screen modulation again...

[AB2EZ]

Brett

I would guess that you had one of the two series capacitors in the wrong polarity. Either that, or you have a batch containing some very leaky ("new") capacitors.

Stu

[N2DTS]

Stu,
ebay specials...
On the positive side, the one that was way overvolt did not explode...

[N2DTS]

I eBay'd some real air system sockets for cheap, and have a collection of new 4x150's and an empty chassis so I might do a 3x 4x150a rig with a real blower (a nice small one on hand) and put up to 2000 volts on the plates and use the screen modulator on it.

The current RF deck does 300 watts carrier plate modulated and 150 watts screen modulated, with loctal sockets and a muffin fan with gentle air flow through the tubes, at just under 1400 volts max.
Nothing seems to get very hot as it is, but I do not know how hard I could push it.
As it is, 1400 volts, about 200ma in, 280 watts input (carrier) and 100 watts out, 180 watts plate dissipation out of 500 for the two tubes. 150 watts out would be about half power.

I suppose I could build it so it would be good for plate modulation as well, I have the parts, but it would be bigger....

 

[N2DTS]

I tried the screen modulator on the pair of 813 rig.
I tried various settings and could not get the audio to sound clean like it does on the 4x150 rig.
I noticed the 813's take a lot more voltage on the screens to get any power out.
I tried various settings of grid drive and bias, plate voltage, input voltage to the modulator and modulator settings.
I could get various power levels out, plenty of modulation, but I could not get it to sound real clean.

The 4d32 rig sounded ok, maybe not as clean as the 4x150's but close.
I wonder if there is any way to tell if one tube screen modulates better then another other then trying it...

[AB2EZ]

Brett

I suggest that you repeat what you did when you were working with the 4x150 rig:

Turn everything on, apply sine wave audio, and then use a scope to observe the waveforms ... starting with the first stage of the screen modulator... and proceeding toward the screen bus of the modulated tubes... and finally to the envelope of your modulated RF signal (using an RF pickup).

With a dual trace scope, it is easy to see where distortiion and/or parasitic oscillations are being introduced.

Keep an eye out for low or high frequency audio parasitic oscillations... and also for RF on the screen or the modulator B+. Also keep an eye out for RF getting into the modulator via the microphone cable.

Stu

[N2DTS]

Stu,
It did not sound bad on the 813's, and more fine tuning would likely improve it, but its easy to get the 4x150 or 4d32 rigs to sound very good.
It was just a test, I would not run it with the screen modulator.

[N2DTS]

Playing around with fixing up the old 813 amp as a 4x 4x150 screen modulated rig.
Here is what I have so far:

http://n2dts.smugmug.com/Ham-radio/i-WDMNfJK/XL

It could be screen or plate modulated, but its mostly for the screen modulator.
The sockets will fit in the holes the 813's used to be in nicely.
Its really big, because it used to be an 813 amplifier and its being recycled.

[steve_qix]

I tried the screen modulator on the pair of 813 rig.
I tried various settings and could not get the audio to sound clean like it does on the 4x150 rig.
I noticed the 813's take a lot more voltage on the screens to get any power out.
I tried various settings of grid drive and bias, plate voltage, input voltage to the modulator and modulator settings.
I could get various power levels out, plenty of modulation, but I could not get it to sound real clean.

The 4d32 rig sounded ok, maybe not as clean as the 4x150's but close.
I wonder if there is any way to tell if one tube screen modulates better then another other then trying it...

The oscilloscope will show all !  What does the scope pattern look like with a predictable waveform such as a triangle wave.  Lacking a triangle, a sine wave will do, but it's much easier to see non-linearity with a triangle.

I'm not sure if you (or anyone) can really adjust a screen modulator (or any modulator at all, for that matter) - at least initially - and do testing, without using a scope and a predictable waveform.

If your modulator can cleanly modulate a 400hZ triangle wave, it has an excellent chance of handling just about anything else, at least up to 100% positive.  If you want to test it beyond 100% positive with a triangle, you can do that too.  If the modulator has a negative peak limiter, you're pretty much all set - just keep going..  If not, you could overmodulate in the negative direction for the test, but depending on the modulator, it may or may not be a valid test - depends if the operating point (or some other DC parameter) shifts.

No matter what, the predictable waveform test will reveal absolute evidence of what is going on, and you can trace back through the circuit and find out where the nonlinearities are.

Happy hunting!

Regards,  Steve

[N2DTS]

Yes, I have a scope, and an sdr, and a mod monitor!
I was not really interested in digging into the 813 rig since I would not be screen modulating that.

As far as the modulator goes, once its set up, and it mostly is if you set it like John says, you really have only two adjustments to make on it, the carrier level pot and the voltage input to the modulator.
You could use a fixed input voltage if its high enough, say just put in 700 or 800 volts, then you adjust the carrier level pot to give the carrier level you want.
It always seems to work best is you run the carrier level on the low side, the more carrier you dial in, the less audio you get on the carrier. That is partly offset by increasing the voltage input to the modulator, but only up to a point.

Example, on the 4x150 rig, I can max everything out, 1300 volts on the plates, 650 volts going into the modulator, and I can get 100% neg and 120% pos mod up to about 140 watts carrier.
To increase power I would need to increase the plate voltage and maybe the modulator input voltage I think.
If I turn the carrier pot up, I can get 200 watts to whatever power I want, but the modulation will be lower.

If I turn the carrier pot down, I can get 200% modulation at 50 watts carrier.

Its  subjective but its easy to get the modulator to sound very good to me  on the 4x150 rig and the 4d32 rig, while the 813 rig always had a trace of distortion.

That rig works fine being modulated by a pair of 4x150's in AB1 and can do 700 watts carrier easy, so its kind of pointless to try and screen modulate it and get 150 watts out.
I did notice that the screen modulator sounds even better then that rig to me.
It has always been my best rig, nice mod iron, nice separate power supplies for the rf and mod decks, plenty of clean audio in AB1, everything adjustable, yet the screen modulator going into the 4x150 rig sounds better to me.

I am having a LOT of fun with this thing...

[AB2EZ]

Brett

What is most important is that you are having fun... and you can adjust everything so that it sounds good to you.

However, I disagree with respect to the adjustments that you need to make. This is particularly true because you are experimenting with different values of B+ on the modulator, and the current gains of the transistors you are using can vary quite a bit from lot to lot, device to device (within the same lot), and with device temperature.

You need to set R2 to obtain the correct DC operating point on the first transistor (as per my earlier post).

You need to adjust the biasing on the 2nd transistor (emitter follower) so that the DC level between the emitter and ground is higher than the screen voltage level at carrier... by an amount that is equal to minus the desired maximum negative value of screen voltage needed to produce 100% negative modulation.

Using a scope with a sine wave or a triangle wave audio input, as Steve suggested, is the best way to do this.

Not adjusting these correctly will limit the maximum percentage of modulation you can achieve, and also produce unnecessay distortion.

The scope will also reveal such things as RF getting into the modulator.

Stu

Yes, I have a scope, and an sdr, and a mod monitor!
I was not really interested in digging into the 813 rig since I would not be screen modulating that.

As far as the modulator goes, once its set up, and it mostly is if you set it like John says, you really have only two adjustments to make on it, the carrier level pot and the voltage input to the modulator.
You could use a fixed input voltage if its high enough, say just put in 700 or 800 volts, then you adjust the carrier level pot to give the carrier level you want.
It always seems to work best is you run the carrier level on the low side, the more carrier you dial in, the less audio you get on the carrier. That is partly offset by increasing the voltage input to the modulator, but only up to a point.

Example, on the 4x150 rig, I can max everything out, 1300 volts on the plates, 650 volts going into the modulator, and I can get 100% neg and 120% pos mod up to about 140 watts carrier.
To increase power I would need to increase the plate voltage and maybe the modulator input voltage I think.
If I turn the carrier pot up, I can get 200 watts to whatever power I want, but the modulation will be lower.

If I turn the carrier pot down, I can get 200% modulation at 50 watts carrier.

Its  subjective but its easy to get the modulator to sound very good to me  on the 4x150 rig and the 4d32 rig, while the 813 rig always had a trace of distortion.

That rig works fine being modulated by a pair of 4x150's in AB1 and can do 700 watts carrier easy, so its kind of pointless to try and screen modulate it and get 150 watts out.
I did notice that the screen modulator sounds even better then that rig to me.
It has always been my best rig, nice mod iron, nice separate power supplies for the rf and mod decks, plenty of clean audio in AB1, everything adjustable, yet the screen modulator going into the 4x150 rig sounds better to me.

I am having a LOT of fun with this thing...

[N2DTS]

This is what I have so far:

http://n2dts.smugmug.com/Ham-radio/i-nt9cnzP/A

[N2DTS]

This circuit is amazing, I am having a lot of fun with it.
All my other rigs sound distorted after using the screen modulator, plus I was having fun waving my hand in front of the microphone and finding it modulated the transmitter over 100% both ways at about 1 Hz!
Yes, that is ONE hertz!
And no mod iron to blow up or worry about.

Got the grid circuit done on the 4x 4x150 deck and most of the filament wiring.

[N2DTS]

Gave the modulator a workout yesterday, on 40 meters all day with plenty of my old buzzard transmissions during which nothing much is said with plenty of talking.
I do not know how people put up with it, but anyway,

The circuit is working fantastic, no issues, so today I took the time to take care of a few things.
The meters on the front read output voltage and current.
Current is slight, the voltage is low also, but was up against the edge at 150 volts resting carrier, so I put in a meter to read 0 to 500 volts, not 0 to 150.
My pair of 4x150's runs right at 150 volts for 100 watts carrier.
I took care of that, then experimented with the 100k 2 watt resistor I have across the modulator output.
Without that resistor, the modulator seems unstable as far as the resting carrier (screen voltage) goes.
If the screen voltage is increased past some point, it increases about 50 volts and stays there.
I suspect electrons build up on the screen and do something to the last pass transistor (q3) and its bias point.
I tried a 220K and it still took off. I tried a 120k and it still took off.
I replaced the cap across the last transistor and it did the same thing.
The base of the last transistor is very sensitive, just putting a dmm on the wiper of the pot changes the output voltage quite a bit.

It works great with the 100K resistor, but I would like to know why it happens.
Maybe that pot (r6) is damaged, I ended up with a cheap brand, John says Bourns are the brand to get, made for the voltage. I ordered some of those.

I did not notice the carrier set point changing when I tried the modulator on the other RF decks, the 4d32 and the 813's.

[N2DTS]

I added an isolation transformer on the input of the screen modulator, 600 ohm to 600 ohm, it did not impact the low frequency at all.
I also added a phase reversal switch, easy since its a balanced input to the transformer.
Very interesting, I have a phase switch on the mic input to the audio stuff, and now on the input to the modulator.
The correct selection of both gives loads of positive peaks, 500 watts pep out of a 100 watt carrier.

[N2DTS]

What is the best way to balance the load of four tubes in a screen modulated setup?
In other rigs with more then one tube, I had resistors in the screen feed to each tube.

If I have four tubes in parallel, do I put resistors in the screens and the cathodes?
The most I have done is three 4d32's in one deck, they seem to work fine with some resistors to each screen feed, but I wonder how that would effect screen modulation.
I do not think I have any resistors in the current pair of 4x150 rf deck, and one tube runs a little hotter.

[AB2EZ]

Brett

I would do the following:

With RF drive and B+ applied to all four tubes, with the pi network resonated*, and with the screen voltage set at carrier level: measure the average cathode current in each tube.

*Loading is critical for this purpose.

You can do this by placing a 1 ohm resistor between the cathode and ground of each tube, bypassed by a .1uF capacitor. It doesn't matter if all four resistors are 1 ohm... but they should all be the same value +/- 1%.

Pick the tubes to be reasonably matched, in terms of the cathode current, with the identical carrier level screen voltage.

Add resistance in series wth the screens of the three tubes with higher cathode currents... to obtain equal cathode currents.

Since all four tubes are biased into class C operation (even though the tubes may have somewhat different grid-to-cathode cutoff voltages, and somewhat different transconductances) equalizing the cathode currents, at carrier, should equalize the power output contributed by each tube, at carrier, and also the dissipation in each tube, at carrier.

Stu

[N2DTS]

Thanks Stu.
No way to have them auto balance?

Guess I should build in some test points.

[AB2EZ]

Brett

In theory you could design an auto balancing circuit using op-amps to compare the cathode currents, and (for example) feedback controlled DC grid-to-cathode bias adjustments to each of three of the tubes.

In practice, making the feedback control system stable (vs oscillations at subsonic frequencies) would be a challenge.

"The difference between theory and practice is greater in practice than it is in theory" (Anon)

Stu

Thanks Stu.
No way to have them auto balance?

Guess I should build in some test points.

[N2DTS]

Got the grid circuit and filaments done:

http://n2dts.smugmug.com/Ham-radio/i-kCkgrd7/X2

Next is looking up what to do with the cathodes, they are separate from the filaments and I do not remember what I did on the other deck, ground the filiment on one side, center tap, or what.
I also have to do the rf choke to the plate supply, dc blocking caps to unload the voltage off the pie net, and go to all 4 tubes in some neat looking way.

[N2DTS]

Stu,
I was thinking more along the lines of resistors in the plate, and/or screen and/or cathodes of a value that would equalize the current through all four tubes somewhat.
Looking at old circuits, it looked like they never seemed to do much about it...
100 ohm resistors in the feed to each screen was about it, and I do not think a lot of resistance in the screen circuit is the best for a screen modulated rig...

Brett

In theory you could design an auto balancing circuit using op-amps to compare the cathode currents, and (for example) feedback controlled DC grid-to-cathode bias adjustments to each of three of the tubes.

In practice, making the feedback control system stable (vs oscillations at subsonic frequencies) would be a challenge.

"The difference between theory and practice is greater in practice than it is in theory" (Anon)

Stu

Thanks Stu.
No way to have them auto balance?

Guess I should build in some test points.

[N2DTS]

In between playing with the SLRM receiver, I removed the current 4x150 deck and put the 3x4D32 deck in and tested it with the screen modulator.
That deck does about 50 to 60 watts carrier nicely, it can do more but the carrier starts down shifting under modulation.
Below about 60 watts, the carrier meter on the mod monitor shifts upwards under modulation, above that it starts shifting down. Any shift has to get excessive before you hear anything in the mod monitor.
The 2x 4x150 deck starts the downward carrier shift at about 150 watts at the voltage I can run it at.
Most of my plate modulated rigs downward shift the carrier a bit, the 32v does, the 4x150 modulated by 811's does quite a bit because of the shared power supply between the mod and rf deck.
The big rig does it little, not at all at below maximum ratings.


I have to be careful, as its easy to over modulate in both directions with the screen modulator, its not like a modulator that might sag on peaks some.

I took the bottom off the 2x 4x150 deck to study how I built that one, as it works very well.
It has a 100 watt wire wound resistor in each screen circuit to balance the load, so I will do that on the big rig, and its got .001 disks to ground at the socket.
Is four .001 disks going to be a problem? I will have four 4x150's so it all adds up I suppose.

Slow progress, I have a bunch of flat 20 watt wire wound resistors that mount at their ends, but could not come up with anything that would isolate them from ground, the voltage could get quite high, but the mounts take small screws that wont work with ceramic standoffs.
I tried drilling the holes larger but that is no good.

Winter fun with radio!

 

[AB2EZ]

Brett

Each 0.001uF capacitor has an impedance of -j31.8k ohms at 5000Hz.

Four(4) capacitors in parallel will have an impedance of -j7.96k ohms at 5000Hz

But:

The input impedance... looking into the screens of four(4) tubes in parallel... will be 1/4 the impedance looking into the screen of a single tube.

So, if .001uF per tube doesn't roll off the highest audio frequencies that you care about, it shouldn't matter how many tubes (each with a .001uF screen bypass capacitor) you use in parallel.

Separately:

Since the impedance (at DC and audio frequencies) looking into four(4) tubes' screens (in parallel) is half the impedance looking into two(2) tubes' screens in parallel, you need to make sure that the 2nd screen modulator transistor stage (the emitter follower) is biased with enough current (at carrier) to deliver the larger DC screen current, and the larger peak-to-peak audio frequency screen current swing. The 10k ohm emitter resistor will be larger than the impedance looking into the screens of the four(4) parallel tubes... so the emitter follower's bias pot and the carrier-level screen voltage pot have to be set with the four tubes' screens attached to the modulator.

Stu

[N2DTS]

Thanks Stu.
The screens present a load on peaks, otherwise they are almost open circuit I think, so who knows what happens.
At some point I will have to set things up and measure response and distortion.
I was looking at it and thinking if it was a problem, I could use some 500 pf 5000 volt TX caps, or even 100 pf ones.
I do not like or want it to go out to 20 KHz, 8 KHz is more then enough.


I plan on running the screen modulator on the new RF deck, so I am not spending a lot of time adjusting things on other decks.
Hopefully the new deck will not be a flop like the push pull triode deck was last year.

I have the screen circuit to do, the meter/switch/lights, and the plate choke/blocking stuff, new front panel to paint, and new pots to put in the modulator.

[N2DTS]

Just finished up testing the 4x4x150 RF deck.
I had to work out a few bugs and mistakes, the 40 meter grid coil had too many turns on it and a power wiring mistake.
It did 500 watts carrier out, 1300 volts (most that rig has) at 500 ma cathode current, and with the 811 modulator giving all it could, nothing zorched.
After running it a bit, the tube temps were warm to the touch, easy to hold my finger on them.

More work to do, I need to set the neutralization, remove or reduce the resistors I have in the plates, they are running too warm, speed the fan up and label things.
Then I will put it in place of the 2x813 rf deck, run it at 1500 or more volts and screen modulate it.

[N2DTS]

Hooked the screen modulator up, likes to run about 175 to 200 watts carrier output at the voltage levels I have (plate and screen).
Sped up the fan some.
I put the audio generator into the rig and it looks clean down to about 40 Hz, then the input transformer looks like it hoses the waveform up.
It seems to modulate well beyond 20,000 Hz.