Building a modulator

[Opcom]

Gotcha.

I am sure you realize that the freq response of the old UTC mod iron will be somewhat less than the "8 ohm"
iron that you will be testing with.

If I were building it, I'd so some things to limit the passband of the circuit of the modulator, on the low end and on the high end, as compared to the hi-fi bandwidth of the Altec.

                              _-_-

That makes sense, it is not BC iron. Thanks to LTSpice it should be easy to insert such a filter anywhere.
Isn't it good practice to set the bandwidth in the voltage amplifier stages before the driver?

[W4RFM]

Progress. Also put in the diagram for the modulator.

[KA2DZT]

Looks good,  schematic shows the CT of filament xfmr connected to the cathodes of the driver tubes.  Why??  doesn't look right.

Fred

[W9ZSL]

Bob, I've had another thread going about a speech amp/driver.  Since we're both building basically the same modulator around the S-22, I Googled the schematic and it looks like it would work with my project. What are you using for driver iron? Seems to me 6L6's could be replaced with 6V6's but other than that, it's pretty sweet.  I'm also wondering about the 4700 resistor in the feedback loop considering the original Altec took the feedback from the 16 ohm winding of the output transformer.

[W9ZSL]

Here is the complete diagram of Bob's driver and YES, that filament xfmr center tap to the 6L6 cathodes is correct.  See the note under the specs.

[W4RFM]

-Most of you folks know more about circuit theory then me, but I assumed that was some sort of bias for the 6L6's.
I have built this amp twice before for hi fi use, and it works well, so my plan is to just change the output transformer, cut the voltage back a little and let it drive those 811a bottles.
Mike,
 I thought I would try to find an appropriate place to get that feedback.  I have also tried that circuit without feedback, and it still sounds good.

[w8khk]

-Most of you folks know more about circuit theory then me, but I assumed that was some sort of bias for the 6L6's.
I have built this amp twice before for hi fi use, and it works well, so my plan is to just change the output transformer, cut the voltage back a little and let it drive those 811a bottles.
Mike,
 I thought I would try to find an appropriate place to get that feedback.  I have also tried that circuit without feedback, and it still sounds good.

The push-pull 6L6 tubes derive their bias (25 volts, in this case) across the cathode series resistor.  In audio amplifiers, it is common practice to elevate the center tap of the 6.3 volt filament winding to a positive potential, to reduce hum.  The net effect is to raise the filament of the microphone preamplifier stage positive with respect to the cathode.  This prevents the cathode from attracting electrons emitted from the filament. 

Think of the filament and cathode of the 6SJ7 as a diode.  If the cathode is negative with respect to the filament, then current will not flow from the filament to the cathode, thus reducing hum modulation from the 6.3 volt AC heater supply.

[W9ZSL]

Bob, I thought that might be the case because even though the feedback resistor taps into a higher secondary impedance, the power output is the same regardless of the actual load be it a speaker or a grid.  I think we are basically building the same thing.  That Altec clone looks like a great driver to me.  I looked it up and have the complete schematic.  The cool thing is the power supply tap for a mic preamp.  I have the diagram for an Ampex MX-10 preamp with an input transformer to match.  Rock and roll!

[Opcom]

If you want a better distortion performance without additional cost and can supply 5A for the pair's heaters, it could be as well to sub in 6CD6's for the 6L6's.

The regulation over the half cycle of audio at the 811 grids will be better because 6CD6s will deliver much higher current at low plate voltages and allow more plate voltage swing because they need only 150V on the screens.
6L6GC G1=0, G2=300, Eb=150: Ib=240mA
6CD6 G1=0, G2=150, Eb=150: Ib=475mA

The bias voltage will be in the same basic range and fixed bias will help make a few more peak watts in the driver if you have the time to do it.

The overall FB resistor 4700 Ohms, it may be worthwhile to experiment with a 10 turn pot and use the most NFB possible before instability or lack of gain comes in. It will help the regulation.

In case of instability or problems due to the combination of variable impedance of the 811 grid interacting with the driver transformer characteristics, the feedback could be taken from the 6L6 plate before the transformer.
  With or without blocking cap depending on how it might affect the first tube's cathode bias,
  Using much larger R like 200K bypassed with 20-80pF to tailor the HF response) before the transformer.
example: http://www.montagar.com/~patj/rcami-12188a_03.gif

Sweeping the audio using a resistive load (instead of the grids) can help set up the response and the gain. Response should be flat at least an octave above and below the bandwidth you want. Then verify it's still good enough with the 811's in place.

[Steve - K4HX]

You may find these useful. From RCA Ham Tips.

[W4RFM]

Steve, thank you.  That is excellent information. I knew I would eventually cut back the B+ since only 10-12 watts of drive is necessary.  I realize I can jut lower the input gain but I would rather kept the noise figure in check as well.  This is good stuff, thanks again.

[Steve - K4HX]

Good luck with it Bob. Hope to hear it on the air in the future.