Antenna not resonating.

[Opcom]

Roasted the 4:1 current balun today. Did not seem to hurt it but the nylon cable ties made a stench. Everything was checked with the MFJ analyzer. A handy 270 Ohm CC resistor put across the balun was close to good, was OK or as-expected from 2-50Mhz. balun is ok.

Checking the antenna with the balun and also with a 1:1 current balun, it does not resonate anywhere. It had been around 100 ohms at 7160KC. R is almost always zero, X is off the scale in most places. SWR also >25.

It is dark now and I can't check this. The whole thing is a balanced open wire feed line going across the room and through the wall to a big knife switch in a box outside. From there, through two standoffs, then more of the outdoor open wire feed line goes on up to the dipole. The dipole is just a dipole.

What is very weird is there is no resonant place at all except there is some funny business up high>70Mhz. What could be the deal?  Open somewhere?  Tomorrow when it is light I will check everything at the switch box, clean the connections, etc.. but this is not small hardware, it ought not go bad. The antenna feed is smallish wire.

[flintstone mop]

Antenna shorted right at the feed point? The point where the ladder line connects to the dipole?
I forget your QTH...ICE across the feedpoint?
You were describing some readings that "Normally you see 100 ohms at 7160, R is almost always 0 and X is off scale" or are these readings when things seem to be operating.
Your antenna system is open ladder to a tuner?
Maybe I"m way off base here trying to think this out.
Fred

[Steve - K4HX]

Does sound like a short or nearly so, somewhere. That would explain a hot balun.

[Opcom]

R is almost always 0 and X is off scale is now when it is broken.

100 Ohms at 7160KC was what it was when working.

Yes - There is a tuner, then the balun. It is open ladder from the balun terminals in the shack, all the way to the tower, then up the tower 55 FT, to the center of the dipole. It is a single unbroken open wire ladder line except at the large grounding switch, a double throw knife switch. No short I could see in there. Should be a fairly normal setup.

The balun was so hot the nylon wire ties inside melted and the PVC case was a bit soft.  The cores could not be touched. It is a "2.5KW COMTEK" balun from DX eng. Looks like it's OK once it cooled off. This is annoying because the 'tuner' tuned it.. I had no indication anything was wrong until the match started going bad, a little at a time, that was probably the balun heating up beyond it's ratings. I didn't think about it and re-tuned.

There is no ice or water here now. There was a storm 2 weeks ago.

Maybe something to add is a port to use the MFJ analyzer for checking the Z before using the system, not just relying on a good report from the TX-swr meter-tuner arrangement.  

[W3RSW]

Now that it's daylight, what did you find?  Sounds interesting.

[VE3AJM]

It does seem that the balun is fried or a short somewhere. What is the tuner and the amount of power that you were using with it? Thought of using a link coupled tuner and dispense with the need for a balun?

Al VE3AJM

[K1JJ]

"R is almost always 0 and X is off scale is now when it is broken."


This is a long shot and you probably have already checked....  but put a 50 ohm non-inductive resistor across the MFJ-259 to be sure it reads R=50  J=0.


Those pin diodes are fragile and forever failing. I once blew them out when switching between antennas on a windy / static day.

T

[John K5PRO]

Ferrite cores might have reached curie temperature, and changed, cracked. Take a good look. Sometimes their properties alter when overheated. But that's not the problem that started it all. What was it?

I was in your neck of the woods for 2 weeks, just before the big ice storm and then during the following thaw. At Continental Electronics, accepting two new high power amplifiers.

[Opcom]

To answer, here it is, and also some things I learned while doing this 'fix'.

The dipole center insulator is supported by a rope and pulley. The ice storm's build-up on the 4 legs of the 'dipole' was considerable and had stretched the rope about a foot.

At the very top of the ascending open wire ladder line, the rope was against the feed wires just below the center insulator and pushed them together, shorting the line there.

I was able to see it with binoculars. Somehow the pulley has turned and its position is not favorable to the way the rope is routed. I can fix that later. For now I just pulled the thing back up and moved the rope's ground-side fastening point enough to workaround.

So, it was a short as suggested. The balun seems OK. I inspected it and the cores are not cracked. The insulation on the wires of it is some high temperature stuff chosen for its dielectric constant to smoothout the response, so the web site says.

The tuner is a Murch 2000 and the power level was a 700W carrier for testing - I was hunting audio feedback problems. (Why was there no RFI to the audio chain using a hacked together nasty audio setup, and endless issues when using proper 'audio' equipment?? I believe it is a big joke).

I have thought about a link coupled tuner but I want to try a balanced-balanced tuner. That way I can still use a current balun on the input and it can be an isolating 1:1 design. I have not ruled out a link coupled tuner though.

I did test the MFJ analyzer. It's fine. The balun performs well now. Thankfully I did not break it.

The problem that started it was the short at the center insulator of the dipole.

The short also caused the audio RFI problem.
Unknown why but the voltage must have been very high somewhere. So, having JUST set up the new audio chain and killer speech amp, I was in "DUH"-land looking for that problem, while the balun cooked.

-----

I also blew up a B&W "KW input rules" LPF. The dielectric sheets are punctured on one side of it. I don't know what they are made of. Thin, transparent, flexible, clear like a  gallon zip-loc bag.

Also blew up the new TR switch I just made. It had two contactors, each 600V 8A. Arced and burned I guess, have to tear it open and see. So much for using contactors. I thought they would be fine. so large. hah. How do you blow up something like that?

-----

Its disappointing because I have been trying to make the whole thing blow-up-proof. What would have alerted me to this problem?
An RF ammeter between the tuner and balun? OK if I know what the values are supposed to be. I have a 5A and a 15A one somewhere.

-----

For now -TR switch wise -I am using an old 'KW sized' B&W amplifier bypass switch to go between the transceiver and the transmitter (TX is really a modulated RF amplifier) until I can start on the new TR switch and station control panel. The audio problems are gone, apparently.

----

The next step might be to use the transceiver only as a driver for the transmitter. And use an electronic TR switch with a 75A-2 I was recenly given in trade for a Drake 2B. I have a Johnson electronic TR switch already and it is rated 4k PEP (supposedly) It has been exposed to the wrath of that transmitter before with no ill effects. This will do until I can start on the new TR/control box.

----

Not to whine but the transmitter gives, either bad efficiency or poor ability to modulate fully, when the carrier is much less than 500W. It loves 700-800W, but I can't use that much becaue of the rules, unless I modulate only downward. "Bad efficiency" is not so much an issue. 300W out is 600W in, on a 4-1000. Who cares.

----

Something else learned. The loading is very coarse on the TX. It has not a variable loading cap but two sets of switchable fixed RF caps, "coarse" and "fine". There is a gap where I want to be, yet the values of the caps say there should not be a gap. I thought of  removing the 'fine' set and putting in a medium sized bread-slicer.

-----

Lastly, although the final stays cool, the pair of 3-500Z's in the modulator get orange with 100% modulation of a 500W carrier. meaning, a 5-10 minute test into a dummy load and they are bright orange.

They are bright orange at the bottom and only a dull red near the top. I thought they ought to be uniform.

Can anyone comment on that?

They run about 3300V and 300mA or so. It's not the  color, but the un-even distribution of it from top to bottom. That is what I am uncertain about. It is the first time using those tubes hard.

----

And thanks for the suggestion about the short. Sure enough. it could only be seen with binoculars.

[flintstone mop]

Even with professional audio equipment there will be RFI issues when it is really over the top exposure. Broadcast stuff can deal with broadcast freq. Short wave freq start another problem. I have been down those roads too. Balanced audio should not be problems hmmmm.
But my last two QTH's the antennas are almost 150 feet away from the shack, that is in the basement. Never been any RFI even with a 1 kw......For Testing purposes.

I read that you chose to support the center with rope. Probably a better approach for support. I am using the ends of my dipole to support the center feed. I have the TrueLadderLine...ladder line turns into antenna at the so called feed point...no connections to go bad.
http://www.trueladderline.com/

Very durable and it has never failed. They use a very nice flexible weave type of wire.The PVC "center feed" is hanging in there with wind and ice load.
Any way glad you are back on the air.
Fred

[WQ9E]

Op,

One of the things you  should do for "blow-out" prevention is the same thing that protects the final and power supply, trip circuits for excessive plate, grid, and screen (for tetrodes) current.  Most antenna faults will cause one or more of these to go outside of their normal range and a fast acting trip circuit will protect a lot of things from damage. 

The last big ice storm here cause the G3SEK "tetrode board" on my homebrew amp to trip and avoid damage after part of the inverted L succumbed to ice during a 160 meter contest.

[WD5JKO]

Pat,
 
   I am glad you got things going, and the antenna issue understood.

Others have given you some good advice. I have a couple of other bits to add:

1.) When bringing the station on line, always key up with the exciter first to verify the antenna / tuner are behaving themselves. Then go QRO. Realize that QRO can still have a flashover, but at least the SWR was good when keyed.

2.) When you notice the tuner is "way-off" from where it's supposed to be, resist the temptation to tune it again. Walk outside, look at the antenna, use your antenna analyzer, etc...

3.) If the rig tunes well at 600 watts, then consider that on 40m daytime conditions (where I usually hear you) often have 20 db or more in propagation fading. Nobody will notice if your Tx power is 2-3 db stronger or weaker.  This won't be like driving at 110 mph when the speed limit is 55, instead it will be like running at 75 when the rest are at 55. Add in the QSB, and no-one will notice.

I'd love to hear how the audio stuff is going...There are numerous threads on that topic you started..

Jim
Wd5JKO

[WQ9E]

Jim,

Three excellent points!

Far too many hams miss the point of slightly reducing power below the equipment's stated limit in order to greatly increase component life.  I don't think many (any?) hams have ever paid attention to the rule requiring using the minimum power necessary to maintain communications but since we are paying for expensive components it behooves us to operate at conservative ratings when possible.

[Opcom]

Pro audio gear:
I feel better now, that the BC audio gear can act up on short wave. So its normal. If the set of three bolted-together racks were all enclosed to where they had metal side panels in between them, and rear doors, some of this would probably not have been so much trouble. Side note I am not yet using the CBS Volumax 400 or the Marti compressor limiters. The Behringer just got thrown in for some reason, probably because it gave me 20dB extra gain, needed to not have to mouth the mike. I can see that I need one of those screens in front of the mike because it picks up the unwanted speech noises.

Ladder line antennas:
I do have a bare wire ladder line like that http://www.trueladderline.com/ one. Instead of stranded, it s some kind of very hard solid wire that looks like copper. The legs of the antenna are made of the same ladder, with very heavy woven and soldered joints. It starts outside at the enclosed grounding switch. From there it is a common 450 Ohm insulated ladder line coming into the room which is a metal builing. This enters through two ceramic tubes through the wall and hangs from the ceiling and is about 9FT above the floor.

I do not know what effect if any the transition from 450 Ohm wire to 600 Ohm wire has. SWR efects maybe, but being line it does not seem to care and the antenna works well.

Blow outs:
About blow out prevention, the transmitter has the overloads and they work, but the meter readings and the like are not disturbed much by bad loads. The screen overload is the one that would actuate, but it didn't, because they are calibrated for the design power level which is about 900W. A bad load will make the screen current go above 100mA, but the trip is about 150-200mA. Plate current doesn't care. I have never gotten it to over 500mA and it trips at 800-1000mA. These are 'fault' protections, not really designed to catch 'slow burn' things. I was supposed to catch those.

So, I want to consider something on the post-transmitter side of things. W5FRS suggested that I could put an RF ammeter in the coax between the tuner and balun and note the readings with standard power levels on each frequency of interest. I only use 7160 for now. I would have seen the meter was way off, if there had been one there.

Minimum power necessary:
As far as using the minimum power necessary, there are several technical problems with using less than 300-350W carrier on that transmitter including screen dissipation and distortion. It's simpler to run the equipment as it was designed, but at a 'low' setting of 350W. I don't have a 100W AM transmitter that is working on 7160KC.
100-200W would be the minimum necessary in the noise environment. We hear those with 20-25W. but they are pretty noisy and not always understandable.
I doubt anyone would notice the difference between a 350W carrier and a 600W one. Sometimes I test at high power only for troubleshooting things like RFI and intermittent arcing. I do my best to always obey the power rule we are stuck with.

The fault for the blowouts was mine for not resisting the temptation to retune the load.

Audio:
The audio stuff got a good test today. Where was everyone? The band seemed noisy but I had a good session with W5FRS. He said the audio was much improved and had a full sound.

Before, the bass was lacking and I am sure it was because of the speech amp's inadequate power and the EQ needed to compensate, and the poor match. The 150W speech amp with a load on it has fixed this and has super good regulation over the entire audio cycle. It's a class B driver thing.

I should use cathode follower 6080's as suggested, and there is a circuit in the Gothic AM handbook. In another thread I worked on the schematic to add NFB from the modulator plates back to the grids of the 6080's.

For now I am enamored of the 150W Stromberg-Carlson PA amp as a driver.

Forward:
Its been a while since I took pictures of this whole setup. Maybe can do it soon. It's very changed from the online pics. I avoided it because the shack is so messy with all the work.

There is an instability in the transmitter if, while keyed and driven, the grid tuning is moved off from resonance. I want to put back in the 300? ohm NI swamping resistor that I had in there before. I won't do that until after I have built the power supply for the interna modulation scope. Then I only have to remove the RF deck 'once' to do all of it, maybe including changing to a variable cap for fine loading, because the RF deck is so large. About 30x30x24" COL Tucker built this very spaciously.

Lastly, today I went back to using the TS430 as just an exciter and the Sunair GSB900 as a receiver, and have it on the shielded loop. The loop is now 'keyed' to quit the reception when the key is activated.

I am now to just one switch to go from RX to TX, and with the compressor I don't have to super-regulate the distance to the mike. The GSB900 is not muted, but I am using comfy Califone 'language lab' headphones. I can hear what the TX sounds like and that is very satisfactory.

I can throw another switch to put the TS430 on the air by itself for SSB on 14300 the maritime Mobile net. The 75A-2 needs put in, instead of the GSB-900. The 900 is a fine set but lacks many useful controls and does not seem to have a limiter or noise blanker, the KISS principle for military radios.

It seems satisfactory now until I can work on the aforementioned TR switch and station controls assy. This month, since I was laid off, I have put about 50 hours of work in the shack on this station. It has been worth it. Still need to find a permanent full time job with benefits in Dallas TX. You guys talked me out of seeking BC engineer work a long time ago because the position is not respected in the industry. No matter what I do, I want to have the chance to earn the respect of colleagues and management.

[N2DTS]

Below is a quote from an early post....


Not to whine but the transmitter gives, either bad efficiency or poor ability to modulate fully, when the carrier is much less than 500W. It loves 700-800W, but I can't use that much becaue of the rules, unless I modulate only downward. "Bad efficiency" is not so much an issue. 300W out is 600W in, on a 4-1000. Who cares.

I suppose you are just loading the TX up light to reduce power?
If you run the same voltage and load up light, the mod transformer match and the Q of the pie network goes out the window.

The only good way to be able to adjust the power output is to make the plate voltage adjustable.
Another reason I like home brew, many rigs are set up to run at some power output point, while my home brew rigs have everything adjustable, plate voltage, screen voltage, fixed grid bias, grid leak bias, screen dropping resistor.


When I operate, I tend to look at the power meter (pep) and the mod monitor, both will tell me if something goes wrong with the antenna.
My rigs run close to ccs ratings, so if anything went very far off, something would trip, as I set my screen current protection circuits just above where I normaly run it at, and the 15 amp breaker in the outlet strip will blow on the big rig with excessive plate current.

One reason to not run really high power all the time is that the guy running 50 watts 1000 miles away will hear you well, while you can barely hear him. That makes for a hard qso. Also, you may be heard well 1000 miles away, and be on top of a qso you can not hear, or that seems very weak at your location.
With REALLY strong signals, you have to use care to hard limit the high frequencies or you tend to be really wide, strong hifi signals can take up 20 to 30 KHz, because even though the highs are down, the signal is strong enough to put a lot of power out far. Its not bad with 100 to 300 watts, but above that, it can get nasty.

200 to 300 watts is a nice power level to run, except maybe on 80 meters at night.
My antenna's are not great, so I tend to run about 300 watts all the time.
I do not think I would want to run 800 watts carrier all the time, that would cause trouble in many ways.
My rig will do that, but I have never run it there.

I do not think anyone pays any attention to what goes on with ham radio power levels unless you were to go nutz and get complaints to the FCC for running many KW of carrier power.
300 watts and a very good antenna is likely stronger then 900 watts and a poor antenna, so who is to know?
 

[WQ9E]

Is the screen trip not adjustable on this transmitter to set it just above where you normally operate it?  If not consider adding your own trip circuit in addition.  High sensitivity relays are cheap and paralleled with an adjustable resistor you can set the trip current to whatever you need.  Tie the relay into the existing system or build a latching setup if needed.  One of the benefits of using a tetrode as a high power final is the ability to provide protection given the sensitivity of screen current to operating conditions.

[W3RSW]

Good comments and potential fixes.

Also furthering an idea:

Its disappointing because I have been trying to make the whole thing blow-up-proof. What would have alerted me to this problem?
An RF ammeter between the tuner and balun? OK if I know what the values are supposed to be. I have a 5A and a 15A one somewhere.

Well, in the heat of playing with new audio equipment or forgetting tune-up
 things after being away for awhile;

You need a 'stall' warning (like airplanes, -always seems to go off when landing light planes.)  for your slow burns that the operator is supposed to catch.

An audio buzzer or tone gen. coupled to your MFJ analyzer when SWR goes very low and Z goes very high.  Audio enunciation could even be pitch and amplitude changing depending on conditions if the pick-off/osc. circuit is fancy enough.  Maybe even voice, "your door is ajar," if your tricky enough. 

 I like the add-on screen current shut-down idea too. But that's sudden and can be caused by a bunch of things.  The Analyzer alarm is more weird load/antenna specific.

[N2DTS]

Its a problem on a big rig run very light.
The antenna can screw up, plate current can be excessive, and screen current will be LOW, not high.

I suppose open wire line and a tuner does not help, since the antenna can screw up and something will still be radiating, and the swr might not look that bad, plus the tuner with open wire line might not be close to the operating position, so you can not see the meters.

I avoid all that by using a resonant dipole, coax fed, no tuner, right into the station control, then a watt/swr meter.
Any change at all will show up as a different power output.
No power lost in tuners, very little in the coax, no balun's, nothing to go wrong or fail, but its not an all band with one antenna setup.
No one watches TV over the air anymore, or if they do its digital, so I have not used a low pass filter for many years.

Its sort of like flying an aircraft I suppose, you need to keep an eye on all the meters, modulation, plate current, PEP power output frquency, and do a scan.

[Opcom]

Below is a quote from an early post....


Not to whine but the transmitter gives, either bad efficiency or poor ability to modulate fully, when the carrier is much less than 500W. It loves 700-800W, but I can't use that much becaue of the rules, unless I modulate only downward. "Bad efficiency" is not so much an issue. 300W out is 600W in, on a 4-1000. Who cares.

I suppose you are just loading the TX up light to reduce power?
If you run the same voltage and load up light, the mod transformer match and the Q of the pie network goes out the window.

The only good way to be able to adjust the power output is to make the plate voltage adjustable.
Another reason I like home brew, many rigs are set up to run at some power output point, while my home brew rigs have everything adjustable, plate voltage, screen voltage, fixed grid bias, grid leak bias, screen dropping resistor.


When I operate, I tend to look at the power meter (pep) and the mod monitor, both will tell me if something goes wrong with the antenna.
My rigs run close to ccs ratings, so if anything went very far off, something would trip, as I set my screen current protection circuits just above where I normaly run it at, and the 15 amp breaker in the outlet strip will blow on the big rig with excessive plate current.

One reason to not run really high power all the time is that the guy running 50 watts 1000 miles away will hear you well, while you can barely hear him. That makes for a hard qso. Also, you may be heard well 1000 miles away, and be on top of a qso you can not hear, or that seems very weak at your location.
With REALLY strong signals, you have to use care to hard limit the high frequencies or you tend to be really wide, strong hifi signals can take up 20 to 30 KHz, because even though the highs are down, the signal is strong enough to put a lot of power out far. Its not bad with 100 to 300 watts, but above that, it can get nasty.

200 to 300 watts is a nice power level to run, except maybe on 80 meters at night.
My antenna's are not great, so I tend to run about 300 watts all the time.
I do not think I would want to run 800 watts carrier all the time, that would cause trouble in many ways.
My rig will do that, but I have never run it there.

I do not think anyone pays any attention to what goes on with ham radio power levels unless you were to go nutz and get complaints to the FCC for running many KW of carrier power.
300 watts and a very good antenna is likely stronger then 900 watts and a poor antenna, so who is to know?
 

I agree with those points. the amp was a hi-fi amp, the frequency response was cut off sharply at 3.6KC by a filter I made.

I am not touching the tuning or loading to go QRP, just reducing the final's plate voltage, and when it gets to where the screen current is high due to that, I turn that down too for <100mA. I think 2200V gets me 250W carriers. -but the 4-1000 does not seem to work well there.

This brings up a point that the efficiency ought not be lower, right? but it is. maybe something else is not right. There have been several weird things found in this transmitter before.

About the Q, I think it is OK. It has an continuous adjustment PI coil, and adjustable vacuum cap for the plate. I confess I do not know the Q but it seems about right, maybe 15. These two things are not ganged. If they were, maybe it would help that remain fairly constant.

All power supply controls are separate variacs.
But the modulation match may be an issue. The transformer is supposedly from a RCA KW BC unit, with 833's modulating 833's. What I have is 2.5KV on the plate and 300mA. That is pretty high load resistance. 8.3K. I run 3000V on the modulator for a little more step-up. Mod current at 100% at the higher power 600-700W is only about 500mA.  I think that is what  remember.

The best trade off is to run 2500V on the RF side, 3KV on the modulator, and then adjust the screen voltage on the RF amp to set the carrier. On TX with no modulation the screen is at 200V, but the supply is at about 600-700V no load. There is a big resistor maybe 2K in series with the poorly regulated screen supply, then a 8H choke. So that is how it is modulated. 8H is enough, it works very well. I tried 20H and 48H, did get more screen swing but little improvement in modulation quality.

For the tuning and loading, I run it full bore (3500V) to a dummy load and tune it for max power then efficiency and the last touch is re-dip the plate. Then I reduce the voltages and send a tone in, and touch up for max peaks -but nothing is generally needed there except excitation adjustment.

My antenna is a sloping, horizontally arranged 'fan' dipole, it has only 4 legs total. so it is like an X from above. electrically it looks like ><. It is cut for 80M supposedly, but it seems a bit long. It's 55FT high in the middle and about 20FT at the 4 ends. Each leg is made of OW ladder. the feed is the same ladder coming down. I don't know if that is a 'good' antenna or not but it is all I have. Is it a 'good' antenna?

[Opcom]

Its a problem on a big rig run very light.
The antenna can screw up, plate current can be excessive, and screen current will be LOW, not high.

I suppose open wire line and a tuner does not help, since the antenna can screw up and something will still be radiating, and the swr might not look that bad, plus the tuner with open wire line might not be close to the operating position, so you can not see the meters.

I avoid all that by using a resonant dipole, coax fed, no tuner, right into the station control, then a watt/swr meter.
Any change at all will show up as a different power output.
No power lost in tuners, very little in the coax, no balun's, nothing to go wrong or fail, but its not an all band with one antenna setup.
No one watches TV over the air anymore, or if they do its digital, so I have not used a low pass filter for many years.

Its sort of like flying an aircraft I suppose, you need to keep an eye on all the meters, modulation, plate current, PEP power output frquency, and do a scan.

Yes exactly, big rig run very light. If it could do over then it would be as robust but not as fussy at lower power. Instead of the 4-1000, what would be better? That is really the thing. The 4-1000 isn't happy with low plate voltages. if I raise the volts and load at reduced current to get the same power of 300W, the modulation Z is very messed up. goes very high.

The tuner is in front of me, as are the FWD+REFL power meters, dual scale ones, they are OK. One watches the excitation, and one watches the stuf going from the TX into the tuner. That's all I have, except a scope. The strange thing is that the scope looked OK when the antenna was shorted.

The TX is 10 FT away, and I can't see the meters. I have to get up and walk over there. It's not facing me. This is an issue with my layout. The operating pos. is three racks bolted together with a rackmount table top. I lined up all the racks in a pretty row when I started setting all of this up, but it has turned out to be a little impractical. Maybe I should swap positions on a couple of the pieces, but I would really like to do a semi circle or a 2/3 circle with them and have the operating position in the middle. I need my friend Jacob to come help me move things but his mom's car died and she has his right now. About tuners, i don't really like them. It would be nice if the dipole was resonant more closely on the desired frequencies.

8-)

[N2DTS]

I never had a 4-1000, so I do not know about them, but remember the globe king 500 running 1800 volts on a 4-250/4-400 got 300 watts out for 540 watts in.
I would think you could make something just as efficient at lower voltages if the Q and other things were set right, but maybe not?

Another thing about running a big rig very light is the filiment power used.
Big tubes, like using three 4-1000's might use as much power running the filiments as my 300 watt rig uses at max power output.

Maybe do some tests since you have a variac on everything and run higher voltage and lighter current on the rf deck, and adjust the modulator to different voltages to get a good match.

I really dislike high voltages, over 2000 volts is another world needing huge parts I do not have access to.
I do jam 2500 volts on my push pull paralel 100TH mod deck to get the power out with lower drive, but use the 2000 volt 4cx250b mod deck most of the time.

My setup has three matching enclosed racks bolted together on the right side of my operating desk with the big rig in it, plus other equipment in the end rack closest to the desk, home brew RX #1, 32v3 exciter, master power switch, station control, antenna tuner if I was to use one.
On the left of the desk is the 40 meter rig in an enclosed rack cabinet, and it includes a vfo/exciter, power supply, mod deck, one of two RF decks, rf control deck.
Makes a cozy U with everything close to hand.
On the desk is the audio rack (cabinet), O scope, home brew RX #2, mod monitor, audio amp, computer and sdr stuff, watt meter, big display for the sdr sits in the middle on top of the home brew receiver.
 

[Opcom]

I did some experimenting, it's right about the Q and the volt/amp operating conditions of the tube. The tuning all has to be redone. These results ought be interesting. Here are the operating details with a 200W carrier and 70% mod.. It is possible but the adjustments are touchy to get it to do so and modulate cleanly, and quite different from the happy 300-400W level and up.

Noticing the envelope, which is pre-LPF, there is something there about 100MHz. It's only a few watts? but I certainly don't want it going out. The LPF stops it.

It's related to drive level. Too much or too little and it's there. Drive has to be just right for it to be mostly gone.
And that point is not where the SWR of the TX input is good (class c stage, SWR changes with the drive level, go figure). Reflected power from the grid tank looks good on the 150W scale, but on the 15W scale the reflected needle is higher. The jaggies in there are not an instability but is more likely a parasitic tuned circuit in the plate somewhere. Something else to work on.

In all of these, the "line voltage" meters reflect the voltages coming from the 20A variacs to the pole pig primaries.
Pigs are 115V->4800VCT, 3KVA in oil, certainly less in air, but they have never even gotten warm.
The RF side line voltage meter is on the left, modulator supply line voltage meter on the right.

The DC KV RF plate supply meter is above, tan face.
the DC KV modulator plate supply meter is below, black face.
They are not perfectly accurate, +/-200V or so.

In each case of the RF power meter images, take note of the range setting on the lower one, the drive.

[Opcom]

Here are the nice 450W settings. note the drive reflected power, and the cleaner output envelope.

[Opcom]

The 400W setting, with the plate Q the same as with 200W (same settings of the continuously variable setup).
Note the much higher screen current. 100mA is a enough, for the tube. The screen is rated 75W but still.

[Opcom]

600W and 125% positive peak modulation with a single-diode keep-alive. That means during the negative peaks, the modulator drives high current right into the keep alive, and this is reflected on the modulator current meter, really loads it.. It holds fast though. The 3-500's will get orange under a long test in conditions like this. It looks ugly but people say that voice sounds great with this modulation, and that when they tune off >3.5KC they do not hear any splatter or other nasty noises from there out to many KC, it is clean. The sine wave makes it seem uglier than it is.. with voice the negative peak clipping is quite seldom, as long as the modulation polarity switch is in the correct position. The TX can be turned up quite a bit more, 800-900W, by running the plate voltage up and retouching the screen to keep it up around 80mA, but there's no point and the dummy load is only a 600W one, so..  I do not intentionally run 125%, but the headroom slack and the keep-alive is good for avoiding splatter. The transition to the keep-alive is gentle, more than it looks, the angle there is soft, so it does not create noise outside of the voice bandwidth.

In any case, this operation is with the plate supplies modified for a low voltage, 3500 max. the original way could make >5KV and had 6KV filters, but that's scary, esp. since the mod iron is from something that ran 3300V, and >60 years old. (RCA 833's x 833's.) Play nice with one's toys.