811 project

[K9ACT]

I finally got the 3rd tube I needed to complete the rig and fired it up today after jumping a few hurdles.

It worked a lot better than the 40 meter band.  I called CQ and listened off and on from about Noon till 10PM and made exactly two contacts, both of whom said it sounded ok and great.

I need some more feedback to determine precisely what "ok and great" really mean.

The only obvious problem is that I can’t seem to get 100% modulation out of it.  The envelop peak never gets much beyond about  50% above the carrier peak and there is lots of audio left when I reach the negative peaks.

This is probably a dumb question but, is this one of the symptoms that the mod trans is not hooked up right?

I set it up as stated earlier with a 1.5:1 ratio with the intent to diddle if it didn’t work.

For those familiar with the iron, I have 1 and 2 on the plates with 3-4 for center tap on the primary.
The secondary is 9 and CT to B+ and 10 to the PA plate with 7-9-10-12 tied together.

This is supposed to be 9K:3.2K and the measured voltage ratio is 1.49:1

One other issue is that if I swap the grid connections at the mod deck, it squeals when I turn up the audio.

Any thoughts,

js

[Bacon, WA3WDR]

RF in the audio will cause that squeal thingie with the polarity reversal.  Put the TX on a dummy load, if you have one, and see if that fixes the squeal issue.  If it does, then stay on the dummy load and set yourself in the right polarity if it makes a difference.

Try the antenna again and see if you can puzzle out where the RF is getting onto the audio - usually it's the mike section, because there is where it connects to the outside, and it also has a lot of gain.  Little bypass caps and small millihenry range chokes and sometimes some ferrite cores are the answer, usually.  Sometimes different or better mike shield grounding helps.  But caps and little chokes can do wonders.

It's good that you can get 100% negative.  Can you set up a trapezoid pattern on a scope?  Mod audio on the horizontal, modulated RF on the vertical?   That will show modulated stage nonlinearity.

Unless the mod transformer is unable to accept the plate DC, which is possible, the positive flattening is most likely insufficient RF grid drive, or weak output tubes... uh, or something else.  Try some more grid drive and see if it helps.  Is the plate current dropping with audio, more than the plate voltage is?  You're on one supply, so it will sag a bit.  But if the plate current clearly drops more than the plate voltage, then the finals can't handle the peaks; maybe because the grid drive is too low, or maybe because the finals are soft.

[K9ACT]

>RF in the audio will cause that squeal thingie with the polarity reversal.

An RF choke and by pass did fix that.

>It's good that you can get 100% negative.  Can you set up a trapezoid pattern on a scope?  Mod audio on the horizontal, modulated RF on the vertical?

Never quite worked that out to produce what the book says so I just look at the envelope

>Unless the mod transformer is unable to accept the plate DC, which is possible, the positive flattening is most likely insufficient RF grid drive….

Not sure what you mean by “accept”.  The grid drive produces the correct current at the specified voltage so I assume that can be eliminated.  I tried more and less and it didn’t make any difference that I could tell.

Is the plate current dropping with audio, more than the plate voltage is?

The voltage drops about 15v and the current drops about 5 ma on when I talk it up so that does not seem too serious.

So, back to my question about the mod trans hook up.  Does this effect the mod percentage?


js

[Bacon, WA3WDR]

It can; there is only so much DC to work with on the primary, so the primary to secondary ratio affects the maximum output voltage possible, and therefore the primary to secondary ratio affects maximum modulation capability.

But it should not affect the positive versus negative capability.  The DC acceptance I mentioned is the magnetic saturation capacity of the mod transformer.  DC adds to the positive magnetization and then positive modulation adds more to it.  If the core material saturates, you could have flattened positive peaks, although it tends to be a messy flattening.

You can sometimes get better modulation by tapping down on the primary and secondary windings so that there are not as many amp-turns in the secondary winding, and therefore not as much DC magnetic flux in the core.  However, you will lose some bottom end response, because you will also have less inductance, and also there gets to be more voltage swing on the ends of the windings.

Try a little more grid drive on the modulated stage, and see if it makes a difference.  Also try a little more or less filament voltage on the RF finals, and see if you are close to emission saturation - 5% higher or lower filament voltage should not affect anything significantly.

Also, what is the mod transformer?  Does it have a "gap" in the core, or is it cross-laminated?  It should have a gap in your design, I believe.

[K9ACT]

Also, what is the mod transformer?  Does it have a "gap" in the core, or is it cross-laminated?  It should have a gap in your design, I believe.

It is a UTC SS22 and I know nothing about a gap other that the spark gap I put across the HV on top.

js

[Bacon, WA3WDR]

I'm almost certain that an S-22 has a magnetic gap.  The problem is somewhere else.

Just as background: power chokes and modulation transformers have a magnetic gap in their cores.  You see a straight line break between the E and I laminations.  This gap allows the transformer to handle a whole lot more DC than it possibly could otherwise.  It also reduces inductance, compared to a cross-laminated core such as you see in power transformers and typical push-pull audio output transformers.  However, the decrease in inductance is small, compared to the huge increase in DC handling capability.

Amateur mod transformers like the S-22 are all built with this gap.  Broadcast mod transformers are usually built with cross-laminations to get more inductance and therefore require smaller windings, with less leakage inductance, stray capacitance, etc, and the DC is kept off of them by using a separate inductor, called the modulation inductor, to handle the DC.  A surprisingly small coupling capacitor, like around 2uF depending on inductances and impedances, is typically optimum for coupling the modulation transformer secondary to the modulation choke and RF final.  Even a gapped mod transformer can benefit from this, but sometimes hams will remove the laminations and put them back in cross-laminated to get more inductance and bottom end when the transformer won't have to carry unbalanced DC.

OK, maybe you have a soft modulator tube.  Try swapping the two tubes; if the positive modulation is better, then that's the problem.  Maybe you can leave it that way, but it would indicate that one of your modulator tubes is bad.

If you have something that can give you an audio sine wave, you can use that as a signal source and then make a resistive divider and look at the modulated B+.  If you see a good sine wave there, but your modulated signal is no good, then your RF amplifier is not linear.  If you see a seriously distorted sine wave,  then the problem would seem to be the modulator.

I assume that this is still the modulated 811s... with no screen grids.  If you used 813s and you did not allow the screens to get modulation somehow, it would behave as you describe.  But if you are using the 811s, as I believe that you are, then this is not the problem.

If the RF amplifier has too little grid current, it might not be able to support positive modulation peaks.

If I recall, you are using two 811s modulating one 811.  Or maybe two...   Anyway, try swapping the 811s in all possible combinations and see if anything helps.

Make sure you have 6.3V on the 811 filaments.  If somehow they are getting 5V, it could cause something like this.


   Bacon, WA3WDR