4-1000A grid driven transmitter layout

[N4LTA]

I have finally got most of the parts together for a 4-1000A transmitter. I plan to run the tube at 3000 volts. The power supply will have a variac and I can go up to probably 3500 volts. I plan to use a stout 400-500 volt screen supply also with a variac. I mostly will run it on 80 meters but will build it for 160 and 40 also. Plan to use a pi net with vacuum caps. Probably wind the inductor with 1/4" copper tubing. I want it to just loaf along at the legal limit.

What I am most worried about with a grid driven final is stability and neutralization and also layout.

Does anyone have a layout of a typical 4-1000A final. The grid circuit will be below chassis and the plate and pi net above chassis obviously but what else should I look out for. I understand that the 4-1000A will turn into an oscillator easily.

I have a 5" x 17" x 14" Hammond chassis for the RF deck. Anyone tried the QRO 60uF 4000 volt Russian filter capacitor?

Any tips will be appreciated. Thanks in advance.

Pat Bunn
N4LTA

[The Slab Bacon]

Complete transmitter or just an amplifier? ? ? ?

[N4LTA]

Amplifier - I will drive it with a seperate exciter.

Pat
N4LTA

[The Slab Bacon]

Why not make it easy on yourself and make it grounded grid with a tuned input?? At 3500v it will be happy and be much easier to build / design? ?

[N4LTA]

That is a good idea as it should be more stable. I had a 4-1000A that I ran on 160 meters years ago that put out about 2K watts with my IC551 running full bore.

Pat
N4LTA

[KM1H]

What mode is this for and what is the driver?
As a linear or Class C you can run swamped grid input and do away with neutralization which may be a problem on 40M. Its a touchy tube to layout for stability in conventional grid driven. I built many swamped grid 4X1's in the 70-80's, one and two hole and 160-10 for most.

In GG it becomes a serious drive hog, requires a tuned input for any decent efficiency and IMD even as an AM linear, especially at such a low voltage

Carl

[K3YA]

I have both a GG and a plate modulated grid driven 4-1000 on the air.  The grid driven final was also built to be also run as a swamped grid AB2 linear but that proved to be a troublesome and dirty mode of operation and was abandon long ago.  I talked with several other folks running 4X1 in grid driven linear service, including Tim HLR who ran his big rig in that way sometime, and it wasn't as simple as it looked to make it work well.  Here are a couple other thoughts based on my experience.

GG is a LOT easier to build.  However 3500 plate voltage is a little on the low side.  At that voltage it will be need a lot of drive.  A Ranger would probably drive it OK and you may "get away with" no tuned input if the connecting cable was short.  A solid state rice-box would need a tuned input and would be straining.  I have 5.5KV on mine and at that voltage it's quite easy to drive.  For easy, Grounded Grid would be my first choice.

Class C plate modulated would do a much better job overcoming summer static. With a modulator, screen and bias supplies it's several time as much work to build.  3.5 KV is still not very high for that tube and I wouldn't expect much trouble with neutralization or otherwise taming it.  I use the "bridge neutralization" circuit from the ARRL and west coast handbooks and it requires no retuning going from 160-20M or even changing tubes.  With 40M as your upper limit lead length and layout is not that critical.  For QRO this would be my choice.

Parts layout is determined by the size of the components on hand.  I do like to mount the tube at the front of the chassis and provide a window in order to monitor the plate color and show off that cool looking bulb. Also think about keeping heat sensitive parts away from the tube or protect them with a heat shield.

[The Slab Bacon]

Down on the lower bands the 4X1 is a pretty docile tube. My big rig is grid driven with no neutralization and is completely stable. But..............
It is only built to run 160, 80, and occasionally 40m. It sees 90% of it's duty on 75/80m phone and runs along happily on 3400v.

[DMOD]

Here is a companion thread in which Clark is building a linear Amp with a tube I sent him in Trade:

http://amfone.net/Amforum/index.php?topic=31647.0

Phil - AC0OB

[N4LTA]

This will not be a linear amplifier. It will be run Class C and either plate or screen modulated.

It should loaf along at 400 watts carrier. I was planning to build it grid driven. I will also build a driver and would think 60-70 watts would be plenty either GG or swamped grid driven.

It will be mainly run on 75 meter am.

What do you think?

Pat
N4LTA

[K3YA]

I think Satan will stand behind you, whispering in your ear "Turn it up.  Look how the tube glows at 1500W carrier out."  He will say it over and over again to you.  Just like he does to me. 

Do you have the mod iron?

[N4LTA]

I have had that happen to me before with a 6 meter 4CX1500 amp!

I have a Gates mod transformer good for at least 300 watts removed from a Gates BC transmitter - enough to get a whisper from ole Satan.


Pat
N4LTA

[The Slab Bacon]

Think what you want, say what you want, the 4X1 is probably the best bang for the buck out there. Big power, usually cheap to find good used ones, and rather undemanding and very forgiving as long as you stay on the lower bands with it.

They are easy to work with, and look soooo sexy with a nice glow on the plate.
For the high bands those "ceramic jet engines" are the ticket, but down on the lower bands the 4X1 will always be my favorite. The biggest problem of building with them is that they are a rather large bottle. (but that just adds to the kool factor)

[N4LTA]

Slab,

Is yours grid driven and swamped. I have seen circuits that use a 200 ohm swamping resistor and a 4:1 transformer to feed the grid.


I have the tube, sockets and the chokes and transformers and the variable caps. About all that I need are the PS filter caps and I think I have them covered. I could probably use a spare tube.

I'll likely use a pair of 813s to modulate it unless you convince me to go with screen modulation.

Pat
N4LTA

[KM1H]

I think Satan will stand behind you, whispering in your ear "Turn it up.  Look how the tube glows at 1500W carrier out."  He will say it over and over again to you.  Just like he does to me.

I sucumb to that urge whenever a LK-800, AL-1500 or similar comes in for service   There is just something about having 1500-2400W of Pd at your disposal that makes me forget to turn the coax switch to the big Bird load

I was stress testing a repaired LK-800NTC a few times last week on AM; I dont think the guys I worked believed I was just limping along at a 300W carrier from the LK-500ZC by the reports I was getting   

Getting back to the 4X1; it requires a lot of RF voltage swing for Class C grid driven and a lot less in AB.  AB2 works OK IF you are using the GM3SEK tetrode boards or similar to keep the IMD manageable. AB1 makes the input circuit even easier as youre not dealing with drive power; just voltage. Ive built them both ways 160-10M and they put out lots of power, stable, and reasonably clean IMD. Several are still in use on the bands altho I quit building them around 1993-95 or so as local pulls dried up. Then did some 3CX3000A7s for a few years...now there is a fine tube for any mode you want, even brand new the Chinese ones are around $700 and very reliable in BC service.

And if you ever plan to use a MRI pull of any type you better start collecting them now as MRI is rapidly going all SS and sources are drying up fast and prices going up. Ive already put a little stash aside

Carl

[The Slab Bacon]

Slab,
Is yours grid driven and swamped. I have seen circuits that use a 200 ohm swamping resistor and a 4:1 transformer to feed the grid.
I have the tube, sockets and the chokes and transformers and the variable caps. About all that I need are the PS filter caps and I think I have them covered. I could probably use a spare tube.
I'll likely use a pair of 813s to modulate it unless you convince me to go with screen modulation.
Pat
N4LTA

Grid driven, but NOT swamped. Just biased way down in class C. (-150 or 200v IIRC)
The heavy biass and low resting carrier screen voltage is what keeps it so stable. there isn't even ANY control grid circuitry in the final deck. The grid lead comes right off of the socket into a piece of RG-142 double shield coass and goes out up to the exciter where the drive and biass is sent down the line to it.

A single 6146 with 500v on the plate drives it nicely to 45-50mA of grid current. The resting screen voltage is somewhere around 300v and swings the rest of the way with the applied audio munky swing. By adjusting the screen voltage and applied audio, you can play all sorts of games with assymetrical modulation. I usually run somewhere around 125% positive peaks. And it ALL fits comfortably in a 40" tall Budd cabinet with room to spare.

Plate or scream mod, well that is up to you. Hmmmmmmmm.................. 2 6L6s and a 5lb mod tranny vs 2 of something big and a 100+ lb mod tranny. How's yer back? ? ?    

But.............A 4X1 modded by a pair of 4-400s or maybe 833s would sure look pretty through the peek-a-boo window! ! ! ! !    

[K3YA]

Swamped grid is only for grid driven linear service.  You will have a tough time driving the tube into hard class C with a 200 ohm resistor grid to ground. 

Plate and screen modulated use essentially the same circuit as far as the 4-1000 is concerned.  The differences are in the screen or plate supplies. 

Use a parallel LC network to drive the grid with link coupling to the input coax connector.  Wind it on a pill bottle or pvc. Tap the coil and link for each band and use a switch to choose the band.  You can probably get by without neutralization, but the neutralization circuit is only 2 more components, so I would add it.  Use grid leak bias with a sense relay to prevent application of B+ in the absence of grid drive.  Or, use mostly grid leak with a little safety bias, but that's more parts and maybe less linear modulation.

Plate modulation can be real simple to implement if you use zero bias mod tubes and a hi-fi amp as a class B driver.  Let the 4x1 screen self modulate by way of a choke which can be a 125 watt mod transformer primary (secondary not connected).

OR, Frank can advise how to make a cute little featherweight screen modulated  4X1.  I've never done a screen modulated final.  Having recently moved 400 pounds of mod transformer and reactor into the floor a relay rack I can understand the appeal of this approach.

[W3RSW]

Frank, That is a very classy way to use an AR13's mod transformer.

I've heard it and used it.  Your 4x1 screen mod. rig is transportably rugged too.
You did de-ruggedize several plate transformers along the way though, heh, heh.  

In this day and age of restricted carrier, ...a neat way to go.

[N4LTA]

The 4-1000 needs to run class C for plate modulation. It sounds like maybe that if I stay below 8 Mhz, I may not need to swamp the grid but a 4-1000A can oscillate easily I am told. BTW _ it is printed in many places , the ARRL HB for one that a plate modulated stage must run Class C or D  (or have a square law relationship)  for linear modulation. I have always heard this, yet some tell me that it is OK plate modulate a linear biased stage. Never heard of anyone doing so.

One thing that needs to be understood is that I will never drive the 4-1000A hard, because I don't have the power to modulate it at full power and it would be illegal. I am looking for it to loaf along at 300-400 Ma max. At 3000 volts, I am still looking at a KW input and 600 to 700 watts carrier out. My modulator would be really pushing to modulate it at that level.

The screen supply will be a good stout choke input 600 volt supply  with a variac for moving the voltage from maybe 400-600 volts. I don't plan to regulate it and will have a safety circuit that will disable it if the plate voltage is removed. I'll use a 300 Ma Hammond choke as a self bias choke for the screen. I likely will have an electronic overcurrent trip for the screen supply (I have a board layed out, that I used in a previous amp.

I also will have a safety bias /self bias combination power supply for the grid.

My plate transformer is good for over 500 Ma at 3000 volts or a little more.  I have chokes good for 350 Ma CCS and a big 60H modulation reactor and Gates modulation transfomer that is CCS conservatively rated at 250 Ma. Got to believe it will loaf at 300-350 Ma.  Unless I get that whisper mentioned earlier and get real long winded, I think the iron will be OK.  If not I have several 100 watt mod transformers to run screen modulation.


I likely will build the deck for 80 meters only to start with no band switch and possible make it easily doable to change the pi net inductor if I ever need to change bands. It will be 3/16 " or 1/4" copper tubing.

I am close to cutting metal and again , my main concern is layout and having the 4-1000A get nasty. maybe I am worrying too much?

[The Slab Bacon]

You are worrying way too much!! Build it and let it fly!

With my setup, you dont need a protective circuit. With the lowered screen voltage, and the it biassed below class C if you lose grid drive, The plate current falls immediatly to zero and the 4X1 just goes to sleep.

As far as stability goes, I can't make mine oscilate even if I deliberately try to.

[N4LTA]

OK. Then I'm going to lay it out and start drilling and cutting. Tube up front near the middle so I can have a window, Plate tune cap to the right, load beside it to the right and the inductor at the rear, plate choke behind the tube, Filament transformer to the left side, below the chassis. All the grid circuit under chassis.

Meters for plate current, screen current and grid current to the left side.

Screen voltage meter will be on the screen supply panel and the same for plate voltage.

I'll build a perforated aluminum cage around the RF section up top for safety.

I will wind a plate choke since this is a monobander - will test it for self resonance - on phenolic tube or teflon rod, probably phenolic tube because I have quite a bit.

Pat

[kb3ouk]

I don't know about other types of linear amplifiers, but you can plate modulate a grounded grid amp as long as you also modulate the driver tube in phase with the final.

[The Slab Bacon]

I don't know about other types of linear amplifiers, but you can plate modulate a grounded grid amp as long as you also modulate the driver tube in phase with the final.

If the GG amp is running class-B or AB2 it will be difficult to modulate properly, you would prolly achive most of yer mod swing from the modulated driver. I've never seen or heard of a GG amp running in class-C, but I guess it might be doable if you putz with it long enough. You would prolly need to add biass and screen supplies to do it, which defeats the simplistic beauty of a GG amp.

[N4LTA]

All of the technical literature states that the amp must have a square law relationship with regard to power output and instantanious plate voltage. This is the case in Class C, D and E but not in a linear amp.

It would be an expensive waste to plate modulate the final if the driver was already AM modulated and the final was linear.

A GG amp should be able to be biased class C by adding bias at the filament transformer center tap I would think. The tube shouldn't care as long as the grid is negative with regard to the cathode at the proper bias voltage.

Pat

[kb3ouk]

The way I saw it done was use grid leak bias. With a modulated grounded grid amp, you want to modulate the driver just enough so that the final will modulate fully. If you don't modulate the driver, the final won't modulate fully because there will be some unmodulated carrier from the driver that passes straight through the tube.