6V6 Voltage divider G2

[Mats]

This time Im having a look at a Challenger (Bogen) CH18 amp.
The two 6V6 have their G2 connected to a voltage divider (15k and 100k).
The divider is then tied to the common cathode res and cap.
Pros and cons for this ?
/Mats

[HotBluePlates]

Pros and cons for this ?

Repeating my usual suggestion: If you're thinking about modifying this into a guitar amp, I'd suggest keeping everything from the 6SL7 (phase inverter) through to the output transformer, and connected parts of the power supply stock. It's already well-designed to work together.

The two 6V6 have their G2 connected to a voltage divider (15k and 100k).
The divider is then tied to the common cathode res and cap.

Notice also there is a screen voltage filter cap, not just a voltage divider.

Bottom-line, Bogen chose to use a screen voltage below the plate voltage for the 6V6's. We don't have a screen voltage indicated; based only on the 6V6 plate & cathode voltage, as well as the 15kΩ & 100kΩ resistors, I would guess the screen is sitting ~360vdc while the plate is at 420vdc. The math would say 367v at the screen, but I suspect a little more is dropped across the 15kΩ for idle screen current draw.

A designer would pick the screen voltage with an eye to the peak plate current they need for the power output desired, OT primary impedance used, and plate supply voltage available. Obviously, Bogen felt they could get the 6V6 to deliver the power needed with the OT present while having the screen at a voltage less than the plate voltage.

Lower screen current also leads to less bias voltage needed for a given idle current target. If the bias voltage is small, then the drive signal from the phase inverter can also be smaller while still driving the amp to full output power. So the output section may be seen as having "greater power sensitivity".

• Said another way, if screen voltage were raised to plate voltage in this amp, you'd need more bias voltage to tame the idle current, which means you'd need a larger cathode resistor, and corresponding higher driving signal to get the same output power you have now; the higher screen voltage might provide more plate current capability, but you wouldn't be able to take advantage of that without changing the OT to lower primary impedance. That's a lot of cost & effort for perhaps only a watt or less of added power output.

Returning the screen voltage divider to the top of the cathode resistor also adds to the current passing through that resistor, increasing the bias voltage (by ~1v in this case). If Bogen really, really  wanted 21v of bias and didn't use the screen divider as they did, the cathode resistor would need to be ~316Ω and some change...

There are some other effects I can imagine, but I don't think they would have much impact in this case. Either way, I'd want to test & measure with the 100kΩ in/out of circuit to know for sure...

[Mats]

Thanks for reply.
No I dont have that amp, just had a look at schematic (still try to learn).
Yes, it would be interesting to test in/out the 100k (and also that filter cap)
and leave that 15k to screens.
/Mats

[HotBluePlates]

... it would be interesting to test in/out the 100k (and also that filter cap) and leave that 15k to screens.

No, as-designed, the 15kΩ and 5uF cap are integral to the am's circuit. They are the series dropping resistor & filter cap which make up the screen node of the power supply. Take them out & the amp doesn't play.

So look at those parts as a parallel branch off of the main power supply, not unlike a Vox power supply. This amp does not have series screen resistors, as you'd typically find in a guitar amp. If you feel you need them, then add them.

The only curious feature is the 100kΩ to the output tube cathodes. Ordinarily, old-tube practice would say this is a bleeder on the screen supply to help set & stabilize the screen voltage by way of pulling current through that 15kΩ resistor. Except the 100kΩ is too-big a resistance to do that effectively. I can envision a way in which such a connection would work with cathode bias to keep the output tubes well-controlled, but I'm not sure how much impact it would really have (which is why I said I'd want to test the concept by probing the actual amp to see).