4-400 screen supply

[N4LTA]

I am looking at building a plate modulated 4-400 transmitter and have most of the parts together and the power supply built. I plan to use the 4-400 as a tetrode at 2000 volts on the plate  on a single band (75 meters)

In the past I have built a 4CX1000 tetrode 6 meter amp and the screed supply was very important. How important will the screen supply be on this simple 4-400 amp? Do I need a shunt regulated FET type screen supply for high quality plate modulated AM?

Thanks in advance.

Pat
N4LTA

[Jeff W9GY]

Pat, I use a transformer with a small variac on the primary which allows me to adjust screen voltage and thus power level.  The supply is just a simple capacitor input filter.  I put a large choke in series with the screen so that the screen will swing at audio frequencies and not strip off the modulation.  Seems like it was in the henry range, but DC-R isn't a big issue since the screen current is small.  A .001 bypass right at the screen pins seems adequate at RF and not large enough to bother the audio.

Be careful not to apply screen voltage without the plate supply being on!

Jeff

[N4LTA]

Does the 4-400 require a shunt regulated supply or something like a zener string  or shunt resistor to absorb negative screen current like some of the large ceramics tetrodes?

It would be nice if you could get by with a screen shunt resistor to absorb the negative screen current.

If necessary - I can build a MOSFET high voltage shunt regulator but I don't want to if I don't need to.

Pat
N4LTA

[W1AEX]

Hi Pat,

I run my 4-400 RF deck exactly as Jeff has described it. An unregulated variac controlled supply with a series choke to function as an audio reactor. I control the screen current by adjusting the variac. It works fine, nothing fancy, and it has been foolproof over the years.

73,

Rob W1AEX

[N4LTA]

That sounds like the way I will go.

I once had about ten panel mount 1 amp variacs in my Junk box and threw them away. I have been kicking myself since. They were used in a large printing press motor operated pot that controlled the drive speed.  The motor was a 125 volt DC motor and the variacs and a DC Bridge rectifier were selected by relays for ramp up speed  , ramp down speed and emergency stop speed . The old technology was replaced with a PLC and we had about thirty headed for the garbage can. I kept about ten for 15 years and threw them away. I knew I would need them for something!

I just wanted to make sure that I didn't need to get fancy with the 4-400 like some of the ceramic tetrodes require.

[W1AEX]

I once had about ten panel mount 1 amp variacs in my Junk box and threw them away. I have been kicking myself since.

Isn't that the way it always seems to happen? Heh, instead of throwing stuff out now I try to give it away to someone nearby. That way I can at least beg for something back if I need it. Of course if they throw the stuff away I can blame them and that lets me off the hook!

Good luck with the project!

73,

Rob W1AEX

[N2DTS]

You only have to  regulate the screen supply if you are going to use it as an amplifier in other than class C.

What I did on my 813 pair was a variac choke input supply, then into a wirewound pot to drop some voltage, that allows it to self modulate, but also allows some protection, screen current goes up, the voltage drop through the pot increases, lowering the screen current.  It helps make things a lot less touchy.
I also included an overload relay, adjustable to open just above the recomended screen current.

With that setup, I COULD turn the pot to 0, adjust the variac to give the correct voltage, and use the deck as an amplifier...choke input, heavy bleed, good regulation.

I wont build a cap input supply, I like the choke input, for regulation, and to limit the inrush current, makes life on the diodes easier, plus you can use less capacitance in the filter.

Brett

[N4LTA]

Brett,

I am with you on that. I have an internet company (a hobby) that sells small cheap guitar amps. (www.stf-electronics.com) I have a Hammond dealership and use Hammond parts for the amps. They all use capacitor input supplies - in fact most all guitar amps have them. I have learned to hate them and their lousy voltage regulation. They are probably fine for a guitar amp adding to the distortion and compression but I get a lot of people who don't understand what is going on. They select such and such Hammond transformer to give them X voltage. They usually select a transformer with three times the needed current capacity and then scream about the voltage being too high and say that the transformer is bad.

I have learned that chokes are your friend when it comes to a quality power supply.

BTW If anyone needs Hamond Parts and you need $80 or so, I can get them shipped direct from Hammond to you at a decent savings. The $80 is about the minimum order. If they have to come to me and get shipped out again, it is hard to make a good deal.

[Jeff W9GY]

Ah yes, Brett's post reminded me of the fuse I put in the screen circuit, just "in case" .... I think I'm using around a 100 mA fuse if, for some reason, the plate voltage goes away while excitation is on.

[N2DTS]

Well, its not just the plate voltage going away, tuning up will change the screen current a lot, as will not having the antenna hooked up, the wrong antenna, wrong band, and other 'senior' moments....
I have to go to great lengths to protect the equipment from ME...

If you use a fuse, put the holder on the front panel, and have some sort of indicator, or you will be tearing the thing apart when the power output goes away sometime and you dont think of the fuse...

I did an overload relay that lights up a reset button/light, hard to miss that....

Brett

[WA1GFZ]

You do want a load resistor on the screen supply for safety and negative screen current . The 4-400 is less sensitive to screen variations than the 4CX1000A and I think it can handle more screen dissipation.

[N4LTA]

Probably best to put a screen load resistor at the screen that draws 10 Ma and let the screen current meter read zero at 10 Ma - then you can read up to -10 ma on the meter if the screen current goes negative.

I have a couple of transformers that I plan to put  bridge rectifiers on and some chokes to get two good stiff supplies - one for the grid and one for the screen. I also found a 2 amp variac on the auction site for the screen.

I may use a zener string for the grid bias  - is that worth the time or is a stiff choke input supply good enough?

I have a 10 H 75 MA choke to install in the screen lead also.

[N2DTS]

What I do for bias is a variac choke input supply, with a diode and a grid leak resistor.
That way you have protection bias, but when you drive the grid with rf, the grid leak resistor supplies the bias.

I think that is the better way to do it, as far as modulation goes.

I have never seen any reverse screen current on my pair of 813's, only normal screen current.
I put a resistor in series with the screen, the screen supply has a bleeder resistor, then the big wirewound pot,
then the overload, then the meter shunt and voltage takeoff for metering.
I meter everything, grid current and voltage, screen current and voltage, plate current and voltage, and meter the currents for each tube and combined...

Brett

[KM1H]

Are you going to modulate the screen from a seperate winding on the mod xfmr, dropping the voltage from the HV or just not worry about achieving 100% or more modulation?

Each way has its own concerns.

How will you control the 4-400 on receive?

Carl
KM1H

[N2DTS]

With most tubes, a series dropping resistor works fine as far as modulation goes.
Chokes also work, the B+ dropping resistor does not always give good results by itself, the 32V rigs do it that way but seem to modulate much better with a voltage divider between modulated and unmodulated B+, there was an article about that someplace.

Bill Orr always had the screen dropping resistor off the UNMODULATED B+ I think....

I just use a seperate supply, a dropping resistor, and have no problem getting 130% positive at 90% negitive....
 
Brett

[Tim WA1HnyLR]

Hi Pat,
My take on it. I have a similar transmitter to what you are planning on there. A single 4-400 with 2Kv on the plate. The transmitter is built upon an old BC610-E carcass. The intermediate HV supply is about 600 volts. I used a 50Hy reactor that I had.This is placed in series with the screen supply to allow for self modulation of the screen. I also placed about 2K or resistance in series with the reactor as well.  This is for current limiting under fault conditions as well as dropping the screen voltage to about 500 volts when operating under a load of 40-50Ma.I use a VR 90 from the biass supply through a current limiting resistor to allow 10Ma of current flow through the VR tube. A 5 K wirewound resistor is in series with the -90V source to the grid circut. This provides grid leak operating bias when drive is applied.The -90V is protective bias should the excitation fail . Typical operating parameters: Ep 2Kv,Ip 250Ma. , E scr 500V I ,scr 50MA. I g 20Ma.The tube runs ver efficiently this way. No need of variacs. I subscribe to the K.I.S.S. system( keep it simple stupid). A small power transformer may be used to obtain screen voltage. A 600 volt CT old radio or TV power transformer with a bridge around it and choke inpoot filter or an 700-800Vct transformer with full wave rectification and cap inpoot filter would do the same thing. An old TV set vertical outpoot transformer makes a damn good screen reactor. Plenty of inductance. You might use 3 -10Hy 75Ma or better filter reactors in series to get the required inductance. If you really want to subscribe to the Kiss theorem use a 30K 200 watts dropping resistor from the modulated B+ to the screen and a 10K 25 watt grid leak resistor. That works quite well . Look ma ! no other power supplies.
De Tim ,WA1HnyLR

[N4LTA]

I have been looking at a 4-400 KW amp in the 1965 Handbook for ideas. The article states that the amp will work for AM with a 10H choke in series with the screen for self modulating the screen.

The circuit uses a solid grid bias supply and no grid leak  -  if you lose the bias supply - you lose the tube.

correct?

They also use a solid (very solid - 200 MA transformer - more than would seem necessary) screen supply seperate from the plate supply. The screen supply has a overcurrent trip-

That was the idea I was using as a guide. I realize that using a dropping resistor off the plate supply makes a safer situation to protect the screen.

I guess my main question is  - is a 10H choke large enough to make the screen self modulate?


Also - during receive - I had planned to cut the tube off with high bias -350 volts or so - would this be the normal manner?

Thanks for all the input.

[WQ9E]

Pat,

Also grab a copy of the Johnson Desk KW manual from Bama for further ideas.  It uses a pair of 4-400's and has provisions for running in linear mode for SSB but you can focus on how it is set up for Class C.  It uses a separate screen supply with a 12H choke to allow the screen to "self modulate".

Overload protection for the tube/plate supply is furnished by a mechanically latching relay that senses cathode current in the final amplifier.  This is easy to implement either with the "old school" relay used by Johnson or a modern electronic substitute. 

The screens in the 4-400 are pretty robust compared to much of the later 4CX series tubes but a fast acting screen fuse would offer some protection.  If you decide to build the final so it can run in linear mode then seriously consider the G3SEK "tetrode board" since it provides a nicely designed and convenient control system, overload protection, and regulated grid voltages.  I used this in my homebrew amp (triple 4CX800) and the tetrode board works beautifully.  In any case, make sure you meter the screen since it is the best indicant of proper tuning/loading of your tetrode.  I added a small case with an external meter to my Johnson Desk since the designed neglected to include provisions for screen metering; this was corrected when the T-bolt was introduced.

Rodger WQ9E

[N2DTS]

I agree, simple is nice, but you loose the ability to play with things (real time) or set it up as an ssb amplifier, or a low level AM amplifier.

I dont do ssb, but included the ability anyway.

Theoreticly, I could build some low level very hi fi exciter and put it into the 2x813 rig and get 100 watts or whatever out.
That seems pointless, but I could do it....

Brett

[WQ9E]

Brett,

I have run my Yaesu FT-817 through my Heath KL-1 "Chippewa" so it is nice to have the option of using a low power exciter with a high power amp.

I am with you on this, given all the work in putting together a nice final it is worthwhile to be able to use it wit a variety of gear.  For example, a Central Electronics 10 or 20 series puts out a nice sounding AM signal when run through a good amp.

[N4LTA]

I agree - letting it be able to run in AB is a good option. I have a 10 watt K3 that could use this for a little kick.


I had originally planned to build a MOSFET Screen shunt regulator will all the trip circuits and bells and whistles but I think that is a little too much for this amp.

As of now - I am looking at a solid screen supply with a variac fully metered in voltage and current. I'll put a 10-20 H choke in series with a good stiff shunt resistor at the screen bypassed with a good quality .001uF cap.

A regulated (zener string ) bias supply with a selector switch to pick off the "top" 5 zeners for setting the bias on the control grid. As someone said earlier - I need to protect this thing from me more than anything else.

I have a variac controlled HV supply that will deliver a good solid 2300 volts at 350 MA or so - I am thinking 1900 -2000 volts on the plate.

I have a nice Hammond 200 watt Mod transformer and I guess two 811As would do the job.

[KM1H]

Im using the G3SEK tetrode boards in my 6M NCL-2000 and for a military UHF amp for 432 that uses an almost irreplacable 2500W Pd tetrode. The 8122's aint $29.95 no more either so its all cheap insurance. Distortion is also more noticable on VHF and above where there is no band noise or other cruddy signals to hide it in.

Since the 4-400A has excellent IMD specs; using vintage tech wont hurt on SSB, thats the way Eimac tested them. Its still GIGO with most exciters anyway. Ages ago I built single and dual grid driven AB2 4x1's and used a couple of triode connected 6550's for the screen regulators. With a SA borrowed from work the IMD was in the - mid 30's worse case. That equaled the 100V driver so I was satisfied.

One of the 60's HB's (ARRL or W6SAI) had a dual 4-250A or 400A grid driven amp built around the new B$W network. Details include the switching from Class C to linear. Might be a good place for ideas.

Carl
KM1H

[N4LTA]

Carl - It was a 4-400 and that is the HB article (1965)  I was looking at. It used a couple of VR tubes to switch from Class C to AB and opened the ground on the VR tubes to cut the amp off. Also mentions adding the 10 h choke in series with the screen for plate modulation and a switch to short it out for CW or SSB.

Its getting time to start drilling and punching holes.

BTW - That old 1942 HB  superhet receiver is working pretty well with no regen in the IF or Mixer.  The AVC could be a lot better and there are a lot of tweeks that need to be done. It turned out better than I expected - but I didn't expect much out of it. I used it to monitor the OFN AM net here on Sunday - I could hear everyone but one of the guys is about 5 miles away running a Globe King and the RX  near jumped off the bench when he came on.

Pat
N4LTA

[N2DTS]

1942?
Dont you know there is a WAR on, what were they doing playing radio?

I dont have a 42 handbook, I would like to know about the design.
That would be octal tubes?

I found the miniture tubes drift a lot less...

262 KHz or 455 KHz?
They make filters for 455, but not 262...

Brett

[N4LTA]

Do a QST search from 38-40 on a regenerative superhet receiver. It has 7 tubes and I was lucky enough to find a couple of Millen IF Transformers and an old Hmmarlund 455 Khz BFO inductor. The original design used a regenerative mixer stage and a regenerative IF stage. If have so far  - left the regenerative stuff out.

I guess I ought to post a photo of it but it is still in the research stage. It is all octal tubes with a 6C5 LO and 6C5 BFO  - I changed the LO after having a time with parasitics. It now has a Hartley tapped inductor rather than the tickler type coil in the article. A 6AS7 Mixer, 6SK7 IF amp, and a 6SQ7 detector , and 1st audio with a 6V6 Audio output tube  - I used a 6V6 instead of the 6F6 th earticle called for. It also has a VR105 for regulating the voltage on the two oscillators and the screens of the other tubes. Plug in coils - which I have wound over many times since I used a different oscillator and different tuning caps. Also has a seperate power supply - with an 80 and two chokes - Its a choke input supply .

Was a lot of fun and is actually the first true tube superhet  that I have ever build. I have a friend who is building a sister to it using some different IF cans. We hope to have an AM QSO on them when finished.