The AM Forum
March 28, 2024, 06:59:57 PM *
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: Class B 811A Modulator driver  (Read 13044 times)
0 Members and 1 Guest are viewing this topic.
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« on: June 05, 2020, 10:27:32 AM »

I have a nice 350 watt modulation transformer and am thinking on building a class B modulator.

I have all of the parts needed except the driver transformer and these seem to be very hard to find.

I have seen MOSFET drivers discussed and a reversed audio output transformer discussed driven with a solid state high Fi amp.

Any suggestions on finding a transformer or other best practice driving method?

Also looking at running a couple 807s in class B for a smaller modulator and have the same problem. looking at 120 watts out. Is a Hammond 124 series 3:1 driver suitable? It looks like it may not have the power rating needed.(4 watts)

Thanks in advance.


Pat
N4LTA
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


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


WWW
« Reply #1 on: June 05, 2020, 02:10:24 PM »

A transformer is absolutely not required.

There are three methods to drive directly with cathode drive.
You'll find info on that by looking at tube audio amplifiers, especially
811a tubes. That plus "cathode drive".

Before the driver, of course you need a phase splitter.
That can be a transformer, or other circuits.

Depending on the B+ voltage required, a bias supply may not be needed.
811a tubes are nominally a "zero bias" tube, class B... the driver ought
to provide power to the grids, not be cap coupled for best results, imho...
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« Reply #2 on: June 05, 2020, 03:36:19 PM »

Sorry

I am building a class B Plate modulator using two 811As.  I can't find the proper driver transformer and am looking for an alternate mean for driving the 811A grids.

I have seen a MOSFET circuit used and described in a past post, and have most of the parts I think. Some of the stuff that I had put together was destroyed in the shop fire.

Is there an updated circuit or circuit board layout available are driver transformers for driving class B grids available - or a phase inverter circuit?

Pat
N4LTA
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #3 on: June 05, 2020, 03:47:34 PM »

Hi Pat,

I just built one of the GFZ MOSFET audio drivers for the 813 rig. My second one. No problems building it and testing.   I should take some pics of it and post.


Take a look at John, W9JSW's 813 thread and you will see all the latest board and circuit info.


There were about 20 boards sold and I think the majority did not build them yet - so there are probably some for sale on this BB. This board will eliminate all transformers except for the modulation iron, which is a good thing for reducing phase distortion and the ability to implement NFB. These finer details all add up in the end.

The other alternative is to find a 20 watt or more HI-FI (20Hz - 20K type)  tube-type audio output transformer on eBay and drive it with a good 8 ohm class A/B audio amp.   (8 ohms to 3K C.T. grids or whatever)  I see them from time to time for sale, but may cost as much as building a GFZ board in the end.  (add in the SS amplifier cost too)


T

Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
WBear2GCR
Member

Offline Offline

Posts: 4135


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


WWW
« Reply #4 on: June 06, 2020, 09:08:51 AM »

Sorry

I am building a class B Plate modulator using two 811As.  I can't find the proper driver transformer and am looking for an alternate mean for driving the 811A grids.

I have seen a MOSFET circuit used and described in a past post, and have most of the parts I think. Some of the stuff that I had put together was destroyed in the shop fire.

Is there an updated circuit or circuit board layout available are driver transformers for driving class B grids available - or a phase inverter circuit?

Pat
N4LTA

Pat,  the Mosfet is fine. If you want that.

If you want tubes, as I said, you don't need a transformer.
Search out the Gates modulator that worked without a transformer...
Also look for the schematic of an Altec 1570B.

Either method will work nicely, easy to scale in the case of the Gates, the
Altec requires no modification.

Those are two of the three methods.
The third is a direct connection to the grids of the 811s.

As mentioned a little search for something like "direct cathode drive schematic"
or similar ought to yield links to various pages with examples...

                  _-_-

Attached is an example of the Gates scheme - I've seen a simpler Gates shcematic, this
one is more complex on the surface, but the basic scheme is the same. Can't find the
nice schematic. It might be in this forum, as cathode drive has been discussed before.
Maybe a good idea to search.

Ok, so I put up the Altec one too... ask questions, if you want... parts are dead simple.

* Gates_BC-1G_schematic.pdf (4375.33 KB - downloaded 357 times.)
* 1570B.pdf (628.85 KB - downloaded 311 times.)
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
WBear2GCR
Member

Offline Offline

Posts: 4135


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


WWW
« Reply #5 on: June 06, 2020, 09:31:38 AM »

I must be in a good mood...

http://amfone.net/Amforum/index.php?action=dlattach;topic=25569.0;attach=22374


An example of the resistive loaded method (not the best, but Gates did this also, and it
does work.)

Good thread too!

http://amfone.net/Amforum/index.php?topic=25569.0

Wow, that was 10 years ago now!!
Good ol' Rip Van Winkle at work again...

              _-_-bear


EDIT: and copied from another post, this simplified version of a direct coupled driver.
https://www.qsl.net/wa5bxo/driver1.html
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2722



« Reply #6 on: June 06, 2020, 11:14:31 AM »

If you can find a schematic of the Clegg Zeus, you can see a pair of 811A modulators driven by a push-pull cathode follower (I forget the tube type used) set up.
Logged
KD1SH
Member

Online Online

Posts: 738



« Reply #7 on: June 06, 2020, 08:31:05 PM »

12AT7 phase inverter driving a 6BX7 cathode follower driver.  Love my Zeus!  One of the few commercially made plate modulated ham rigs where the modulator was designed with a huge surplus of audio power.  A pair of 811's modulating a 120 watt carrier from a 4X150 PA.


If you can find a schematic of the Clegg Zeus, you can see a pair of 811A modulators driven by a push-pull cathode follower (I forget the tube type used) set up.

Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
—Robin, in the 1960's Batman TV series.
W2PFY
Contributing
Member
*
Offline Offline

Posts: 13312



« Reply #8 on: June 08, 2020, 12:13:31 PM »

 1. Here is a copy of the Clegg Zeus transmitter & modulator
 
 2. Improving the class B driver. Look up the author on Google, he was a very interesting man.

 3. Home brew cathode follower by W1FRM GUY. I actually have used this cathode follower pictured in his description (built by Guy) and it worked just fine! If I were to build it, I would use 807's in place of the 6L6 tubes just for the look of it. Same tube except the 807 has the anode coming out the top of the tube He built it to drive a Johnson 500 modulator with 572B Yubes in it.

* W1FRM CATHODE MODULATOR DRIVER.pdf (273.59 KB - downloaded 368 times.)
* CLEGG ZEUS MANUAL.pdf (3787.96 KB - downloaded 441 times.)
* Improving_the_Class_B_Driver(1).pdf (3551.25 KB - downloaded 374 times.)
Logged

The secrecy of my job prevents me from knowing what I am doing.
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« Reply #9 on: June 09, 2020, 08:30:20 PM »

Thank all for all the info. Very interesting stuff.

Looking forward to getting back to building something. Hope to do some shop layout with my builder tomorrow  (my son) He has the basic footprint done  adding a couple linear feet in each direction. I need to put the interior walls in. May get a new foundation poured next week and some brick work started. After that I hope framing is fast. May try to put a 20 meter ground plane at the roof peak and the ground plane in the attic. Also lots of PVC conduit to get coax outside and to the roof.

Pat
N4LTA
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


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


WWW
« Reply #10 on: June 10, 2020, 07:44:39 AM »

Guy's method is essentially the Gates method, resistors in the cathode of the driver.

It's ok, but wastes power unnecessarily in the resistors. The choke is slicker. Cheesy

The other aspect is that depending on the B+ on the 811s, they can be run "zero bias".
Above a certain B+ voltage some negative bias may be needed. The choke puts the grids
on DC ground, no bias supply needed.

If you use something like a 22k resistor it adds a Vdrop on the cathode of the driver tube,
which would lift the drive point up in voltage, when one wants down in voltage... 

One only needs a 6W4 as a driver, no more than a 5 watt tube. An 807 is major
overkill. WAY too much power there. A 6V6 is much bigger than required, for example.

Note the small dual triode in the Clegg?

Cool stuff...






            _-_-
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
W2PFY
Contributing
Member
*
Offline Offline

Posts: 13312



« Reply #11 on: June 10, 2020, 01:19:21 PM »

Quote
One only needs a 6W4 as a driver, no more than a 5 watt tube. An 807 is major
overkill. WAY too much power there. A 6V6 is much bigger than required, for example.

That's true but if you wanted to swing the grids on a couple 833A's, 807's would be happy. The circuit would need to be tweaked a bit and some supply voltages may have to be changed fer the 833's. But about 807, my thought is not about what is practical but rather what looks cool to me. At age 79 I want it my way:)
Logged

The secrecy of my job prevents me from knowing what I am doing.
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #12 on: June 10, 2020, 10:59:54 PM »

I think you guys are referring to the 6W6. The 6W4 as I recall is a damper rectifier tube. I use a 6W6 in a HV regulator, and a isolated 5V winding to power the filament. Works very well as a series regulator.

I attach three files. One is for the Millet A2 Driver PCB, and the other is a work in progress that I m installing into a Viking II. You can buy the Millet PCB on Ebay.

The setup I show can swing each output to 140v peak, or 280V P-P when the B+ is 350v. The low source impedance of the FET source followers will provide ample grid current. The 4500 ohm resistor feeding the Millet PCB would need to drop in size to about 1K 10W if driving 811A's. The FET's and CCS's need heatsinking.

The ECC99 tube is a 12BH7 on steroids. Bigger cathode, Plate, and much higher Gm. The circuit as shown will work with a 12AU7, 12BH7, 6CG7 (filament rewire) just fine, but the ECC99 achieves a greater P-P output and higher gain. The CCS on the cathodes sets the operating point, so no adjustment is necessary.

The modulated envelope attachment is from the Viking II using 6DQ5's as modulators, and driving G2 (instead of G1) with the Millet A2 board.

Jim
Wd5JKO


* V2_Audio_Driver.jpg (561.69 KB, 3510x2550 - viewed 731 times.)

* 100%_Mod_SINE WAVE.jpg (241.45 KB, 4160x3120 - viewed 319 times.)
* Millet A2 Driver Notes.pdf (785.65 KB - downloaded 241 times.)
Logged
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2722



« Reply #13 on: June 11, 2020, 10:35:29 PM »

Good stuff Terry. Thanks for sharing.


1. Here is a copy of the Clegg Zeus transmitter & modulator
 
 2. Improving the class B driver. Look up the author on Google, he was a very interesting man.

 3. Home brew cathode follower by W1FRM GUY. I actually have used this cathode follower pictured in his description (built by Guy) and it worked just fine! If I were to build it, I would use 807's in place of the 6L6 tubes just for the look of it. Same tube except the 807 has the anode coming out the top of the tube He built it to drive a Johnson 500 modulator with 572B Yubes in it.
Logged
KD1SH
Member

Online Online

Posts: 738



« Reply #14 on: June 12, 2020, 09:21:42 AM »

   Very interesting topic and information, especially since I've been considering modifying my B&W 5100B - which has a bad driver transformer - to use a phase inverter/cathode follower direct drive setup.  My question is this: since the B&W modulator was designed to run the 6146 modulator tubes in class AB1, which means that, in theory, there's no grid current, couldn't you just take the output of the driver tubes right off the plate circuits and capacitor couple it to the 6146 modulator grids, rather than run the driver tubes as cathode followers?
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
—Robin, in the 1960's Batman TV series.
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #15 on: June 12, 2020, 11:13:52 AM »

   Very interesting topic and information, especially since I've been considering modifying my B&W 5100B - which has a bad driver transformer - to use a phase inverter/cathode follower direct drive setup.  My question is this: since the B&W modulator was designed to run the 6146 modulator tubes in class AB1, which means that, in theory, there's no grid current, couldn't you just take the output of the driver tubes right off the plate circuits and capacitor couple it to the 6146 modulator grids, rather than run the driver tubes as cathode followers?
   
   Running in class Ab1 does not mean there is no grid current. On many multi grid power tubes, some grid current occurs as the grid 1 peak swing approaches zero volts. Then there is the plate to grid 1 capacitance which gets multiplied by the gain of the stage due to Miller capacitance.

https://www.aikenamps.com/index.php/what-is-miller-capacitance

    Cin = Cgk + Cgp*(A+1)

where: Cin = input capacitance
           Cgk = grid-to-cathode capacitance, composed of the internal tube capacitance plus the stray capacitance
           Cgp = grid-to-plate capacitance, composed of the internal tube capacitance plus the stray capacitance
             A  = stage gain

  Therefore, a low impedance driver is desirable, even when driving big power grid tubes in class Ab1...

Now consider what happens when an audio peak on the driver tries to extend into Class Ab2, and grid current is drawn. The R-C between the audio coupling capacitor, and the output stage grid resistor will cause a bias shift that could last 10ms or longer depending on the RC time constant. On a good amplifier, recovery time from overload should be minimal, or zero. A prolonged overload following a big voice peak can sound really bad.

  My post that had sand in the signal path seemed to at best receive a groan, so I attach another schematic using 6146's with a RC coupled Tube driver. This uses several 6S4's which are a dirt cheap power triode. Notice the plate load resistors are 40K. Imagine how the Miller capacitance would destroy the driver if the plate load was 470K. Realize though that the 6146 grid bias is regulated, as is the 6146 screen grids. This should work very well, but if you let those voltages wiggle with audio, the results will be degraded., and the bias point will shift as the line voltage varies.

  The mention of the ECC99 built by by J&J is a new tube design, and in my opinion based upon my tests is a superior tube. Could use two of them instead of the 6S4' in the attached schematic. Not cheap though, around $20 apiece +/-.

Jim
Wd5JKO


* Triad HF-40.jpg (83.96 KB, 773x398 - viewed 574 times.)
Logged
KD1SH
Member

Online Online

Posts: 738



« Reply #16 on: June 12, 2020, 12:21:45 PM »

Thanks - I like that circuit.  Yeah, I've noticed that with a few Ab1 6146 amps I've been able to see bit of grid current using my scope probes in differential mode across the series resistor.  And, I figure that unless I'm using some pretty tight limiting, peaks are going to get through and result in grid current.  My original concept was to use a pair of 6AQ5's in conventional grounded cathode, downstream of a phase inverter, but I was thinking that the lower impedance of a cathode follower would give me more elbow room and less worry about driving the 6146's into class B.  I like the 6S4, nice tube.
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
—Robin, in the 1960's Batman TV series.
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #17 on: June 12, 2020, 02:03:50 PM »

Thanks - I like that circuit.  Yeah, I've noticed that with a few Ab1 6146 amps I've been able to see bit of grid current using my scope probes in differential mode across the series resistor.  And, I figure that unless I'm using some pretty tight limiting, peaks are going to get through and result in grid current.  My original concept was to use a pair of 6AQ5's in conventional grounded cathode, downstream of a phase inverter, but I was thinking that the lower impedance of a cathode follower would give me more elbow room and less worry about driving the 6146's into class B.  I like the 6S4, nice tube.

Here is another audio setup that carries things to a higher level. The center tapped choke (L1) could be the primary of a small audio output transformer. The center tapped choke idea increases the output swing, and lowers the driving impedance.

http://www.turneraudio.com.au/80W-AB1-amp-2014.htm

Jim
Wd5JKO
Logged
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2722



« Reply #18 on: June 12, 2020, 02:30:20 PM »

"Running in class AB1 does not mean there is no grid current."

I think that's exactly what AB1 means - no grid current. Yes, a tube can go from AB1 to AB2 (grid current) with just a bit more drive. For many amateur grade modulators, it probably doesn't matter.
Logged
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #19 on: June 12, 2020, 03:12:18 PM »

"Running in class AB1 does not mean there is no grid current."

I think that's exactly what AB1 means - no grid current. Yes, a tube can go from AB1 to AB2 (grid current) with just a bit more drive. For many amateur grade modulators, it probably doesn't matter.

  Steve,

   Ideally what you say is exactly true. That said, many bigger power tubes such as 6L6, KT88, EL34, 7591, etc have a Max spec on grid resistor size when running on fixed bias. Some can go as high as 470K grid to bias tap, and others maybe 100K maximum. This suggests to me that some sort of low level grid current is flowing (no signal), and a runaway condition can occur. The solution is to either lower the Rg value to a lower amount, or switch to cathode bias. A lower Rg adds more loading to the driver stage.

   With this situation occurring, the grid could already be a few volts higher (more +) than the bias at the bias tap, or pot wiper, and then the crossover for the drive to cause grid current comes premature. This likely happens to tubes with a little gas inside.

   Then we have the driver loaded with the output tube Miller capacitance, and this causes an elliptical load on the driver stage. It takes current to charge and discharge a capacitor. You cannot drive a 4-1000 G1 in Ab1 with a 6AU6!

Jim
Wd5JKO
Logged
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2722



« Reply #20 on: June 12, 2020, 05:01:50 PM »

I completely agree!
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


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


WWW
« Reply #21 on: June 13, 2020, 09:06:03 AM »


That effect, the cap causing grid current to flow, is a cause of some audio tube amps
making the output tubes into toast.

So, not a good thing to want.

Once there is a positive bias current created, in the manner described, IF there
isn't sufficient means to discharge the coupling cap and get rid of the positive bias,
what happens is as the tube ages and the bias point shifts (taking MORE negative
bias to cut off) the tube tends to run more current. So, the short term momentary
bias causes excess peak current, and the tube's ability to bias off gets worse.
This leads eventually to run-away, and the tube cherry's out. Especially a problem
with "self-bias"/cathode biased circuits.

There are several high priced, high end audio amps that suffered from this issue.
Many expensive NOS 6550s turned to junk.

Patrick Turner's 807 amp looks like a choke fed driver stage. He's using a single
choke per tube fed from the CT to balance out the effect of the DC through the choke.
Which is fine.

But, imho, in that schematic, unless one needs the gain, it's better to use the same choke in the
cathodes of the driver tubes and run in Class 2. The use of a small output transformer's
primary for the "choke" was something I suggested earlier... Although it could
just as well be a power transformer's CT winding, as long as the inductance is high
enough. A bit depends on the gain required, as the cathode drive stage could add
a pair of tubes - in a commercial design, it would have been extra expense. But in
ham use? Not a worry.
Logged

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

Offline Offline

Posts: 1996


WD5JKO


« Reply #22 on: June 14, 2020, 12:04:26 PM »


That effect, the cap causing grid current to flow, is a cause of some audio tube amps
making the output tubes into toast.

So, not a good thing to want.

Once there is a positive bias current created, in the manner described, IF there
isn't sufficient means to discharge the coupling cap and get rid of the positive bias,
what happens is as the tube ages and the bias point shifts (taking MORE negative
bias to cut off) the tube tends to run more current. So, the short term momentary
bias causes excess peak current, and the tube's ability to bias off gets worse.
This leads eventually to run-away, and the tube cherry's out. Especially a problem
with "self-bias"/cathode biased circuits.

There are several high priced, high end audio amps that suffered from this issue.
Many expensive NOS 6550s turned to junk.

Patrick Turner's 807 amp looks like a choke fed driver stage. He's using a single
choke per tube fed from the CT to balance out the effect of the DC through the choke.
Which is fine.

But, imho, in that schematic, unless one needs the gain, it's better to use the same choke in the
cathodes of the driver tubes and run in Class 2. The use of a small output transformer's
primary for the "choke" was something I suggested earlier... Although it could
just as well be a power transformer's CT winding, as long as the inductance is high
enough. A bit depends on the gain required, as the cathode drive stage could add
a pair of tubes - in a commercial design, it would have been extra expense. But in
ham use? Not a worry.

Bear,

   I attach an Altec design that uses a pair of triode connected 6W6's to drive a pair of 811A's They use a center tapped choke on the cathode side as you describe.

If you are into a bunch of tubes in the audio chain, then there are many circuits out there. I still like the Millet A2 board for it's functionality, and simplicity.

Jim
Wd5JKO


* al1570b.gif (117.09 KB, 3285x2360 - viewed 536 times.)
Logged
WU2D
Contributing
Member
*
Offline Offline

Posts: 1800


CW is just a narrower version of AM


« Reply #23 on: June 14, 2020, 01:38:21 PM »

"Also looking at running a couple 807s in class B for a smaller modulator and have the same problem. looking at 120 watts out. Is a Hammond 124 series 3:1 driver suitable? It looks like it may not have the power rating needed.(4 watts)"

Oh on the hammond 124D - that is a 5 Watt transformer and it would be great for a push pull driver.

You are right about finding driver transformers as called out by Collins, Stancor and the like - pretty rare.

But small modulation transformers can be used. I have used old TCS 1:1 CT 20W transformers and Johnson Mobile Modulation transformers as driver transformers. I made a 1625 modulator driven with a 6SN7 into a TCS modulation transformer as the driver and for the modulation Iron itself, one of those $6.95 jobs from an old 618T, all military iron.  Feedback straightened the response out nicely. Pretty much anything from 1:1 to 3:1 can be made to work and if you do push pull, 4:1 and 5:1... 

Logged

These are the good old days of AM
WBear2GCR
Member

Offline Offline

Posts: 4135


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


WWW
« Reply #24 on: June 14, 2020, 09:25:38 PM »

Jim, I already posted the 1570 schematic... iirc. Roll Eyes
But yes, that is a nifty way to do it.

The best part is that a choke doesn't need to be "good" in the same
way that a transformer needs to pass high freqs and go low...

I own a pair of those amps! Cheesy

They be good.

         _-_-


Other topic: Last time I looked a pair of 807s is NOT going to make anything like
120watts... a quad, ok. Did we talk about that before now? thought it was mentioned.

EDIT: The 1948 ARRL Handbook, I happened to grab - it does show a pair of 807s @750vdc
claiming 120watts, into a specific load. At CCS they call for 80 watts. I've seen 807s do
100watts in an audio amp @ ~750volts, AB2, but they were really pushing the tubes.
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
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 0.152 seconds with 19 queries.