The AM Forum
March 28, 2024, 07:24:28 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: Using 4-400s at lowish anode voltages  (Read 3967 times)
0 Members and 1 Guest are viewing this topic.
G4OJY
Member

Offline Offline

Posts: 9


« on: January 31, 2023, 04:12:44 AM »

Hello

I want to build a plate modulated transmitter for 10 and 15 metres. Here in England we can only use max 400W pep (100W) carrier so it doesn't need to produce the kind of power used over in the US. I don't want to use 813s in the final as I find they run out of steam at higher frequencies. I have some 4-400s in the junk box and would like to use them if possible.

I already have a homebrew 2X813 modulating 2X813 transmitter for 80m AM, and my plan was to use the modulator in this to modulate a 4-400 in grid driven class C.

Now for the problem. I'm not sure what HT voltage I can use without zorching the mod transformer (Woden UM4). There is no manufacturers data on max HT voltage. It currently runs with 1250V in the 813 transmitter, and seems quite happy with this.

So has anyone here used a 4-400 as a plate modulated amplifier with a HT voltage of about 1.3kV? If so what bias and screen voltages did you use and how much output power did it achieve?

Thanks

Andy G4OJY
Logged
KA2PTE
KA2PTE
Member

Offline Offline

Posts: 118



WWW
« Reply #1 on: January 31, 2023, 10:03:19 AM »

I got an old Johnson Thunderbolt linear going a couple yrs back, it uses 2 of those tubes in parallel.
Plate volts from the factory was I think only 1800v, quite low than what the tubes can actually handle.
Screen voltage was about 500V, but it too can go higher, but I beleive your plate must also increase
proportionately. Grid was biased at I think -75v for AB2 but mine was modded for AB1 and they have
it somewhere like -80 I think. They have a CW mode where they have it at +150V I guess bringing it near class C operation.
Logged

==================
EF Johnson RF Tube Linears
==================
https://groups.io/g/johnson-tube-linears
K4NYW
Contributing
Member
*
Offline Offline

Posts: 160



WWW
« Reply #2 on: January 31, 2023, 09:24:00 PM »

FWIW the US Navy had a series of HF transmitters introduced in the 1950's that used a 4-400 with 1050v (AM) or 1300v (CW/FSK) anode supply. Driven by a 5933 (807W) and modulated by a pair of 5933's.  100w output.

There was an optional "booster" with 2400v/3000v supply and a pair of 4-125 modulators for 500w output.

Autotune AN/URT-2, -3, -4 and Manual tune AN/SRT-14, -15, -16
The SRT-14 manual is here -
https://www.navy-radio.com/manuals/srt14-manual.htm
Cheers,
Nick K4NYW
Logged

G4OJY
Member

Offline Offline

Posts: 9


« Reply #3 on: February 01, 2023, 05:17:46 AM »

Thanks for the info Nick. I had a look at the manual for that transmitter, it looks ideal for what I want, so I should be able to adapt it to make a 10m AM rf deck.

Cheers

Andy
Logged
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #4 on: February 01, 2023, 07:51:52 AM »


Many years ago, I remember on the air discussions about the low 1800v'ish B+ used on the RF final of a Globe King 500. Many reported that a 4-250 did a better job at low B+ than a 4-400 did. One theory kicked around is the plate geometry differences between the two tube types. The 4-400 plate zig zags /\/\/\/\/\ in a way that the average distance between the filament to plate is a greater distance with the 4-400, hence the less plate current off resonance than the 4-250.

I could be all wet, and FOS...maybe those reports, including mine, were from an OP with a strong 4-250, and a weak 4-400.

Jim
Wd5JKO
Logged
W1AEX
Un-smug-a-licious
Contributing
Member
*
Offline Offline

Posts: 1488


Apache Labs SDR


WWW
« Reply #5 on: February 01, 2023, 09:57:10 PM »

I used a variac on the plate supply for my 4-400 RF final x pair of 833 tubes as modulators and since I had variable regulated bias on the modulators I could run the 4-400 and 833's at less than 1000 vdc to run around 75 watts output just for the novelty of it. The tubes never complained whether I was set for less than 1000 volts or to the max setting of 3200 volts. You should be fine.

73,  Rob W1AEX
Logged

One thing I'm certain of is that there is too much certainty in the world.
w8khk
Member

Offline Offline

Posts: 1200


This ham got his ticket the old fashioned way.


WWW
« Reply #6 on: February 02, 2023, 09:03:25 PM »

I used a variac on the plate supply for my 4-400 RF final x pair of 833 tubes as modulators and since I had variable regulated bias on the modulators I could run the 4-400 and 833's at less than 1000 vdc to run around 75 watts output just for the novelty of it. The tubes never complained whether I was set for less than 1000 volts or to the max setting of 3200 volts. You should be fine.

73,  Rob W1AEX

Rob, I find that information very interesting.  I wonder if I may ask a few questions.  I am curious how much input and output power you get when running at around 2000 plate volts?  Are you using a single 4-400A, or a pair?  And how do you provide the screen voltage, via a series resistor, audio choke, or a separate supply? 
I am looking at building an RF deck that will have between 1000 and 2000 plate volts, and I am not sure whether I want to use 4-250As, 4-400As, or possibly a pair of 304-TLs.  Your experiences will help me formulate a plan. 
Thank you,
Rick W8KHK
Logged

Rick / W8KHK  ex WB2HKX, WB4GNR
"Both politicians and diapers need to be changed often and for the same reason.”   Ronald Reagan

My smart?phone voicetext screws up homophones, but they are crystal clear from my 75 meter plate-modulated AM transmitter
w9jsw
Two shots of Whisky
Contributing
Member
*
Online Online

Posts: 592



« Reply #7 on: February 03, 2023, 07:21:57 AM »

Rick,

I am just now about ready to power up a pair of 4-400s RF deck. The design point is 2 voltages, 2000 and 2500. i am using dropping resistors for the screen. One 225W one for 2000v and add a series 75W one with a vac relay for 2500v. The design point for input power is 2000 at 525ma. At 75% efficiency that should yield around 800W output power.
Logged
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #8 on: February 03, 2023, 08:38:17 AM »

To add my half a cent worth here...

I've been working on and off for a year on a Broadcast conversion, a RCA BTA-1R1.  It uses 4-400A's, a pair modulating a pair.  As designed, it has three power levels, 250, 500, and 1000 watts carrier.  To get these, it runs at 1550, 2150, and 3000 volts.  This is with a very deep into class C bias (datasheet is 220 volts, RCA has 400 volts bias), and a double pi output that is lossier than a simpler output network.  In my conversion, with a standard pi output (using vacuum caps and a 30 amp roller) and datasheet bias levels, it makes well above those levels -- up to 340 watts at 1450 volts (that's the minimum I can adjust it to).

So -- low plate voltages seem pretty doable. And if this thing could spend half its life doing overnights at 250 watts/1550 volts, I would not worry about the low voltage hurting anything...

Ed
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #9 on: February 03, 2023, 11:35:13 AM »

Maybe the only real things you may have to deal with will be screen grid dissipation and control grid dissipation.

A low plate voltage usually begets higher screen current, so that the scren voltage has to be reduced. Heavy loading of more plate current may help there.

As suggested in a previous post, lots of grid bias (deep class C) will help the grid and improve efficiency by narrowing the grid conduction angle so the plate current can be as high as possible during the grid pulse. The grid drive has to be enough to drive the tube during the crest of the modulation cycle just like the usual high voltage operation, and that is where the dissipation can be a little high.

These comments are from operating a 4-1000 at 2000V Eb, which is ridiculously low for that tube. A 200V screen voltage works well on that situation. It's quite a balancing act but once the best combination of voltages is found it's gold. I get about 300W carrier out of that.

---

maybe some TL;DR related comments about parts;
If you can find a modulation transformer T201 from the MD-141/GR modulator-power supply for the T-282/GR transmitter, it could be useful. I read you have a good Wotan, but if you like the lower voltage stuff and find one of these it's worth picking up.

They are not very popular because they're the size of a 250W unit and have a 125W rating. Therefore cheap, if they can be found.
Despite the stated frequency response of 200-10000Hz +/-2dB, it does very well for those in the basso profundo voice range because the unit is oversized for the 125W power rating.
It was used with a pair of 811s with 8K Ohm p-p winding, modulating a pair of 4X150s with IIRC a 2800 Ohm winding.
All of this ran at 850-900VDC and the output was 100W carrier in the 400MHz military aircraft band. The transformer also has a center tapped audio feedback winding. I ran one on 1250V with no trouble. It's 1957 military grade so maybe it's well-made.

The schematic for the screen modulator of that transmitter might be interesting for the 4-400. The T-282/GR uses a very simple series modulator with a 5763/6AQ5, for the screens, operating at 190V output to the screens which is adjustable. The DC to the modulator comes off the 900V supply and is dropped by three VR tubes to the plate of the 5763. The screen mod chassis with front panel plugs into the T-282 transmitter, not the MD-141, and includes the low level audio stages too. That module is also a good find.

The "Training Book" here has schematics etc. for those units:
https://bunkerofdoom.com/lit/mil_uhf_01/index.html
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
W1AEX
Un-smug-a-licious
Contributing
Member
*
Offline Offline

Posts: 1488


Apache Labs SDR


WWW
« Reply #10 on: February 03, 2023, 03:56:41 PM »

Rob, I find that information very interesting.  I wonder if I may ask a few questions.  I am curious how much input and output power you get when running at around 2000 plate volts?  Are you using a single 4-400A, or a pair?  And how do you provide the screen voltage, via a series resistor, audio choke, or a separate supply?  
I am looking at building an RF deck that will have between 1000 and 2000 plate volts, and I am not sure whether I want to use 4-250As, 4-400As, or possibly a pair of 304-TLs.  Your experiences will help me formulate a plan.  
Thank you,
Rick W8KHK

Hi Rick,

My RF deck used a single 4-400 and when run at 2000vdc and 200ma (400w input) it would produce pretty close to 300 watts output. That was actually what I typically ran for power on all three bands (160/75/40). Initially, I used a dropping resistor for the screen but pretty quickly switched over to a separate screen supply that used an inline 10mH choke to self-modulate the screen. The 4-400 screen supply had its own variac to allow setting up the screen current within ratings. I used a 4-250 for awhile until a local broadcast station gave me a few 4-400 tubes that were pulls. With a 4-400 in the PA it would run fine right up to 3000vdc plus a couple hundred more and 300ma current but that was pretty rough service for a single 4-400. That being said, it never arced over and never failed but the plate of the 4-400 sure looked pretty!

For better or worse, the rough and dirty drawings of my rig are archived at this link:    

https://www.w1aex.com/4_400/4_400.html

From way back when with my WB1AEX call sign, the rig can be heard here:  

http://www.w1aex.com/mic/AEXAM.mp3

Good luck with your build!

73,  Rob W1AEX
Logged

One thing I'm certain of is that there is too much certainty in the world.
w9jsw
Two shots of Whisky
Contributing
Member
*
Online Online

Posts: 592



« Reply #11 on: February 03, 2023, 05:33:10 PM »

Rob,

Did you do anything to prevent over-current on the screen supply if the plate voltage drops inadvertently? That was the reason I went with the dropping resistor.

John
Logged
W1AEX
Un-smug-a-licious
Contributing
Member
*
Offline Offline

Posts: 1488


Apache Labs SDR


WWW
« Reply #12 on: February 03, 2023, 06:03:00 PM »

Hi John,

That would certainly be advisable but in my youth I stupidly enjoyed living in the danger zone! That scenario never happened to me but it would be a certain quick death for the 4-400 if the plate supply failed. Glad you mentioned that for the benefit of anyone going down that path with their project.

73,  Rob W1AEX
Logged

One thing I'm certain of is that there is too much certainty in the world.
w8khk
Member

Offline Offline

Posts: 1200


This ham got his ticket the old fashioned way.


WWW
« Reply #13 on: February 03, 2023, 06:39:37 PM »

I do not wish to monopolize Andy's thread with my questions, but it seems that so far the detailed answers are in line with his query, so I will continue.

John, Thank you for sharing your plans and specifications.  My goal is to get a reasonable amount of power out, while limiting myself to around 2KV plate voltage.  I was leaning toward a pair of push-pull neutralized Eimac 304-TLs thinking that the tremendous emission and lower plate resistance would serve me well with limited high voltage.  And I already have the RF deck that my dad built in 1949, which I have fully restored.

You have renewed my consideration of using tetrodes, either a pair of 4-250s or 4-400As.  It seems they will play well with the available voltage, the only down side is the need for forced-air cooling, which is already available for the modulator.  Just trying to keep things quiet.  

As far as screen grid safety is concerned, I usually employ a double interlock for the screens on expensive tubes.  First, excess screen current will trip an interlock that must be manually reset.  Second, a relay provides screen voltage ONLY when sufficient plate voltage is present.  This, to me, is just as safe as using a dropping resistor, but allows more flexibility in the screen supply design.

Ed, I have an RCA BTA-1MX, but I often wish it was a 1R1 instead.  The 833As look cool, but the 4-400As are much more friendly when trying to push the frequency up to the higher bands.  I was not aware of the voltages used for the lower power settings, and that is exactly the detail I was hoping to learn, thank you!

Rob, you went above and beyond the scope of my questions, but it is much appreciated.  Looking at other folks designs in detail, I never fail to learn something new.  Gracious sharing of all the inner details is remarkable, thank you.  The drawings are not the least bit dirty, most excellent!  I have compared the Eimac '250 and '400A tube specifications, and other than a slight difference in interelectrode capacitance (1/10 pF) they look identical except for the max voltage, current, and dissipation limits.  On another thread, Jim, WD5JKO mentioned hearing that there are additional subtle differences, but he has not yet confirmed that as fact.  My assumption was that the 400 had higher dissipation ratings due to the finned outer structure of the anode, thus the ability to radiate more energy.  Based upon all the input, I look forward to comparing the performance of the 250 and 400 in the same circuit.

The reason for the interest, and the 2000 volt limit, is that I am trying to match the RF deck with the capabilities of my Class-A series modulator, employing a single Eimac 3CX3000A, and the only challenge remaining is how to cool it without it sounding like a topical storm in the shack.  I will be running 5000 volts on the plate, using power supply components from a Collins FM broadcast transmitter, providing 5000 volts DC at one ampere.  The modulator tube, functioning in a similar fashion to the series tube in an electronically-regulated power supply, will provide a nominal 2000 volts to the final, with 3000 volts across the modulator tube.  It will be adjustable, such that the final plate voltage may be raised or lowered, but the 2000 volt target leaves some headroom for positive peaks.  It seems the Eimac tetrodes in the final will play quite nicely, can operate in parallel instead of push-pull, and will be somewhat easier to drive.  They may not even need neutralization, but the layout will allow it if found to be necessary.

Thanks all, for sharing your respective experience and knowledge.

Gotta shift gears now, it is almost AM Rally Time!
Logged

Rick / W8KHK  ex WB2HKX, WB4GNR
"Both politicians and diapers need to be changed often and for the same reason.”   Ronald Reagan

My smart?phone voicetext screws up homophones, but they are crystal clear from my 75 meter plate-modulated AM transmitter
G4OJY
Member

Offline Offline

Posts: 9


« Reply #14 on: July 30, 2023, 01:13:04 PM »

OK so after a few months of work I've finally got the transmitter built. The driver and PA are based on the Navy SRT14 transmitter operating in low power mode. 807 driving a single 4-400.

However test results have been very poor with low output (about 40W), and I seem to have some RF getting back into the mains supply.

I'm using about 1200V anode, 400 screen and -100 grid bias on the 4-400. This is giving me anything up to about 200mA anode current, about 75mA on the screen and 10mA grid current. What I am wondering is, do the 2 capacitors between the 4-400 filament and ground have poor RF performance? They are 10nF ol style transmitting micas. These should normally be ok at lower frequencies, but will they work at 29MHz.

What type of capacitor are others using here, and what value for operation at 29MHz.

Cheers

Andy
Logged
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #15 on: July 31, 2023, 03:57:22 PM »

However test results have been very poor with low output (about 40W), and I seem to have some RF getting back into the mains supply.

I'm using about 1200V anode, 400 screen and -100 grid bias on the 4-400. This is giving me anything up to about 200mA anode current, about 75mA on the screen and 10mA grid current. What I am wondering is, do the 2 capacitors between the 4-400 filament and ground have poor RF performance?

The bias voltage seems rather low, while the screen current is way higher than I see on my 4-400 RCA.  At various voltages on the plates, varying resistances keep the screen current in the mid-twenties per tube. I shoot for exactly the same 10mA per tube grid current, but I'm seeing -230v bias to do that -- modified from the stock -400v.  How are you achieving the bias? Strictly grid leak, or something else? How are you achieving the screen voltage? A resistor, or something else?

I don't have experience with those micas at 29MHz. But I'm more concerned that the screen is already melted on your 4-400...

Ed

Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
G4OJY
Member

Offline Offline

Posts: 9


« Reply #16 on: August 01, 2023, 08:43:33 AM »

Hello Ed

yes you make a valid point about excessive screen current, and I only saw it that high for a short while. I should really reduce the screen voltage.

The US Navy An/SRT14 that I've based my transmitter on uses about 1kV on the anode, 300 on the screen and -120V on the grid, when operated in the low power mode. I should really reduce my screen voltage to this level. The screen and grid voltages are derived from fixed supplies.

I have checked the filament decoupling caps I used with both my MFJ259 and VNA. They aren't particularly good at 29MHz. I'm being sent some other caps from someone who works with high power SW broadcast transmitters so I will see how these perform.
Logged
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.096 seconds with 19 queries.