now lets talk abt something really important - "self modulation" of screens

[N3DRB The Derb]

Going to have a fixed screen supply in all my rigs. educate me on aspects of this. how many henrys,
how much current handling does the choke need to do, only the current drawn or more?

Never built or really fooled with a tetrode/pentode tube before. whole new world.

[N2DTS]

Its not critical.
I have used a 12 henry choke, or just a resistor.

Experiment, looking at the peak power or the mod monitor.
The choke only has to be good for the screen current (very little in most cases).

Some tubes seem more critical, like the 4D32.
4-400, 813, dont seem to be critical.

My pair of 813's just has a resistor, and has no problem doing 4x pep the carrier power.

If you dont plan on using it on ssb, the screen supply need not be regulated.

Brett

[N3DRB The Derb]

ok. but exactly why does this work the way it does? What purpose does the choke serve? Its just a bunch of turns of wire. What is it doing ? I understand it's purpose in a PS, but not here.

This one of those really rare times when I actually want to know the underlying theory about a subject. Oh lord, look out. why does a choke in series with your screen supply get you so much more positive modulation?

and your screen bypass cap still goes right at the socket pin after teh choke?

[KD6VXI]

ok. but exactly why does this work the way it does? What purpose does the choke serve? Its just a bunch of turns of wire. What is it doing ? I understand it's purpose in a PS, but not here.

This one of those really rare times when I actually want to know the underlying theory about a subject. Oh lord, look out. why does a choke in series with your screen supply get you so much more positive modulation?

and your screen bypass cap still goes right at the socket pin after teh choke?

The choke introduces R into the circuit.  As someone else said, even R inline can perform the same function.

It makes the tube slide between bias classes, near as I can tell.... Or at least, it causes the gain to change according to input power (screen current drawn)?

Anyway, works similiar to the "modulator" CBers use...  I can't understand how it wouldn't introduce distortion, though....  Maybe because our tanks are already designed to clean up a Class C signal?


--Shane

[N2DTS]

Yes for the screen bypass cap, at the socket.

As the plate voltage changes under modulation, from nominal to double, to zero, the screen current will vary a lot also, at zero volts on the plate (100% neg modulation) the screen current would be very high.
When the plate voltage is very high, the screen current will be very low.

If the screen voltage is fixed and stiff, the plate current does not quite follow the modulation.

A choke or resistor allows the actual screen voltage to follow the modulation, more plate voltage, more plate current, less screen current, and if the screen voltage floats, the actual screen voltage will increase.

Less screen current, less voltage drop in a resistor or choke in series with the screen supply, more screen voltage. The screen voltage follows the modulation.


Some rigs use a seperate winding on the mod transformer for the screen, the 30K1 I had did, but it worked just as good with a choke.

Some tubes NEED the mod transformer winding, the 4cx250b is one that the manual says needs it, it has something to do with the layout of the grids, the 4cx250b wont self modulate the screen (per RCA)...


Brett

[Steve - WB3HUZ]

No bias class changes.

The screen current changes with plate modulation. The screen draws less current when the plate voltage is high (modulation positive peak), and more current when the plate voltage is low (modulation negative peak). Placing a resistance in the screen lead, of course creates a varying voltage in due to the varying current. Thus, screen modulation.  Proper levels of screen modulation will actually reduce distortion.

The tank circuit won't clean up modulation distortion, unless the distortion products are FAR from the carrier frequency  - 100  multiple hundreds of kHz.

[KD6VXI]

No bias class changes.

The screen current changes with plate modulation. The screen draws less current when the plate voltage is high (modulation positive peak), and more current when the plate voltage is low (modulation negative peak). Placing a resistance in the screen lead, of course creates a varying voltage in due to the varying current. Thus, screen modulation.  Proper levels of screen modulation will actually reduce distortion.

I was under the assumption that bias class was set in a tet by the screen, grid and plate voltages..

IE, if the voltage on the screen is allowed to follow changes in something else, it isn't linearly amplifying....  As evidenced by the need to regulate it for an SSB amplifier.

Might not be changing an entire CLASS of operation, but I'd be willing to bet it's moving from one class to close to another, no?

--Shane

[k4kyv]

With a screen grid tube, the screen needs to be modulated along with the plate.  Running the screen at fixed voltage and just modulating the plate would produce very non-linear modulation.  On negative peaks the screen voltage would be higher than the plate voltage!

The choke allows the screen to self modulate when fed with a fixed voltage source.  You can use a dropping resistor from the modulated plate voltage to feed the screen, or use a  dropping resistor from a fixed source, letting the dropping resistor take the place of the choke.

I have always used triodes, so don't have much experience modulating screen grid tubes.  I would probably try all three methods and see which gave the best modulation linearity.

When using the choke it's kind of like using a modulation reactor.  If the choke  doesn't have enough inductance you don't get good modulation linearity at low frequencies.

[W2PFY]

This is topic that is much talked about on this board. Just do a search on screen modulation or modulation of the screens and stuff comes up all the way
back to 07.

Here's a link

http://amfone.net/Amforum/index.php?topic=12222.0 

[Steve - WB3HUZ]

The class of operation is defined by the angle of conduction and that is largely set by the bias on the control grid. So it's how long the tube conducts over one cycle of the operational RF frequency. Another way to look at it is that it's a DC thing not an AC thing. So there is no varying of classes unless the DC voltage on the grid is changed.

Sure, during part of the positive portion of the RF cycle a Class C amp conducts like hell, just like a Class A amp. But other than at the peak of the RF cycle, the tube is cut off - no conduction.

[W3SLK]

Brett said:

Some tubes NEED the mod transformer winding, the 4cx250b is one that the manual says needs it, it has something to do with the layout of the grids, the 4cx250b wont self modulate the screen (per RCA)...

I thought I read somewhere in an old QST that you could put a choke on the screen and modulate it.

[N4LTA]

From an older ARRL handbook

"The choke L1 is the audio impedance in the screen circuit: its inductance should be large enough to have a reactance (at lowest desired frequency) that is not less than the impedance of the screen. The screen impedance can be taken to be approximately equal to the d.c. screen voltage divided by the d.c. screen current in amperes."

So calculate your screen impedance   =   Screen Voltage/ Screen Current

and then decide the lowest frequency and calculate the value of the inductor.

For my single  4-400  it works out to   Screen voltage = 500    Screen current = 30 ma

500/.030 = 16700 ohms = Z

Z = 2 x pi x f x L

L =  Z/ 2 x pi x f

L = 33h   for f = 80

L = 6.6 h for f = 400


So to get true low frequency response , you need a fairly large inductor - but the current requirements are low.

[N3DRB The Derb]

I hope I have something in the junkpile with enuff inductance. I seem to remember a little bittie thordarson thing back there with 2 leads. it's almost o mall to be anything power supply related. I'm gonna go grab it right now and lookit up.

My 1941 handbook has nothing on the subject. need to get about a dozen covering from 32 to 54.

[N2DTS]

I just remember that from the RCA tube manual.
The 4cx250 and others of its ilk are odd in some respects, its likely the grid and screen layout, but you can bias them at 1000 volts, get 90 ma resting current, and then turn the plate voltage up to 2000 volts without having to change anything else, bias or screen voltage, and the resting current stays at 90ma......

That seems to show the screen and grid have a lot of control, and are different from tubes like the 813 or 4-125/400.

Brett

Brett said:

Some tubes NEED the mod transformer winding, the 4cx250b is one that the manual says needs it, it has something to do with the layout of the grids, the 4cx250b wont self modulate the screen (per RCA)...

I thought I read somewhere in an old QST that you could put a choke on the screen and modulate it.

[N4LTA]

Hammond makes a 157G - 30H 40 Ma  That's what I probably will use.

It sell for $23 or so on the net.

[N3DRB The Derb]

I'm trying to find this thing to see if I got a winner. damn bench looks like a bomb hit it.

[flintstone mop]

I'm from an old school too and 4CX250B seem like a tube more suited for VHF and UHF.
They were very hard to work with in the MOTO UHF base stations.

I have seen a cousin to the 4CX250B, the 4CX300B used in linear amplifiers, but still not in the service of a plate modulated TX using the good ole glass pubes.

KEEP LOOKIN TIM it's around there somewhere
Fred

[ke7trp]

I have run my 4-400 several ways.  Self modulating with the CHoke and with a big Dropping resistor. With the Dropping resistor set to 300 volts I had the best audio and pattern on the scope for AM.  If the Rig was just used for C AM, I would use this method.  I would also employ a way to lower Screen volts down to half for tune up.  Otherwise, You will be tuning at Full power. 

On my Dual 4-400 rig, The screen is supply is sperate with its own trans, chokes, caps, Variac.  I have screen volt and current meters. It has over current trip and a a trip if I lose Plate. It also has a big Red warning light and a reset button on the front panel.

This is because I use the Rig as an amplifier For SSB and for a ranger.  This alllows me to set the Bias to what ever I want.  In Class C am, I would like the screen to be modulated along with the plate, So I will probably find a large Switch and mount it on the deck. When I go to AM, I can Clunk the switch over to a Big Drop resistor off the plate supply so my screen is mudulated by the plate volts.


C

[flintstone mop]

More memory banks kicking in.
It was always suggested to use the modulated DC going to the plate of the final to go to the screen through the dropping resistor. No separate screen supply needed.
Chuck K1KW (big rig) had a separate screen supply for the 4CX1500A and he used a  choke for that tube. Can't remember how any audio got to it.

Tim, there may be little advantage of using a choke. The choke may help with some better low end response.

Fred

[Jim, W5JO]

You can modulate the screen with the resistor or choke as described.  My books say to put the resistor at the plate side of the modulation transformer.  I have seen a diagram of two resistors and a cap.  One resistor to the HV side of the mod transformer and the other at the plate side with the cap in parallel.

With the choke and a separate supply as the plate of the tube(s) go more positive the SG current decreases.  The decreasing SG allows the magnetic field of the choke to collapse, inducing a more positive voltage on the screen, which causes the plate and screen to be more positive at the same time.

Either one simply makes the screen more positive when the plate is allowing the tube to reach 100% modulation.  Your choke should be high inductance, low resistance.  In the Globe King the screen choke is 12 Hy, 80 ma. at 375 ohms.  So if you don't have a choke that will work, you might do the two resistor/cap thing to accomplish the same thing without a screen supply.

[N4LTA]

I am building a single 4-400 now and plan to run it at 1.9-2KV on the plate and 500 volts on the screen. I have a 2400 volt 400 ma plate supply with a variac.

It will have a totally separate screen supply with a variac and a screen trip circuit and also a plate voltage sensing circuit.

I just finished the PC board for the relays and power supply rectifiers and filters. The board has a 12 volt relay power  supply, a time delay relay to delay start the screen until the plate, heaters  and bias is "on", a separate zener regulated grid bias supply that will deliver -220 volts or -150 volts (for linear if I ever need to run it linear).

For relays I am using the Zettler AZ755 series, rated at 20amps at 277 VAC. That works out to 391 VDC at 20 amps. I am thinking that 500 volts at 30 MA will work OK with this relay. I have seen them used to handle 1.5KW at 50 ohms.

I hope to get the board built tonite and test it. Then I will get the control chassis started - hopefully this weekend.

If I can fit it , I will put the modulator power supply transformer and choke on that same chassis. It's gonna be heavy.

[N2DTS]

I think its been shown that a dropping resistor off the modulated plate voltage to feed the screens is not a good idea. I think the handbooks all say to use the unmodulated plate voltage.

Sometimes a dropping resistor off the unmodulated B+ works, or some combo of modulated and unmodulated voltage, might depend on the tube. The 32V3 works better with about a 50/50 mix.

I never liked the dropping resistor setup, that is a lot of voltage to drop with some tubes, plus playing with the tuning can cause excessive screen current.

I like a seperate supply, run it at 100 or more volts above the normal screen voltage, then drop 100 volts in a resistor, that allows self modulation AND some screen current protection.
It makes tuning up much less critical as the screen is self limiting.

I dont like the 4cx250 as a plate modulated class C amp, but they make good AB1 audio amps, and were also used in things like the Collins KWS1 as ssb amps.

Someone was running a single 4cx250 screen modulated and it sounded really good....


Brett

[Jim, W5JO]

My memory is none too good today, but didn't the Valiant use a big WW resistor for the screen voltage on the 6146s? 

I don't like it either but if the plate voltage goes away so does the screen voltage, not only eliminating a power supply but a control circuit or clamp tube.

[ke7trp]

There are many ways to handle this.  Some more complex then others.  If its big sound you want, You wont beat the Screen dropping resistor.

Adjustability and protection will require a seperate screen supply. If I was building a Class C AM Transmitter, I would use the Dropping resistor and be done with it.  On a 4-400 you are talking 25 to 50MA of screen current depending on loading.  Thats nothing and big resistors are cheap and easy.

Brett, I am not sure what books you are reading. But I have a bookcase full of Handbooks and nearly all of them show a Dropping resistor (sometimes a cap) for the screen supply if the design is for Class C am. Only when I see the design for dual use as an amplifier or SSB use, do I see the seperate Screen supplys come into play.

I think there is real merrit to have the screen modulated with the Plate for AM.  The reason, I say this is because I have tested it three ways just last month. Fixed, Self modulated and modulated.  The minute I used the Dropper, I got a wonderfull pattern on the scope, more peak power, headroom and the sound was just super smooth out of the box. The sound quality was just incredible..  But sound is what really matters to me so I am probably "biased".
 
Clark

[k4kyv]

I never liked the dropping resistor setup, that is a lot of voltage to drop with some tubes, plus playing with the tuning can cause excessive screen current.

I would think that would be more of a problem with a fixed screen supply.  With the dropping resistor, if the screen tries to draw excessive current, the dropping resistor drops the voltage down to near zero, resulting in little screen dissipation and therefore protecting the tube.

I recall an article about using some kind of combined modulated and unmodulated voltage source, employing a voltage divider and an audio by-pass capacitor, and they claimed it gave the best modulation linearity.