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

[AB2EZ]

This doesn't necessarily contradict what others have already said. Just another way to understand what's going on with "self modulation" of the screen voltage.

Remember... there are electrons being emitted by the cathode... which head toward the positive screen. However, when they reach the screen, they pass right through (at least most of them do) and continue on to the plate. They do this because the plate is at a higher voltage (relative to the screen), and there is an electric field pointing from the plate toward the screen. [Electrons (negative charges) are accelerated in the opposite direction v. the direction of the field... by definition]

If you lower the plate voltage, more of the electrons emitted by the cathode try to stop at the screen. If the screen has a low impedance to ground, those additional electrons stopping at the screen will cause lot's of screen current to flow.

However, if the screen is at a high impedance to ground at the modulation frequencies of interest (e.g. 50 Hz - 10 kHz), these electrons "pile up" on the screen... thereby reducing the voltage on the screen, until the screen voltage is (once again) significantly below the (modulated) plate voltage. [Remember when charges (electrons) accumulate across a capacitor (the screen to plate capacitance) they produce a voltage across the capacitor: V= Q/C]

The same thing happens when the voltage on the plate increases. A smaller percentage of the electrons will stop at the screen grid (instead, they will pass through the screen and continue on to the plate). This will reduce the amount of electron charge that has accumulated on the screen grid... and the screen voltage will rise up. 

For the above reasons, the screen voltage moves up and down as the plate voltage is modulated up and down.

Stu

[ke7trp]

Don, I would like to see that article if you can dig it up. 

I like keeping things simple. 

Now that I have my Dual 4-400 box up and running, I find the adjustable Screen supply is a mess. It needs constant adjustment while tuning, Changing Plate volts, bands and Drive.  Plus, I had to build the Trip system to kill the Screen if I loose plate and i had to put in Over current.  All of that gets eliminated with a Screen Drop resistor.  You loose plate, You loose screen. All thats needed is a Screen Current meter so you can keep an eye on the screen while loading. 

Of course, the main reason I did all this is so I can run the thing as an amp. Wondering now if that was worth the effort. Two 4-400s in AB produces a hell of alot of heat in this small room

Clark

[Steve - WB3HUZ]

Two excellent articles on the subject.


http://www.amwindow.org/tech/htm/scrnmod.htm

http://www.bruhns.us/T-P_Mod_Lin/T-P_Mod_Lin.html

[W3SLK]

Clark said:

If its big sound you want, You wont beat the Screen dropping resistor.

I don't think I or Brett are disputing that except when attempting to use it with a 4CX250/300 style tube. If I get a chance tonight, I'll look through my QST CD editions. I think it was in the late '50's where the guy did some experiments with the 4cx250B and a tertiary winding, (the ART-13 mod iron comes to mind) and the audio choke on the screen. I thought the author covered why use of dropping resistors will not work, (efficiently). A mind is a terrible thing when it is wasted!

[ke7trp]

The poster is not talking about 250Bs.   Why are you in this thread? 


C

[N2DTS]

Re: the 4cx250b,
We were talking about different tubes behaving differently under self modulation of the screen.

In order to get the lowest distortion, tests likely have to be done to find out the best amount of self modulation, and some tubes may not self modulate at all.

813's and the 4-125/250/400 work well, the 4D32 needs attention, along with the 6146 I think, and the 4cx tubes may not work at all.

Brett

[N2DTS]

Just put something like a 2k resistor in series.
I use a big  wirewound pot.
That will make the screen self adjusting.

Without the resistance, you need to be REAL carefull!

Brett
 

Don, I would like to see that article if you can dig it up. 

I like keeping things simple. 

Now that I have my Dual 4-400 box up and running, I find the adjustable Screen supply is a mess. It needs constant adjustment while tuning, Changing Plate volts, bands and Drive.  Plus, I had to build the Trip system to kill the Screen if I loose plate and i had to put in Over current.  All of that gets eliminated with a Screen Drop resistor.  You loose plate, You loose screen. All thats needed is a Screen Current meter so you can keep an eye on the screen while loading. 

Of course, the main reason I did all this is so I can run the thing as an amp. Wondering now if that was worth the effort. Two 4-400s in AB produces a hell of alot of heat in this small room

Clark

[KM1H]

When feeding the screen off the B+ thru a big WW the magic word is efficiency.

The power wasted in heat has to come from the modulator and needs to be calculated into your power budget and iron.

A seperate screen supply AND a winding on the iron requires minimum audio power. The trade off is all the parts that make up the seperate PS unless that voltage is used in the speech amp or elsewhere.

Carl
KM1H

[WA5VGO]

According to the Eimac "Application Notes" for the 4-250:

 "Where a reactor is used, it should have a rated inductance of not less than 10 henries divided by the number of tubes in the modulated amplifier, and a maximum current rating of two to three times the operating dc current. To prevent phase shift between the screen and plate modulation voltages at high frequencies, the screen by-pass capacitor should be no larger than necessary for adequate rf by-passing".

Presumably, this would also hold true for a 4-400.

Darrell

[W3SLK]

Mike(y) said:

If I get a chance tonight, I'll look through my QST CD editions. I think it was in the late '50's where the guy did some experiments with the 4cx250B and a tertiary winding, (the ART-13 mod iron comes to mind) and the audio choke on the screen. I thought the author covered why use of dropping resistors will not work, (efficiently). A mind is a terrible thing when it is wasted!

Damn, I couldn't find the article. It may be in an early '60's QST. A fellow heard me talking about modulating a pair of 4CX250's on 160 years ago and was nice enough to send me the article that he had just read about them.

[N2DTS]

Well, that is what RCA says you need to do, have a seperate screen winding on the mod transformer for 4cx250 screens.
That is why I would never use them in plate modulated class C service, there are better choices.

Not a great choice for anything really, as they have gotten expensive, and old pull outs dont work long without arcing over.

Neat small tubes though, great for modulators in AB1 if you can get new ones, 4x150, 4cx250, the later 4x150 was actualy a glass 4cx250, I found four new 4x150's cheap some time ago and run them to modulate the 813 pair. 600 watts out with zero driving power at 2000 volts.

Four in paralell with screen modulation might work very well....


Brett

[ke7trp]

Carl, With all do respect. I dont agree.  We are talking about 300 to 400 volts at 25 ma on a 4-400. Nothing needs to be redesigned to handle that screen current.  The resistor is not producing much heat either.  Depending on Tube the current could be even less.

Factor in the fact that more power output is typicical and its backwards from your thinking on effeciency. Have you actualy tried this so you can see and measure the efffects of a directly modualted screen?

When feeding the screen off the B+ thru a big WW the magic word is efficiency.

The power wasted in heat has to come from the modulator and needs to be calculated into your power budget and iron.

A seperate screen supply AND a winding on the iron requires minimum audio power. The trade off is all the parts that make up the seperate PS unless that voltage is used in the speech amp or elsewhere.

Carl
KM1H

[KM1H]

Huh? A resistor from a 3000V B+ line doesnt doesnt produce much heat?

Eimac calls for 500V on the screen of a 4-400 and lets round the current off to 30 ma. I dont consider 75W something to just toss off as unimportant. Dropping the screen voltage lower as you specify just wastes more power.

Thats fine if you have the overhead in the mod tubes and iron but it certainly cant be called efficient.

For low power the resistor is a viable option but according to any documentation Ive read its the last choice for a good modulator. Its common in cost conscious ham gear.

Carl
KM1H

[N4LTA]

Yes - A resistor off the plate supply would be a real heater and a big resistor.

On my little 4-400 design with only 2000 volts on the plate and 500 volts at 40 ma on the screen, the resistor would disapate 60 watts. That would require a 125  watt resistor to make me feel good although a 100 watt would work.. Thats a big resistor and a lot of heat.

For a dual 4-400 at 3000 volts  - it would be much worse.

Dropping 2500 volts at 80 ma would require a 31K resistor and would disapate 198 watts.

[Steve - WB3HUZ]

I suppose you could put a slider on the HV bleeder and get the right screen voltage. No extra heat there since the bleeder is already in place. Some sort of protection scheme would be in order, just in case the HV line to the plate was broken. Otherwise, smoked screen!

[ke7trp]

Never thought of that Steve. 

To me 75 watts is not alot.  I use big single Resistor. It never gets that hot.  My screen is down at 25 ma most of the time. Not sure why people think this is a big deal. 

If it produces better sound/signal and is easy and cheap to do, why not run it?  I think I wasted my time and money building this big screen supply with the variac, metering, and all the protection systems I had to put in place.

Clark

[KD6VXI]

I suppose you could put a slider on the HV bleeder and get the right screen voltage. No extra heat there since the bleeder is already in place. Some sort of protection scheme would be in order, just in case the HV line to the plate was broken. Otherwise, smoked screen!

I was going to chime in with the seperate winding on the plate supply, but you also brought up another method.  If you're going to use a string of electrolytics, you can tap off a cap to give you the voltage you need, too.  Just size your caps appropriately, and you can get DAMN close.

I've also used multi-tap plate xformers to get screen voltage before...  As well as plate voltage to run an intermediate power amp.  Just because your using the 4800 volt tap for anode1 doesn't mean you can't use the 1kv tap for the screen   Under load, it drops right below 1.4 kv, as the 4x5 needs .  Run it grounded grid, and the 1800 VAC taps run the 250 series of tubes as great drivers.  One xformer, multiple supplies, and supplies a LIMITED amount of protection on the B+ application without screen voltage.

But, I do like your idea...  As well as tapping a cap at the appropriate level.  Any pros / cons over one to the other?  I typically have my bleeders as my eq resistors...   I guess using it as the bottom resistor would be ideal..  The sliding R would allow for a bit of adjustability, without it, your stuck having to purchase an exact value / amount of caps.

--Shane

[KD6VXI]

Never thought of that Steve. 

To me 75 watts is not alot.  I use big single Resistor. It never gets that hot.  My screen is down at 25 ma most of the time. Not sure why people think this is a big deal. 

If it produces better sound/signal and is easy and cheap to do, why not run it?  I think I wasted my time and money building this big screen supply with the variac, metering, and all the protection systems I had to put in place.

Clark

75 watts to us isn't a lot because we have PEP ratings on output, not DC ratings on input.  30 + years ago, 75 watts was 225 watts PEP  

Think of it this way, you have a great adjustable supply for other homebrew projects now.  That's something a lot of people would give a right arm for....  I have a couple HV Lambda supplies here that at the time I was like WTF, I HAVE to take them, but now they make GREAT 200-400 vdc adjustable screen supplies (I could never figure out what they used them for, since Lambda is a HUGE computer / rack system PS mfg...  You guys had some, didn't you?)  The only problem I've had with the Lambda's is that they run HOT.... And as such, they are ALL suffering from cap's going out.  I have a 12V one here I was going to use as a bench supply.. heh...  It has 12VDC and 11.9VAC on the output.  The 60+ thousand uF filter cap measured about 100uF.

--Shane

[KC4VWU]

The neat one is if you have a CT transformer, run a FW Bridge for Plate, use the CT through a suitable diode string tapped in at filter cap bank mid point. Half your Plate B+ there. 2400v at the Plate equals 1200v. at youir CT feed. Now you have a lower potential for the SG to deal with power resistor choice or possibly even regulate it at that level. Very interesting thread; I'm still learning.
Phil 

[N4LTA]

Tapping the electrolytic string only will work if you have large wattage equalizing resistors drawing much more current that you plan to tap off. That means large wattage ones that disapate far more than a single dropping resistor. Otherwise - you will unbalance the equalizing string and blow up a cap.

The slider drops the voltage the same as a separate resistor - anytime you drop voltage with any kind of a resistive scheme - you disapate the amount of voltage dropped  - multiplied by the current that flows through the resistor - That's ohms law.

The economy type power supply mentioned above will get you the voltage - but it won't be modulated  - it is the same as a separate screen supply.

If you drop the screen voltage off the modulated plate supply - it is going to take some power -  the amount of voltage dropped times the screen current.

[KM1H]

The comment earlier about supposedly not using the 4X150 family in Class C AM raised its head this morning while in the "library" room and reading old QST's. I started with 1926 many months ago 

In the May 55 issue Eimac specifically advertises the 4X150A for AM service without any caveats.

Getting back to the resistor subject. I dont have any proof one way or other which method looks best on a scope. My 813 amp uses a seperate screen winding so no experimenting was necessary. I do have a NIB pair of 4-400A's that may get built around but thats off in the distance.

Anyway, throwing away 75 to 200W is wasteful to my Yankee way of thinking. Id certainly not mess with the equalizers or bleeder.

For protection you can use a high value resistor off the HV to feed a DC relay coil to ground. The relay voltage can be zener stabilized. Then feed the screen voltage thru the energized contacts. A loss of HV will open the relay and hopefully save most screen grid tubes. It has the benfit of working with any screen voltage source. Ive used that method many times in the past with large glass bottles minus the zener which wasnt readily available or reliable in the 70's.

Carl
KM1H

[ke7trp]

Shane, I think what these guys dont understand is the benifit in output PEP when you modulate the screen.  75 watts of current lost in my transmitters power supply for another 300 watts pep output?  That seems like a fair trade off to me.  I guess it really does not matter as my PS is built serious overkill.

Neat ideas on screen supplies.. Keep em comming!

I dont believe the Comments on the 4cx250 and other tubes in that family not running am with modulated screens. I know a man that is doing this now and we have not seen any issues.  I looked through my library and found several designs posted in books all showing self modulating screens.  I would love to see the article that explains you cant do this.  I am always looking for information.



Clark

[KD6VXI]

I dont believe the Comments on the 4cx250 and other tubes in that family not running am with modulated screens. I know a man that is doing this now and we have not seen any issues.  I looked through my library and found several designs posted in books all showing self modulating screens.  I would love to see the article that explains you cant do this.  I am always looking for information.

Clark,

I believe the reason you  don't want to run the 250 that way is it is DEFINATELY one of the tubes that likes to pull "negative screen current", so to speak. 

Delicate bastard tubes. 

--Shane

[ke7trp]

I have twenty of them.. LOL


Did you work on that amp anymore?

C

[Ed/KB1HYS]

For my 4-250 rig, I built up a bleeder that is tapped for the screen. Worst case current draw from the tap is designed to be max screen current limit, so the screen is safe that way.  Since this is unmodulated B+ I was going to put a 10H inductor I had in the screen line to help with screen modulation.

Very simple. fairly safe, probably not efficient -a compromise but ya do what ya gotta do.

Someday I'll get it finished and we'll see if it works or if I compromised all the smoke out...