Class B, zero bias - satisfactory circuit arrangement?

[KC2ZFA]

What to do, seems still up in the air.

how about this:

[WBear2GCR]

This amp is a fixed screen bias design, with the driver providing output from the cathode, using a resistor to ground.

It is not the same as was talked about before which has the drive on the screen, and a dropping resistor to keep current down on the grid... in essence driving both the grid and the screen... making it a triode...

Nice amp though.

                _-_-bear

[Steve - WB3HUZ]

I heard W5PYT use the term shark-fin back in the 1950's.

[WBear2GCR]

da QIXmon wrote:

Transformers are linear resistive elements, albiet not large resistances. They are imperfect in their ability to transer energy.  So, when the grid(s) go from an infinite impedance (negative voltage) to some finite resistance (being driven positively), there is a significant change in the load as seen by the transformer & driver, and this will result in a slope change on the driving signal (due to voltage drop) and this means DISTORTION.  Negative feedback will help, but nothing will compare to a source or cathode follower driver.

Also, transformers introduce phase shift - and this in turn limits your ability to put negative feedback around the whole modulator (and include the modulation transformer in the loop).

Not sure what ur thinking about Steve, but transformers are not linear nor really resistive? They have hysteresis for starters... resistors don't. Then as you say they have phase shift...

maybe you meant to write "non-linear"??

                 _-_-bear

[WA1GFZ]

Wow HUZ I thought you were born in the 60s

[Steve - WB3HUZ]

Hatched.

These were on old recordings.

Wow HUZ I thought you were born in the 60s

[steve_qix]

What to do, seems still up in the air.

how about this:

That's a lot better design.  That .02 into the 470k may be a little small, and some of the other coupling caps may be a little small (I didn't go through the whole thing with a fine tooth comb), but it is far superior to the other circuit, hands down !

AB2 won't give quite as much power as triode connected, but it should be sufficient.

The modulator output should be de-rated to account for positive peaks greater than 100%.  The extra power required is rather significant as you get up there...  But I'd use this circuit over the other one any day!

[k4kyv]

I have a question: 400W PEP is my limit and that is power measured out at the antenna terminals. What is 400W PEP in terms of average DC power input to the plate?

So the UK has adopted that PEP rubbish too?

[The Slab Bacon]

I have a question: 400W PEP is my limit and that is power measured out at the antenna terminals. What is 400W PEP in terms of average DC power input to the plate?

(Could be measuring between apples and oranges here). Not terribly important issue.

But anyway, the old set was made to run 150W into the plate, and that measure was the legal limit then.  So, I've basically got a transformer to supply 150W to the plate.  And for some reason an 813 was chosen as the PA.

400w pep output is basically 100w carrier output power at 100% modulation. Which is your basic 100w table top boatanchor. (DX-100, Viking 1 or 2, Heath Apache, etc)

As far as the DC input goes to get that output, it is strictly dependant on what your plate efficiency is. (It usually winds up somewhere around 50%+ so you would need somewhere around 100w of audio for 100% modulation.

The general rule of thumb has always been whatever your carrier output is, you should have about the same (or a little more for good luck) in modulating audio for plate modulated class C AM service.

[Opcom]

What to do, seems still up in the air.

how about this:

That has also been shown as a concept in one of the handbooks (editors& engineers, ARRL, or RSGB) with 750V on the plates making 120W ICAS. I hate that I missed the old 1950's RSGB handbook I used to check out of the school library in my youth. I found a copy, but passed it up, was tired. regrettably.

[G8VOQ]

I've now added the modulator power supply schematic:

http://homepage.ntlworld.com/richards_internet/RADIO/zerobiasmodulator.html

The thing is I'm comming from a certain angle. Which is, that I'm trying to utilise the components to hand from an old AM set.  I'm trying not to have to buy new transformers, or tubes - if I can help it. But, I realise taking that tack, my AM rig might not be the best. But hopefully it will provide pretty good audio.

So, what do we have? I have one transformer available for the modulator. It supplied two  HT voltages. Apparantly 300V for the speech amplifier and driver, and 750V for the finals, which were KT88's in class B, zero bias arrangement. The modulator is touted as a 100W modulator modulating an 813, that was to have 150W into the plate (1000V x 150mA). The UK limit was 150W DC to the PA final. That might be the same as todays 400W PEP limit we have in the UK.

Of course the PSU is simple, there is no separate supply for screen grids to the finals, or independent bias supply.

Given T1, the transformer, it suggested to me that perhaps class B zero bias should be kept. Trying as I am to utilise T1 and if I can help it avoid buying any more "iron". I guess though T1 may not preclude changing from a triode connection (hi mu) to a tetrode connection, but is that sensibe trying to utilise T1 only. (BTW, I really should I think use 807's instead of KT88's).

So, what to do, revolves a lot around T1. And perhaps, but I'm not sure, that suggests keeping class B, zero bias.

But, the thought was, if class B, zero bias is kept, does it produce decent enough audio, or can it be made to produce good audio. I think it's been said that I should not keep those 15K resistors between g1 and g2, but employ zener diodes.

You see, someone might say, if I want to base my modulator exlusively on T1, my best bet is to stick to class B zero bias, but improve the drive circuit and the g1 g2 connection.

Of course, I don't think I'd have much choice but to keep the interstage transfomer. A point of weakness.

But, the question would be, could you actually make a better modulator using tetrode connection - using T1? I could get rid of the interstage transformer then.

If you could make a better modulator class B, zero bias would be abandoned, but at the moment it's not determined whether a better modulator would result using 807's in say AB1 or AB2 - employing T1 only.

It's a bit involved, but I think you get the picture. :c)

[WD5JKO]

R:

   I can certainly appreciate the concept of designing from the junkbox. With some creativity you can make some pretty interesting circuits based upon what you have. If you had more funds, and access to vintage parts you might do different.

  Your circuit as last posted make me reflect on the similarity to your modulator to the Gonset G76 transceiver. In the Gonset we use a 6CM6 (similar to a 6AQ5) to drive via a transformer the screen grids of a pair of 6DQ6 modulator tubes. What you have is essentially a scaled up version of the same thing. The one big difference is the the sweep tubes (mostly) don't need a lot of swing on G2 to make power like a 807/KT88. So here they ground the control grid G1, and only drive the screens (G2). This works extremely well, and efficient. This gets around the 15K resistors from G2 to G1, and therefore the 'Kink' in the transfer curve.

   With the Gonset the limitation is clean audio drive. Like your 6L6, the Gonset drive has NFB to the prior stage with a tapped resistor off the driver tube plate. With the Gonset, the distortion was seen in the G2 drive signal. So how good is it by just cleaning up the drive? I had to do three things to the 6CM6 audio driver. The items below use component numbers from your schematic:

1.) Take the audio pre-driver plate load R11, and increase the % NFB. Look at your circuit, take R14 (top), clip it, and pull down to R15 (bottom).  This puts R14 & R15 in parallel, and essentially adds 6DB more NFB. This will also drop your gain.

2.) Increase C10 cap to ~ 250 uf

3.) Look at lowering R16 to 270 ohms (so long as interstage tranny can take more primary current).

In your case do you have any sweep tubes lying around? Using a pair of 6DQ5's, or a quad of 6DQ6's would be ideal, and not need that 15K resistor. Just ground G1, and drive audio to the screens ONLY. Works FB.

After doing similar changes to the Gonset, using a dual trace scope, looking at the G2 mod drive (ch1), and the same tube plate (with AC 1000X compensated HV probe), the signals were identical except for the 180 degree inversion. So it seems that driving the sweep tubes this way, very good linearity is achievable. The tubes will pass on the driver distortion, so minimizing the driver distortion was the main objective.

On your power supply, I think you will need more capacitance for HT1. Go to at least 25 mfd. For HT2, that 400 ohm resistor might be better to be a 10H or so choke. Also a single 5R4 will be overloaded doing both modulator and RF. Maybe use a pair of them, or solid state the HT1 supply.

Just some ideas from an OM in Texas..

Jim
Wd5JKO

[K1JJ]

I can't say when the term "shark fin" actually came about.  I can say that we used to call it "Christmas Trees pointing to the (left or right)".

All low distortion, low phase shift, wide frequency response transmitters are capable of this pattern, and it is not an easy pattern to reproduce.  I first noticed it during the early '70s when I built my first PWM transmitter when I was in college.

Anyway, that's what it is.  It's just one of many measures of a good modulator.  Many modified rice boxes, flex radios, class E rigs, tube PWM rigs, tube grid modulated rigs, etc. can reproduce this pattern, which is typical of the male voice.

That pretty well tells the shark fin story. Shark Fins are a visual observation using a male voice and scope, showing how faithful the extreme low end is.  The area below 70hz or so, extending down to DC, is what appears to make the difference in seeing the fins or not.

For example, in the real whirl, when I ran my 4X1 rig using a backwards audio transformer and 8 ohm amp driver, I had to run my phase in polarity #1. There were no signs of shark fins. In polarity #2, the positive peaks were being swallowed by the negative peaks – terrible waveform. But it was acceptable overall in phase #1.

When I switched to the GFZ MOSFET audio driver for the 4X1 modulators, there was not much difference in the polarities as far as raw levels - polaritiy #1 and plarity #2 were close in performance, but not quite.  #2 was actually better now.  It wasn’t as obvious which one was the proper phase – EXCEPT when using a peak reading wattmeter or looking for the shark fins. The shark fin phase produced a certain low end audio resonance that was not there before. The optimum phase became PHASE #2! It seemed phase #2 was the best polarity for my voice all along, but I could not use my 4X1 rig to pull it off due to its lack of true DEEP low end fidelity.

Technically, the difference between the old transformer driver and the new MOSFET amp was the ability to sweep an audio tone down below about ~50hz SUPER cleanly, without much phase shift. The limitation was now the modulation transformer, that’s it.

In contrast, my class E rig easily sweeps down to 1 HZ. The shark fin peaks are tremendous and the proper #2 polarity choice is even more obvious than the 4X1 rig with the MOSFET driver and mod transformer.

Bottom line is the lower your freq response sweeps CLEANLY towards  0 HZ, the more obvious the shark fins are, EVEN if your voice does not contain much energy down there. I realize there’s no such thing as “sub-harmonics,” at least that's what I've read, but something is going on that takes advantage of these low frequencies when they are available. It might even be low frequency "intermod" generated by the human voicebox, (if that makes any sense)  but whatever it is, it sure sounds mellow to the ear.

BTW, when PYT referred to shark fins, I think he meant the terribly distorted audio you see when a rig is running poorly. Anyone who has built a rig probably has seen this waveform when first fired up and levels are all screwed up…  

T

[G8VOQ]

In your case do you have any sweep tubes lying around? Using a pair of 6DQ5's, or a quad of 6DQ6's would be ideal, and not need that 15K resistor. Just ground G1, and drive audio to the screens ONLY. Works FB.

Jim
Wd5JKO

Thanks for some further insight. Yes, I see the similarity with the Gonset G76.

Well, since I am not wanting to re-use the KT88's, choice of tube finals is up for suggestion.

I think what you are saying is for driving the screen, I'd really want a sweep tube - like 6DQ6's.

I have no tubes for the modulator saving the KT88's which I don't want to use.

Generally, I know I can use class B, zero bias with a pair of 807's.

Assuming modification of the feedback circuit is "a given" irrespective of what circuit is used eventually, I think you are of the opinion that driving g2 grids, will be an advantage (audio quality wise) over class B zero bias with 807's. (Perhaps only really applies if 6L6 circuit kept).

I think maybe that driving g2 there is less load on the driver? Is that the advantage? And not having g1 connected to g2 via a resistor?

Microphone would probably be a Shure 444 I have kicking around.

[G8VOQ]

Looking at the original circuit and the fact that I'm trying to utilise T1, I want someone to come in and say something like "Absolutely, definitely, 100% - ditch class B, zero bias and do...." (Or the opposite).

I wish things were so simple. :c)

[K1JJ]

Looking at the original circuit and the fact that I'm trying to utilise T1, I want someone to come in and say something like "Absolutely, definitely, 100% - ditch class B, zero bias and do...." (Or the opposite).

I wish things were so simple. :c)

Well, since you axed... I know you want to build it with tubes, but since you're still in the design phase, it's not too late to build up the ultimate hi-fi driver for your modulators:

http://amfone.net/Amforum/index.php?topic=25599.0

I was in the same spot as you about a year ago and opted to go with a MOSFET audio driver - and glad I did. The modulator tubes and final tubes are enuff glow to satisfy my fancy...

T

[G8VOQ]

Looking at the original circuit and the fact that I'm trying to utilise T1, I want someone to come in and say something like "Absolutely, definitely, 100% - ditch class B, zero bias and do...." (Or the opposite).

I wish things were so simple. :c)

Well, since you axed... I know you want to build it with tubes, but since you're still in the design phase, it's not too late to build up the ultimate hi-fi driver for your modulators:

http://amfone.net/Amforum/index.php?topic=25599.0

I was in the same spot as you about a year ago and opted to go with a MOSFET audio driver - and glad I did. The modulator tubes and final tubes are enuff glow to satisfy my fancy...

T

What I will likely do is a first build with tubes, to "employ" the parts.

Then continue my homebrew endeavours, with ever more sophisticated designs. :c)

[G8VOQ]

As to what WD5JKO brought to the story, I did find this comment by Jim:


"Many different pieces of ham radio gear from the 1950's to 1960's use
the technique you mention. This is particularly true for mobile tube type AM
transmitters, where size, weight, efficiency, and overall complexity are all
constrained. One such transmitter is the Gonset G-76 AM transceiver. For
push pull audio modulators, they use a pair of 6DQ6 sweep tubes driven from
a separate CT. winding on the audio output transformer. The audio output
tube is a 6CM6 which drives a speaker on receive, and acts as a audio driver
during transmit. Those 6DQ6's easily put out 50 watts of audio this way. Why
did they do this:

* The modulator tubes need NO screen DC voltage
* Efficiency is very important. RF stages run class C, and Audio modulator
stage is zero bias class B.
* The modulator tubes need NO grid 1 bias
* The audio driver and transformer were already there. It just needed some
relay switching, and a additional transformer winding to work.

It appears to me that treating a Beam Power tube this way results in high mu
triode. The result is a class B circuit capable of serious power, good
efficiency, and reasonably clean so long as the audio driver can effectively
deal with a varying load impedance.

For kicks, the RCA receiving tube manual (1973) shows the plate curves for
6L6GC where G1=0, and G2 is the input variable. Those curves sure look
linear to my eyes!"

So, taking the lead or circuit suggestion by JKO, I'd be sticking with class B, zero bias, but a better arrangement? I mean tying g1 to ground.

[G8VOQ]

I was in the same spot as you about a year ago and opted to go with a MOSFET audio driver - and glad I did. The modulator tubes and final tubes are enuff glow to satisfy my fancy...

T

I know what you mean. In the end I suspect the only tube in my best AM rig will be the 813 or similar output tube. And you will look at it through a window. :c)

[WD5JKO]

For kicks, the RCA receiving tube manual (1973) shows the plate curves for
6L6GC where G1=0, and G2 is the input variable. Those curves sure look
linear to my eyes!"

So, taking the lead or circuit suggestion by JKO, I'd be sticking with class B, zero bias, but a better arrangement?

   I forgot I wrote some of that stuff. Keep in mind that with sweep tubes G2 driven (only), the peak audio level will be 100v or less to each Grid (g2) to ground. Going to a 6L6 variant like 807, or many others, the peak audio level will be several hundred volts (possibly 300-400v). The sweep tubes offer the advantage here unless you really want to build an elaborate driver circuit. Getting high drive amplitude at low distortion is somewhat tricky.

   So keeping a circuit as you have drawn will lessen the drive voltage swing need, but will put a 'kink' in the transfer response as G1 draws grid current, and then the 15K resistors drop voltage in response to the grid current. Steve QIX expertly pointed that out earlier. You could 'counter' this with global NFB, but with two transformers in there, stability will be a big issue due to phase shift.

    If you want to retain the KT-88 tubes, use them in tetrode mode with HT2 on the screens, HT1 on the plate, and make a simple bias (C-) supply to set the DC operating point of each tube to about 30ma per tube class Ab1 or Ab2.

  You have a lot of varied, but excellent input here representing different circuit configurations.

   I suppose your going to build it, and hope it works. You can also look at this as a 'work in progress' where you build, learn, tinker some more, and hopefully have a ton of fun along the way. After all thats what its all about isn't it? :-)

All the best,
Jim
WD5JKO

[WA1GFZ]

A 6DQ5 with G1 at zero saturates with about 60 volts peak on the screen

[G8VOQ]

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

I was looking at the above circuit in the STC application note.

When I saw it, I thought, oh, all those power supplies to make.

But, I do wonder whether the 300v supply from my T1 that I've got, will supply the pre-amplifier, phase splitter driver, and 807 g2 grids.

Because the 6L6 would go, and the available current would go elsewhere.

The 750V also that I have, would supply the plates.

Some adjustments here and there and maybe it could work.

I would just have to make a bias supply.

Well, several ideas presented I can work on.

[G8VOQ]

Wow !  Ok, well - I've been solving the same problem over here for the past week or so.... driving, in this case, a pair of 807s in triode connected mode, and I have FINALLY and successfully done it - with high fidelity, and have learned quite a bit in the process.  Here are the results of this work (and it was a lot of work!!!):

[snip]

Hope this information is useful!

Regards,

Steve

Steve, your post really grabs me. :c)  A voice in my head says, investigate the "QIX triode configuration". :c)

I'm going to experiment and see what I get using class B, zero bias, and seek to put into operation all the stuff mentioned: better feedback, cathode follower buffer, the "QIX triode connection", and whatever helps alleviate crossover distortion. I'll get back whenever, and post how things went.

Thanks all for contributions. Rich