Modulation XFMR Ratio

[WA5VGO]

I'm building a new transmitter. The PA is a 4D32 and the modulators are 809's. B+ will be about 750 volts. The modulation transformer ratio calculates to be 1.57. The transformer I'm using has ratios of 1.4 and 1.6. Which ratio should I use? I'm leaning towards 1.4.

Darrell, WA5VGO

[k4kyv]

That would be better.  It would give you more headroom in the modulator and better positive peak capability.  With the 1.6:1 ratio, using a common power supply, you would most likely just barely be capable of achieving 100% modulation.

[N2DTS]

Sounds like a fun project!

The 4D32 is a good tube, they like lots of voltage, I run mine at 1500 volts!
With a pair of 809's I would run the rig at 1200 volts.

Why not add a 2nd 4D32 and use 811's?
200 watts carrier and 800 watt pep..

Brett
 

[AB2EZ]

The folks who have been building plate modulated rigs for decades know this stuff like "the back of their hand". Some comments... which will be obvious to those folks... but which may be of interest to others who are building plate modulated rigs (of just thinking about it):

To a first approximation, what is important about a transformer is the turns ratio. That is, if you have a transformer that is specified as having a "14.4k ohm input impedance and 10k ohm output impedance", it is (to a first approximation) not important that it actually be used with a 14.4k ohm source on one side and a 10 k ohm load on the other side. 

The turns ratio of a transformer is the square root of the impedance ratio. So in that example, the turns ratio is 1.2: 1. [I.e. 1.2 x 1.2 = 1.44].

If you put a 840 volt amplitude sine wave on the input side, you will get a 840 / 1.2 volt = 700 volt amplitude sine wave on the output side (Note: you can also use the transformer in the opposite direction)

If the impedance ratio is 1.44:1 (as in this case), and you put a 5000 ohm load on the output side (e.g. the modulation impedance of the r.f. stage that is being modulated), then the modulator will "see" a load of 5000 ohms x 1.44 = 7200 ohms.

If you imagine yourself back in 1940 or 1950, and you are designing a plate modulated transmitter to achieve the best performance at the lost possible cost... and if you could use any type of modulation transformer you wished to use... then you would select a modulation transformer that was optimized for the most efficient coupling of audio power from the modulator to the modulated rf output stage. The ratio of input impedance-to-output impedance of the modulation transformer would be selected to match the ratio of modulator output impedance -to - modulated rf stage modulation impedance. You would also take into account such things as whether you wanted to be able to achieve 90% modulation, or 100% modulation, of 125% modulation... and whether you were going to use one power supply voltage for both the modulator and the rf output stage... or different power supply voltages for each of those.

If you are building a plate modulated transmitter today your objectives might be totally different. For example, you might be willing to use a modulator capable of delivering 1000 watts of audio (into a matched load), even though you only need (for example) 400 watts of audio to 100% modulate the r.f. stage.  You might do this just because you always wanted to build a modulator with a pair of 4-1000's in push pull... and you don't care about how much electricity it takes to light the filaments. In that case, you could use a modulation transformer that causes the load impedance on the modulator to be substantially higher than the matched load value. In addition to realizing your dream of a pair of 4-1000's behind a peak-a-boo window... you might get other benefits from this "inefficient design".... such as a better low end frequency response for your modulated r.f. signal (due to the lower modulator output impedance, as seen by the modulated r.f. stage... looking backward toward the modulator), and lot's of reserve audio power modulation peaks (Careful! Don't get carried away and blow up the modulation transformer or the output r.f. stage). As Don pointed out, this approach would probably require a higher voltage B+ supply for the modulator, than the B+ on the r.f. output tubes... but that might not be a concern to you. One of our East Coast AMer's (KC2JXX) used a single plate transformer to drive two power supplies in parallel: one is choke input and one is capacitor input. The capacitor input supply should put out (roughly) 1.25/0.9 = 1.39 x the voltage of the choke input supply.   

Those of us that have used modern 8-ohm output impedance audio power amplifiers to modulate vacuum tube r.f. amplifiers typically use an audio power amplifier with lots of extra audio output power capability (but not too much!), so that the impedance matching is not critical. In my plate-modulated GS-35b rig... I run the tubes at 1700 volts and 300 mA (at carrier) ... so the modulation impedance is 5667 ohms. I use a 25:1 turns-ratio toroidal ferrite power transformer as a modulation transformer [Heising mode]. I have no idea as to how to characterize its input impedance and output impedance... except that I know that the impedance is high enough at low frequencies to produce a flat modulation response down to below 30 Hz. So the 5667 ohm load of the r.f. stage is converted by the transoformer to 5667 / (25 x 25) = 9 ohms. I drive this with a Samson 600 watt (every manufacturer seems to rate their amplifiers a little differently) amplifier, with a recommended output load of 8 ohms. It works fine with a higher impedance load, but the amount of audio power it will deliver is roughly inversely proportional to the load impedance. [This amplifier acts like a voltage source... so its output power is v^2 /R]. I only need about 255 watts of average sine wave audio power to modulate the r.f. amplifier 100%. The 9 ohm load is actually a pretty good match for this amplifier... but it certainly isn't critical... given all the excess audio power I have in reserve. I could put a 16 ohm load on the amplifier, I it would still deliver about 300 watts of audio power. [Anyway, it works fine and it sounds fine... according to the audio reports I receive]

Best regards
Stu

[ke7trp]

Great Post STU. 

We are doing something similiar with my Transmitter.  I am using two large ratio Transformers to couple the 4-400s to the Audio amplifier. I will post pictures once the deck is complete.

The 4D32 and 809s sounds close to my Globe Champion 300A.  It uses two 9909s in the final but some of converted to the 4d32 as the 9909s are hard to find. This rig does 220 watts carrer and 800PEP and has for a couple of years now. Super loud on the air.

Clark

[steve_qix]

I'm building a new transmitter. The PA is a 4D32 and the modulators are 809's. B+ will be about 750 volts. The modulation transformer ratio calculates to be 1.57. The transformer I'm using has ratios of 1.4 and 1.6. Which ratio should I use? I'm leaning towards 1.4.

Darrell, WA5VGO

Ok, the ratio - I'm assuming you mean the impedance ratio    Please let me know if that is wrong.  Assuming the impedance ratio is what is 1.4:1, then the TURNS ratio will be 1.18 (more or less) to 1.  That is a good ratio if you use a common power supply for the modulator and RF ampflier - and you are using triodes, which is even better.  To wit:

The voltage step-down, full primary to full secondary will be 1.18 to 1 and figuring this using half the primary (which is how the modulator tubes are connected - each works against 1/2 the primary), the step UP ratio will be .59:1.

A 750 volt supply is assumed.... if you can drive the tubes fully, you will develop about 700+ volts across one side of the primary which will result in 1267 volts (peak) across the secondary.  This will be added to the 750 volts DC already present, giving a total peak voltage at the highest positive modulation of 2017 volts.  That will give you a positive modulation capability of about 168 to 169 % - not bad !

Of course, all this is assuming you have good modulator tubes, and can drive them sufficiently to bring the plates to near saturation and achieve the highest possible voltage swing.

Be sure to include the 3 diode negative peak limiter!!! (click on this link to see the peak limiter schematic) http://www.classeradio.com/3-diode.jpg

I had a big transmitter (back in the days of vacuum tubes    )  that used a 1.4:1 ratio mod transformer, and I could easily achieve 150% positive modulation using a pair of 833A's modulating a pair of 450TL's.

Regards,

Steve