Half power wiring

[stretch2011]

Just curious to what all of you think is the best method for a half power switch.

For a 4x el84 amp cathode biased, is it better to completely lift one pair of cathodes from ground (left drawing) or parallel 2 resistors to half the resistance (right drawing)?

[martin manning]

The circuit with the DPDT switch does nothing since the two 130R are always connected in parallel, with all four tubes sharing 65R to ground. Halving the cathode resistance would not be a good idea, as the idle current would be far too high. The other circuit which lifts the cathodes of one p-p pair will work fine. Don't forget to double the speaker load.

[LeftyStrat]

I think you'll be disappointed with the results of half power mode. It doesn't equal half-volume. A ten watt amp is half as loud as a 100 watt amp.

Half watt mode mainly shows as less headroom and bass.

[stretch2011]

Martin: Ah I should've looked over it better before I posted. Just drew it on the fly. Would you have to change the speaker load since the plates would still be connected?


Lefty: Duely noted. I probably would never use it, but people seems to like the thought of it. I havnt used one so I figured I'd try it out.

[xtian]

You can try it out just by pulling two of the tubes!

[stretch2011]

Xtian: Now that would def. Need the speaker tap switched to compensate for the missing plate loads. Aside from that, good point.

[sluckey]

Would you have to change the speaker load since the plates would still be connected?

Yes. Disconnecting the cathode behaves the same as pulling the tube. But since the idea is to lower power, leaving the speaker load the same provides a mismatch that will further reduce power.

I would not worry too much with a 4xEL84 amp.

You may consider just disconnecting the grid drive to a pair of tubes. Does the same thing without changing the impedance.

[martin manning]

You may consider just disconnecting the grid drive to a pair of tubes. Does the same thing without changing the impedance.

I have to disagree there. With only half of the output tubes responding, the load impedance will appear to be halved.

[stretch2011]

Ok so there is a conundrum here.

I understand the plate needs to be connected for the ot to see the impedance. Now the cathode also needs to be connected for the plate to see the impedance? Why is this? Is the plate's impedance seen through connecting to the cathode to ground?

Now on to martin's disagreement to sluckey's statement about the grid. I have no idea why disconnecting the grid would change the plate load. Granted I'm still at least one of two shades of green. Please enlighten me. Is the system in a whole (plate, grid, screen, cathode) in operation what gives the plate its impedance that the output transformer sees, or is it individual or multiple connections in the system.

[martin manning]

As in any transformer, the secondary load only appears at the primary when there is a change in primary current. If half of the output tubes are prevented from modulating the primary current (either by disconnecting the drive signal or disconnecting the cathode ground) then the change in primary current is halved. The secondary impedance will then have to be doubled to restore the original current-voltage relationship seen on the primary.

[stretch2011]

That clears up a lot. Thank you Martin.

So would a more effective half power switch be turning the pentode into a triode? I've seen this done with el34's in fixed bias, and in a single ended cathode biased el84 amp.

I'm rethinking the half power switch more now. But I would still like to know.

[sluckey]

After thinking about it I agree with you Martin. Sorry for the misinformation.

[stretch2011]

Wonderful! A conclusion has been reached.

Thank you everyone for the input. Special thanks to Martin.

Upon seeing that the half power switch is a not ALL that great idea, I am thinking of adding a passive loop instead. Gotta have something extra