance ... I suspect it may be a bit light ... something a bit bigger such as a triode connected 807 could fit the bill ....73

Is there a reason why you want to do it the hard and low fidelity way?
The more iron in the path, the more phase shift and frequency response issues.

What is the going price for 6b4's?

6B4's are going for around $100 a pair. They can drive a pair of 811A's no sweat. It's just a thought right now.  I know the more iron the worse the fidelity.  I've been in the sound recording business since 1966.

Here is the schematic.  The 6B4's would be replaced with 6SN7's with each tube's triodes wired in parallel on the socket.  The output iron would be a Stancor A-4416.  I'd increase the value of the coupling caps above .003.

Instead of two 6SN7 or screen grid tubes, why not consider the 6N7, a common-cathode zero bias dual triode, will do 10W to an 8K p-p load @300V and drive those 811's for cheap. I'm surprised it is often overlooked, but is designed for ease in this application.

Two in parallel are not necessary but can be used with no fuss for a little extra current on them 811 grids, drop the maul! Matching is not necessary with them but a 33 Ohm cathode resistor can be used in each one. It only takes about 1/2 Watt to drive a push pull 6N7 all the way.

Drive the push pull 6N7 with the smaller transformer and the 6V6 in triode connection or better yet with a second 6N7 as stated in the datasheets.

There are many useful 6N7 circuits, including driving one set up for zero bias push pull with one set up for class A. The datasheets spec this. There is also a RC amplifier data for it.

I bought 10 on ebay for $4 each. They come also in GT style but the metal might be better for in an RF area. Great tube, the 'phools don't care for them, and cheap. Tons of circuits show up in google images for 6N7 amplifier.


-Opcom, the 6N7 ambassador

So you do not want to drive the 811 grids with an 8 ohm amplifier and want to go old buzzard.
I think I have a bunch of 6b4's, I will have to check when I get home.
Got anything to trade?

Can anything be done with 6080's?
I have a lot of those.

That 6N7 set up looks really interesting!  I think I actually have some of those and the circuit is way simpler than anything else I've considered.  Questions.  The input shows a mic transformer, but since this is for mobile use, I'm assuming the battery shown can be eliminated with the filaments powered off the 6.3 winding and can a dynamic mic be used?  That R1 variable would seem to be unnecessary if the PS was similar to the one shown in my 6B4 schematic.  For all that matter, it seems to me a single 6N7 could be used in place of a pair of 6B4's?

I use a 6bl7. 6SN7 pinout (BBD.) Sweep tube power. If using minatures I use 12bh7.
don

Looks like there are a few better choices for those output tubes other than the 6B4G.  After checking out the specs, the 6N7 looks like a good bet!  The Stancor A-4416 will handle 80 MA total on the primary and it appears a single 6N7 in class B would pull around 70 MA with 300 V on the plates for a solid output of 10 watts.  Using the original diagram with the 12AX7 and 6C4 into the UTC S-8 followed by a single 6N7 operating P/P rather than a pair of 6B4's would probably work just fine.

use as low of a plate resistance as possible in class B driver service to minimize distortion generation ...

Replace T1 with a phase inverter.

http://www.amwindow.org/tech/htm/rangmod.htm (http://www.amwindow.org/tech/htm/rangmod.htm)

Then there's always this one.  Difference would be substituting one 6N7 for the pair of 6B4's.  This uses 811A's in the modulator.  Any suggestions for better frequency response like changing the value of the .01 caps in the inverter output?  The mic pre-amp cap as well?

You can make all the electrolytic caps bigger, and remove the one in the mic preamp cathode.

I have three 6B4's.

N2DTS, what would you want for them?  I have a lot of parts.  You mentioned swapping.  Anything special you can use?
Mike

I will test them and let you know if I come up with anything to swap for or a low price...

imo this is a strange design ... the push pull 6N7 is class a and can be done with a 12AT7 since each half only has 1.5 mA current available ... this cannot provide but a tiny drive current to the 6B4s following which are self-biased into class ab with cathode bias .... I don't believe it works but to a point without clipping the waveform, which may be the designers intention ... reminds me of a military way of doing things

Good deal with the 6B4 tubes.

Here's another 6B4 driver circuit using a 12AX7 and 6C4.

Al VE3AJM

Yeah Al, that's the same diagram I attached early-on in the discussion when I suggested using a pair of 6SN7's with plates and grids of each tube tied together running in P/P with the other tube.  Since 6B4's are basically two plates and two grids tied together internally, I thought I could accomplish the same thing with 6SN7's with the connections external on the sockets.

After that we discussed substituting a phase inverter in place of the inter-stage iron.  I have the NOS UTC S-8 which would perfectly for that purpose.  That's where the last diagram with the 6N7's came in.  Back to square one?

Are you really going to build something?
And use it on what?

I ask because I have a UTC LS something or other, push pull plates to push pull grids at about the 20 watt level I have no use for.
I will get the number when I get home.
It would be nice if someone actually used this stuff...

Hi Mike

Okay on your original circuit. The 6N7 phase inverter looks cool eliminating the interstage xfmr. Any of those circuits will work. Fun to experiment. I believe the external Johnson speech amp for the Desk KW used a PP 6B4 audio driver. The 250-33?

Build one of them up and try it out with the 811a modulator.

Al VE3AJM

I've been planning on building an AM / CW / SSB (?) amp since forever and have been collecting parts just as long.  Then I came across a UTC S-22 mod iron so the final design is being based around it.  The speech amp / driver will be separate.  The mod deck shows the filament xfmr on the left and a 500 ohm line to class B grids Kenyon on the right as an alternative switchable input choice in addition to the modular driver.  That along with switching and a meter will be the only components on the mod deck. The modulator power supply will deliver either 1250 or 1000 VDC @ 300 MA under load. The 811A's will be toward the front and the iron in back.  Filament iron will probably be under the chassis on the right with the Kenyon top left. The SS rectifiers will be mounted directly on the chassis eliminating that assembly with the insulators which originally came from a Gates BC-1T tranny.  The RF power supply is similar also using Thordarson iron.  Both supplies will be on the bottom of the rack.

Mike

FB on the UTC iron Brett!  Bob, W4RFM, is working on a modulator built around the UTC S-22 with a driver based on an Altec A-333A power amp using a pair of 6L6's.  The output is around 20 watts so a driver iron of that rating would be perfect.

I dug up the full schematic for the Altec and it looks like a solution for my driver as well.  Exclusive of the output transformer, the amp is basically flat 20-20 KHz.

Of course, this doesn't include a mic pre-amp, but that is fine because I have that covered.  Looks like this is the way to go.  Problem solved.

Won't need the 6B4G's, but thanks for the info!

The 6b4's tested good, the trans is an LS 22.

Push pull plates to push pull grids, split primary and split secondary.
30K ohms plate to plate primary, 80K secondary, +28 dbm.
20 to 20K Hz.

Brett, I'm familiar with that iron.  I just wonder about matching.  I've decided not to go with the 6B4G's because the Altec A-333A amp Bob (W4RFM) is building using a pair of 6L6's makes a lot more sense.  That amp is rated at 20 watts, more than enough to drive a pair of 811A's.

Here is my problem.  The LS-22 would handle the power and the frequency response is out of the ballpark.  But would it match for the application?  I've attached the driver schematic.  Secondary would just feed a pair of 811A grids with a plate voltage of 1250 DC max.  Little or no bias.

Next problem would be the price.

If I can build a relatively hi-fi modulator around a UTC S-22, FB.  I've been planning this project for over 50 years.  A few days more won't matter.

I do not know, the ls 22 might do well reversed so its a step down.
Ideally you want something that matches the plate to plate load of the 6L6's and gives the needed voltage swing on the 811a grids.
The problem is the 811 grids are class B, and not a steady load.
To drive class B grids you want a very low plate resistance tube operating into a step down transformer so the load the grids hit on peaks do not effect the driver.
Tetrodes make poor drivers of class B triodes.

An 8 ohm solid state amp into an 8 ohm to 5000 ohm center tapped transformer works real well as the solid state amp makes MORE power if it works into a lower impedance load.

I am not sure how much distortion and peak power limiting you are willing to put up with, or how a pair of 6L6's as drivers for 811's will work. You do not have to sound like a broadcast station to sound good.

 

I did manage to score the Stancor A-4416 driver iron.  Primary is 3K to 10K CT at 40 MA / side to P/P class B grids.  I'm going to pass on the LS-22 but am interested in the 6B4 tubes.

$20.00 each plus $5.00 for shipping?

lots of discussion, very fun. 6B4s for cheapish prices, it is good. be sure to stock up some spare too.

1. The 6B4 (or 2A3, 6A3) can be driven by an RC coupled circuit and uses grid bias. The 6N7 is a zero bias dual section tube. For power amplifier applications the 6N7 will work in place of two 6B4s, up to the maximum power output allowed for the 6N7, which is 10W. The 6B4 pair can make 15W.

2. 6N7 in push pull is a zero bias class AB2 situation and the grids require a little current, so a transformer is needed to drive it instead of a RC coupled phase inverter. Due to this requirement the tube manual offers a simple class A Triode driver made of both sections of a 6N7. Alternately a small dual Triode could be cathode coupled to drive a 6N7, if the extra complexity is not minded.

3. That Altec circuit: You can use screen grid tubes for class B grid drivers well, but you need to shoot for a quite bigger amp that that 20W Altec, so that at its half-power level you can still swamp 50%+ of your drive in a resistance, which will improve the regulation. So a 4x size amp is a bare minimum*, if it is a regular screen grid tubes design like that Altec. 40-80W minimum to avoid distortion with that setup, based on 10W into the 811 drive xfmr, for generous losses. 40W cuts it close.

4.) All wattage ratings for screen grid tube amplifier apply to closely-specified nicely driven and polite little resistive loads, and are immediately un-useful for class B driver discussion. For discussion of class grid drive via push pull screen grid tube amplifiers, the output has to be considered in terms of:
a.) peak voltage and peak current at the crest of the audio half cycle.
b.) strict regulation of the drive voltage regardless of a current waveform that is not of a linear relationship but exponential
 b₁.) with added weirdness.
c.) Solid state will do well but that's the chicken way out.)


I use a Stromberg-Carlson 150W PP-PAR 6550 public address amp to drive a pair of 3-500Zs with 25W, When the drive to the 3-500 grids is 25W average during speech, the amp is making only about 50W average for 100% modulation (incl. 25W to the swamping load resistor), and is not called upon for its full current except when the grid current and voltage together reach their peak at the crest of the half cycle. This reserve has the effect of driving from a very well regulated source. It is a brutish way to do it but very simple.*

No you do not need to sound like a broadcast station, but you must be free from amplitude distortion over the entire half-cycle to sound clean and clear and natural.
Choice:sound like space shuttle or sound like speaking in the room.
I prefer to use a switch to select the space shuttle sound (as desired) and not be limited to it by design. Especially if spending precious time building something.

The proof of this advice for those who use screen grid tubes to drive class B grids is what my transmitter sounds like. ;D Ask anyone who has heard my transmitter on 7160 what it sounds like or more specifically if it sounds clear and natural and easy to understand every word. Some of them are on this board.  ;D

I am way too smug in my audio, but worked a lot on it to learn and verify these truths, which are not new but are all implied in the Ancient Volumes.

*power reserve can be reduced to about 2x with a custom class B driver transformer, by using tubes with extreme current reserve and a special cathode feedback circuit to lower the output Z as shown here: http://www.montagar.com/~patj/aph1050.htm

Lastly, for a simple package modulator, the standard Altec 1570B circuit is an extremely simple, cheap, and well proven way to deal with a pair of 811's. It is only 175W, but in experiments I bypassed the filter choke for C input and obtained 255W by raising the plate volts to 1200. manual: http://www.dvq.com/hifi/images/1570a.pdf

Yes, lots of clean audio to dump on the carrier is not so easy to come by.
Class B triodes seem easy, and they are with solid state amps and good iron, not so easy with all tube drive.

AB1 is great, like my 4x150 mod deck, just voltage needed to drive the grids, but you need bias adjustment for each tube, stiffly regulated screen voltage, and a variable plate supply to closely match the load to the modulator.
Can do 600 watts, blowers needed for the tubes.

Screen modulation can be great (no iron, phase shift or frequency response limitations) but it takes big tube blowing off a lot of heat to get big power output.

Why can we not have our cake and eat it too?

I have tried them all, and I think the AB1 modulator sounds better then the class B, and the screen modulators sound best with the right tubes.

Two or three 4-400's screen modulated would be ideal I think.
About 500 watts carrier.

I guess it depends on what you mean by a big tube. A pair of 828s in AB1 will produce over 300 watts in AB1. No air required. These tubes were designed for AB1 audio service and feature low idling current (IIRC, 50 mA per tube) and low distortion. I had a GE BC TX that used a pair of these as modulators. I had never heard of 828s before I got that TX. They are quite unknown in ham circles. The down side is that they are not widely available. Too bad. The 828s are an excellent tube.

I think the power problem could be satisfied if I could score a Stancor A-4763 driver iron.  It's 30 watts and would probably match anything.  The only solid-state amp I have is a Realistic STA-77 receiver and it's good for 16-22 watts per channel output.  I certainly could run it into a hi-fi plate to VC transformer which would probably work.  For now, I'll put the modulator deck together and experiment.  I'm not even going to mount the Kenyon line to grid transformer on the chassis...which is my alternate driver.  I'll just stick a 3-terminal barrier strip on back for the drive input and play with it.  Sigh.

Meanwhile Brett, what do you have for tubes that would work for drivers besides the 6B4's?  At this point 6N7's, and 6L6's are in the running.  Do you have any 811A's?  I have one and need another for certain...maybe a matched pair?

Oh,and as far as broadcast quality audio is concerned, I was a broadcaster for over 35 years and great audio habits are hard to break!  Then there is the recording studio thing.

Mike

You're saying that this isn't your first rodeo?  ;)

Nah. :)

I have some 6n7's, 6sn7,s (lots of those, new), lots of 6080's, and all sorts of hifi tubes, kt88, kt90's, 6ca7's, and all sorts of other tubes.

Audio amplifiers are dirt cheap, all you need is 20 or 30 watts and the right transformer, Hammond sells good ones.
A PA amp will even have a mic input/preamp, there are amps on ebay very cheap, kits, and pawn shops sell equipment as well as music stores.
I use a symetrix 20 watt per channel one rack unit high amp I got at a pawn shop cheap, its a bit light in power, but its small and its been working for the last 25 years or so.
A radio shack PA will even work, and likely be cleaner then anything you could build using tubes without a lot of effort and money.
Its fun to build, but I have never been lucky with building a really clean, quiet HIFI audio amplifier that takes a D104 input and gives 300 to 600 clean watts out using all tubes.
I did build a really great sounding 2 tube screen modulator, 12ax7 and a 6de7.
It modulates a pair of 4x150's at 100 watts carrier output and sounds fantastic.
That is the best all tube audio I ever built.

Much depends on if you are looking for the best sound, the easy fast way, or the very hard, complex all home brew all tube way.

In my book, almost everyone on the air sounds quite good.
Some ssb type rigs sound very nasty, some sound good, but most vintage, homebrew, sdr and class e rigs sound very good.
If you want really fantastic audio on the air, there are about 3 ways to do it, sdr, a modified low level modulator into an amplifier and class E.
A broadcast rig will do on 160 and 80 if you have a lot of space in the garage and own a fork lift.
 

I checked out some PA amps and they certainly are going for cheap!  I take it a plate to voice coil transformer would be used in reverse as a driver?  I've actually done this before for screen modulation.  I know I have a 70 volt transformer lurking around somewhere, I assume using that 70 volt tap would work also?  As for the tubes Brett, I have to dig through my jonque boxe and see what I need.

For now I'm going to build the modulator deck first then decide how to drive the bugger.  Did I mention I will need at least 1, but preferably a matched pair of 811A's?

I never bothered about matched pairs in modulator service.
I am not sure that is important.

I have 1 x 811A Brett so I need at least 1 more.  If I had a larger driver iron, the 6B4's might be desirable so I'm putting a hold on that.  I would like to score some 6N7's.  I'm actually building two modulators; the 200 watt and a 50 watt using a UTC S-20.  That will have either 6L6's or 5933's driven by a 6SN7 using the Stancor A-4416.

I use 811's and have some Russian ones, they work but the base and plate cap glue was no good.
I have some old US made 811's (not A's) and some 812's and 812A's and could actually use some good US 811a's myself.
6n7's I never used and maybe I can send you some if you are going to use them.

Hammond sells good transformers, 8 ohm to whatever, 5000 or 10,000 ohm center tapped, 15 or 20 watts would work very well for 811 grids, and if you wanted to, you can get a 20 watt IC chip to drive it, or any old solid state audio amplifier.

Heck, am I dumb or what?  Sometimes the obvious doesn't register. If I don't run into problems with RFI, I could use my studio amp as a driver!  It's a big honkin' Yamaha home theater amp with DSP.  I'm running around 175 watts/channel stereo into the main speakers BUT it also has outputs for surround speakers at 75 watts a pop.  That pix is old and things have been updated including a pair of 15" MTX speakers. The ham shack is in the studio to the left front of the big mixer. That would rock!  After all, I've been wanting to integrate the two for years. 1,000 lashes with a wet spaghetti noodle.

The reversed output transformer may work. W5FRS uses one rated 20W to drive 808 grids modulating an HK54. The difference is the 808 takes negative bias and less grid current than the 811 under similar conditions.

70V transformer, does it have a CT? What peak voltage is needed? It could work if the voltage is adequate.

Opcom, forget the 70 volts.  I'm thinking driving the 811A's through an 8 Ohm iron might be the way to go since so many of you seem to favor the "old buzzard" approach!

I do not think I would run a 175 watt amp into a 20 watt transformer.
If you just want to get on the air with some good audio, the 8 ohm to 5000 ohm center tapped to the grids of 811a's is a good and easy way to go.
Run a vx2000 or some such cheap mic processor into a small audio amp and away you go.
The S22 transformer will be more of a limitation then the audio equipment.

If you like building, an all tube setup would be fun, but might not work as well.
It takes a LOT of parts and much time to build.

Brett, of course not, but I could run one of the 75 watt outputs into the transformer at reduced volume.  The Yamaha has multiple outputs. The amp would be less than loafing.  It would be dozing!  I have more than enough parts to build anything but until I get my cataract surgery done, I'm literally shooting in the dark where construction is concerned!

I used a single 6L6 with a 25 mod xfmr to drive 811As.  I used the primary for the 811 grids and the stepped-impedance secondary for the plate load of the 6L6.  I think (IIRC) the stepped secondary was 3K, 5K and 10K.  I used the 5K for the 6L6 and the 10K for feedback to the 12AU7 that drives the 6L6.  I used a lot of feedback so it takes about 5 volts RMS to drive the 12AU7.  My audio stages before the 12AU7 were able to supply that much audio.

Seems to work well.  I designed the driver using 100-200 ohm load on the driver xfmr in place of the 811 grids.  I think the ratio was a step-down to the 811 grids from the 6L6 plate.  Using a lot of feedback lowers the plate resistance of the 6L6 which what you need when driving class B grids.

Fred

I was wondering about that concept Fred.  I also have a 50 watt modulator in the works using an NOS UTC S-20.  Could that be used as a driver?  Imagine a modulator driving a modulator.

May not have a low enough secondary impedance to drive the 811 grids.  It may work, you would have try it using a step-down ratio going to the 811 grids.  A S-20 is a universal type mod xfmr.  It has four windings, each with a tap.  The taps are not in the center of the windings, they're offset from center.  Maybe using the taps from center (both secondary windings in series) to drive the 811 grids.  May work, you'll have to try it.

Fred

That's what I was thinking.  Impedance on either side of the tied-together  terminals would not be the same on either side.  It was just an idea.  Frankly I've tracked down an NOS Stancor A-4763 30 watt driver iron cheap. I probably will pick that up and build around it.