Under-modulation

[K9ACT]

Now that my 2X8000 RF deck is working properly, I seem to be short of audio power to modulate it fully.  If I back it down to the loading/plate current/output of one tube, I can get in excess of 100% and exercise the Volumax.  At over 300 w carrier, I only get about 75% modulation.

Here are some data points:

B+ for both mod and Rf about 1550v from common supply.

With sine wave input, the 813's mod tubes draw about 400 ma when the mod envelope peaks.  That's 600 watts input to the modlator.  Guessing at efficiency, that should be about 250W of audio.

Input to the RF tubes is about 350 ma for 542W input.

Mod iron is unknown make with 2403 stenciled on it.  I am using 6700-PRI- 6700 for 813's and 0 - 7700 for RF deck.

This doesn't make much sense to me as this would seem to be step up instead of step down and could be the source of the problem.

The options are for the primary... 6700 and 7900
for the secondary.... 0-2200, 6700, 7700

Is this the wrong iron for the job or could I have it hooked up backwards?

What else might I be doing wrong?

js

[k7yoo]

Try the 7900-7900 on the primary (modulator side)
and the 6700 secondary
Skip

[w5dud]

Hello with out looking at your design , it soundslike your common pwr supply is sagging at peak, if possible use one power supply for the mod deck and another for the RF Deck, some broadcast trans do this when pwer levels get over 5kw, Just my 2 cents worth, Dudley/W5DUD

[W9GT]

I believe your plate load Z is about 4428 ohms....might be a bit of a mismatch.  As for the primary.....I don't know what the pp load Z is supposed to be for 813s, I would guess higher than 6700 ohms, but I don't know.  I would try the various combinations of taps and see how it works and which direction it goes as far as effective modulation %, then optimize for the best performance.  ...obviously you are not going to have a perfect match, but you should be able to get it to work.  The ratio is more important than the rated Z s of the windings.  You should have plenty of audio to do the job.

73,  Jack, W9GT

[WBear2GCR]

Not being an expert at this sort of thing, what comes to mind is reducing the problem to its simplest state: are you running out of voltage swing from the modulator, or are you running out of current? Or both?

If it's voltage swing and not current, then you can reverse the transformer (assuming it is CT on both sides) and make it step up - although I don't know if that is "legal". 

If you are running out of current, but not voltage, perhaps the 1:1 ratio will cure that somewhat. Or some step down ratio that you can get out of it...

Otoh, one could consider using more B+ on the modulator tubes, IF they will take it, to get more swing...

But one thing that did occur to me reading ur post: modulator output = 250w, while carrier output >300watts - can one modulate a carrier fully with power that is < the carrier power??

The other thing that occured to me, is how is your negative modulation compared to your positive modulation? Can you make cut-off and not go to full positive peaks??

One thing to do is to monitor the plates of the mod tubes and see what the waveform looks like under load with the finals, and then with a high power resistor - which you can select (if you have a few around) for different resistances and see what sort of load resistance delivers how much power and that might give you a clue as to which way to go or what is lacking/needed?

The other thing, are the 813s strapped in triode or run with screen bias?
In triode they'll make a bit less power than run with screen bias, but in triode the plate Z is lower... yin/yang??

Someone else will have a definitive answer no doubt.

              _-_-bear

[Bacon, WA3WDR]

I think that this is the wrong iron to use for this job.

The impedance of the modulated stage is about 4428 ohms (1550V/0.35A).

The impedance ratio of the transformer is now 6700:7700, or about 1 : 1.15.

That means that the p-p impedance presented to the 813 is 4428/1.15 = 3850 ohms, and each 813 has half of that winding which is 1/4 of that impedance so each 813 is pulling down from 1550V through an impedance of about 3850/4=963 ohms, and that's too low of an impedance.  The plate current would be over an amp peak, and the 813 can't do that.

Switching to 7900-0-7900 on the primary side and 6700 on the secondary side would be better.  The impedance ratio would be 1.18:1 and the 4428 ohm load would be convreted to 4428 * 1.18 or 5525 ohms, and each tube would see about 5525/4 = 1306 ohms.  That is better, but it's still too low of an impedance for the 813s.  The tubes will run out of emission.

Using 6700-0-6700 on the primary side and 2200 on the secondary side would be a better ratio, but the full winding voltage would be high, and you probably still would not make 100% modulation either, even if you went to fixed screen voltage  (maybe it would just make it with 500V on the screen).  Also the bottom end would roll off a bit high, because the load impedance would be two times what was expected. 

As Bear was saying, you are running out of current capability now, and you would be running out of voltage capability the 6700-2200 way, plus the bottom end probably isn't going to be as good, and the windings would see a lot of AC voltage.

If it had a 3500 or 4000 ohm secondary tap, you could use that.  But you can't get the right ratio even between the 2200 and the 6700 or the 7700 output tap.

Four 813s in push-pull parallel with a 7900-6700 ratio would be better, but it would be a lot of tubes, a lot of filament power, a lot of grid drive and a lot of plate current.

[k4kyv]

As I recall from years ago when I used 813's triode connected, the optimum p-p load impedance was high, something like 20,000 ohms.  They should work at lower impedances, but I think yours is much too low.

Try connecting the plates to the 7900 ohm taps, and then try each of the three secondary load impedance taps and see which one gives the highest percentage of modulation with the least distortion.  The 6700 ohm load might work ok.

The 813's might like it better if you ran something like 2250 volts on the plates, and used a transformer with more step down, keeping 1600 volts on the plates of the final.

With a common power supply, the turns ratio of the mod xfmr should be somewhere between 1.2:1 and 1.6:1, in order to be able to achieve 100% modulation with some  headroom  left over.  If the step-down ratio is more than 1.6 :1, you will not be able to achieve 100% modulation, regardless of what impedance you running the tubes at. 

Some transmitters use a ratio like 1.7:1 in a deliberate effort to make it impossible to overmodulate the final.  However, driving the modulators into saturation causes exactly the same kind of splatter and distortion as overmodulation, so that doesn't accomplish much.

[WBear2GCR]

Bacon's comments made me think about something here.  (rare occurrance?)

Since the secondary is used single ended, and it does have a center tap (?), you might be able to get away with using half of the secondary (the largest half secondary you can manage) rather than a full secondary to get a lower secondary impedance...

The idea is to get more than 2200 on the secondary, but less than now, and maybe just enough to make the ratios work... and again, the 813 strapped as a triode will have a lower plate Z than when run with a screen supply, that might help too...

More Dilithium Scotty!!

                     

                            _-_-bear