Under modulation

[K9ACT]

Had a nice QSO on 7290 with W7ISJ tonite with my more or less working 811 rig.  'Really nice audio but not enough of it" was the latest report.

I have seen the full/under/over mod pics in handbooks a zillion times but there is very little information on what to do if it isn't right.

All the ugly stuff from previous discussions is gone.  What I have now simply seems to be not enough.

With a sine wave input, the positive peak starts to flat top at about the 75% point and at about the same point, the downward peak is also reached and starts to spread sideways.

I guess I really don't know what this is telling me much less how to fix it.

Can someone please explain.

Thanks,

Jack K9ACT

[k4kyv]

I am assuming this is a plate modulated rig, not an AM linear or grid-modulated final. If this is the case, I'm assuming that the modulator tubes are good.

Sounds like your modulator is not capable of generating enough audio power to modulate the transmitter 100%.  At 75%, something (probably the modulator tubes) has reached saturation, and driving the stage further with more audio, simply causes the waveform to clip.

First, make sure you have enough grid drive to the modulators.  This clipping could be occurring in the audio driver or speech amplifier.  If you have adequate grid drive, then there are two ways to increase audio power output from the modulator stage: increase the plate voltage on the modulator tubes without increasing the plate voltage to the rf final, or reduce the step-down ratio of the primary:secondary of the modulation transformer.  In any case, make sure you don't exceed the rated plate dissipation or maximum plate current rating of the modulator tubes.  The modulators will stand a certain degree of plate voltage beyond the manufacturer's rating, if the other parameters are not exceeded.

[WD8BIL]

Jack, are you using the sinewave to tune the amp ? Ya might find the tops of those peaks in the load control setting.

[Bacon, WA3WDR]

If you have enough grid drive on the 811 modulator, try few different impedance ratios on the mod transformer.  I think we had thought of about 10K plate to plate to about 7K.  You were experimenting and came up with 9K to 3.2K which is almost 3:1 impedance, but the voltage ratio was 1.5:1 which should have been an impedance ratio of 2.25:1, so the impedance should have shown  7.2K (3.2 * 2.25) to 3.2K, not 9K to 3.2K, so now I'm wondering if there are shorted turns in the mod transformer secondary.  That would be a disaster.

Try selecting taps for 10K and 7K, and see if that helps.  Connect the taps as prescribed, and remove any jumpers from other unused taps.  I'm hoping that the modulation transformer is OK.  Assuming that it is:

The modulator tubes only have so much voltage swing, so the turns ratio (which is the same as the voltage ratio) determines how much output voltage swing they can possibly produce.  You need an output of two times your final amplifier B+, peak to peak, for 100% modulation, and you want maybe 1.2 to 1.8 times that for decent to heavy voice asymmetry.

But there are limits.  The impedance presented to the modulator tubes determines how much current will flow at the necessary output voltage, with a lower load impedance resulting in more output capability (up to the modulator tube emission limit), but higher current for a given output voltage.

If the impedance presented to the modulator tubes is too high, the tubes will run out of voltage before they can fully modulate the PA.

If the impedance presented to the modulator tubes is too low, the tubes will run out of current (due to an emission limit, assuming adequate grid drive), before they can fully modulate the PA.

At any given modulator B+ voltage, there is an optimum load impedance where the tubes are giving the maximum current they can, at their maximum available voltage swing, at maximum output.  This impedance needs to be transformed to the load impedance of the PA, hence the impedance ratio of the modulation transformer.

In your case, there is the additional limit of power supply current capacity.  So although you could select a very low impedance ratio that would produce lots of audio, the power supply might not be able to provide the resulting modulator current in normal operation.

It's my guess that an impedance ratio of 10K plate to plate, to 7K load, would work well.  That would be a voltage ratio (turns ratio) of about 1.2:1.  (Note: I had written 2:1 here earlier; that was a mistake.)  So if you put 1VAC on the secondary, you would see about 1.2VAC between the modulator plate connections.  If the resulting modulator current is too high normal in voice work, then try a higher plate to plate impedance, maybe 12K, etc.  If the modulator current is low in normal voice work, and you want more audio, then try a lower plate to plate impedance, maybe 8K, etc.

But if the modulation transformer has shorted turns, you're sunk.  The clue is low output despite high modulator current.  Of course this can happen with incorrect impedance taps too, so you need to check that in a new design.  If your modulator current is low at maximum modulation, then your problem is probably inadequate modulator grid drive, and/or too high of a plate to output impedance ratio in the modulation transformer.  Here's hoping, because those are inexpensive to fix.

[flintstone mop]

A lot of good points to check in your search for the clipping. Leaky caps bringing audio to the grids of the mod pubes could be a problem or the mod transformer secondary is saturated. Can the mod iron handle D.C. current for the plates of the RF final + modulation audio???
Keep us informed.
Fred

[K9ACT]

I tried the easiest fix first and it seems to have worked.

I re-strapped the mod iron.  I am now using the 9K to 5.5k connections which gives me 1.25:1 voltage ratio.  This was 9K to 3.2K and 1.5:1 VR.

The positive peaks go right up (almost) to the 100% line using a sine wave.

Things aren't  quite so clear under voice mod but I think I will take the next step and build up a separate power supply for the modulator.  The plate current drop under modulation is much more obvious under sine wave modulation.

Thanks for the pointers.

js

[Bacon, WA3WDR]

OK.  I would expect more than 100% modulation capability, though.  Are the negative peaks flattening at around 100% mod too?