Hello,

I have some questions about amplifiers.

I have seen many amps with negative feeback that use a RC network from the speaker to the cathode of the driver tube. Most of the sets i've restored, has no feedback network.

But this one is strange to me.
A 1 Megohm from plate of 6V6 to the input of the driver.
Is this a positive feedback ?

I took off the 1 Meg and noticed no difference in sound quality. May be at high volume, or in frequency response i would see a change ?

Also most of amplifiers that use cathode resistor bias on the 6V6, also use a bypass capacitor of 20uF/25v.
In this amplifier, there is no capacitor bypass.
However the bass audio response sounds very good to my taste. I just wonder why is some circuits, the capacitor bypass is not installed ?

Last question is concerning the high value resistor we see at the grid of tubes. I saw many 6J5 that has values from 100k to Megohms.
In this case it's a 330k .
Is this value critical ?
what would happen if we double that value ?

Thanks
Eddy

What did this circuit come from? That may help answer some of the questions.

:What did this circuit come from? That may help answer some of the questions.
:

Hi Brian,

This circuit comes from a Viking combo radio set.

I have not been able to find the schematic of it.
So i traced the wiring and drawn the schematic myself.

At 1 meg, that resistor can't be doing much. AFAIK, Both audio signals would be in phase on either side of it. Maybe one the theory guys will help here. Would be interesting to see the differences in plate voltages on the 2 connected tubes, with and without the 1 meg. Also, remember that these sets were wired by average laborers, and that 1 meg resistor could have been a mistake..unless we can see the original schematic.
AFA the cathode cap...They may have left it out as a money saver..which was a common thing then..maybe this probably wasn't a true HI-FI set and didn't need it.
You could try different value resistors in place of the 6J5 330k grid resistor just for fun, and see what you get. I'm sure they chose that value to give a decent grid bias to work with the cathode of the phase inverter...again, the theory guys may have the proper answer.

The 1M resistor appears to have no useful function. From the grid of the "phase inverter" tube to its plate, there exists signal phase inversion, and from the grid of the output tube to its plate there is also phase inversion- so across the 1M resistor there is no net phase inversion, just phase shift due to the propagation delay through the stages.

Connected this way, the 1M resistor would therefore provide "regenerative feedback" which is generally a bad thing for audio amps. Apparently the output impedance of the driving stage is low enough to make the presence of this 1M resistor unnoticeable.

The RC feedback nets that you described provide degenerative feedback which improves tone quality at the cost of some audio output power. They are usually only present in sets that have some pretensions to "hi-fi" sound.

Not using capacitive bypassing in the output tubes' cathode circuit simply means that some of the audio (AC) power that would otherwise have been coupled into the output transformer will be dissipated in the cathode resistor, along with the tube DC current. Cap bypassing of the cathode resistor allows maximum signal power to be coupled into the output transformer instead of being partially lost in the cathode resistor. This improves operating efficiency.