Paraphase to fixed bias?

[EFK]

I actually understand that - thanks! I can use my LCR meter to put a 120hz signal into the amp, so when I measure AC voltage on the 6L6 grids, do I just ground the black meter lead? Well I'll go find out!

[octal]

Assuming your LCR meter puts out something resembling a sine wave at a level appropriate for a guitar amp, yes. Measure the voltage from the grids to ground. I'm assuming you have a high input impedance (10meg?) digital multimeter as well. Good luck!

[Bruce / Mission Amps]

Go look at a an old VOX AC50 or AC100 that uses a similar phase inverter and blocking caps. It is super simple to do this with a fixed bias power stage.
However, my 32-20 harp amp uses a paraphase driver and I do have to have one extra 22uF-100v filtering stage on the end of the bias supply to keep the hum down, as previous mentioned by Loudthud. Whether or not it is for the exact reason is arguable but it is a fact and I do know how to build a bias supply, ha ha....

[octal]

Bruce,

I think the version of the AC 50 in question has a floating paraphase, so that's a bit different than the 'original' style paraphase used in the Supro/Valcos.

Nathan

[redelephant]

I learn someting new all the time
Not to be bugging u about this, and since ure done I guess I'm just posting for discussion...and I'm bored and willing to bore u too...I have no experience with the paraphase, and some day I'll have to try it. But want I meant about creating the LPT was using the negative bias for the tail, and then u can place the grids at ground, and the PI will have just as much swing as the paraphase. I think...havent actually looked closely at the paraphase yet...
This is different than the typical LPT that use no neg rail and therefore have less b+ to play with. . . isn't it?
Not intending to drag this into some long ranty thread about which PI is best, just checking my own six.

[Steve Conner]

I think that works as long as you don't use NFB. The DC level of the cathodes/tail is wrong to connect the NFB loop to it. And you have to be careful putting a DC block capacitor in the NFB loop, it'll eat into the phase margin and maybe make it motorboat.

I don't like the paraphase because it throws away most of a signal and then amplifies it again. I prefer the floating version if I have to use it at all, but my favourite for musical instrument use is the LTPI.

[redelephant]

thanks...but the grids are at gnd level, how is that wrong DC level for a NFB connection? If the NFB is from somewhere other than the output, such as from cathode or somewhere special I see ur point, but isn't the NFB usually from a DC of about zero volts?
Btw, I almost never use loop NFB so have never given loop problems much thought but agree cap values must be considered over a brew or two.

[loudthud]

Steve was referring to the feedback that goes to the bottom of the tail resistor to make it look more like a constant current source. It's not really needed if you can tolerate a slight imbalance in your PI. Fender rarely paid much attention to it (IMHO) because they almost always used 100K and 82K plate resistors without regard to what tail resistor was being used.

[Steve Conner]

Sorry, don't know what I was thinking. You're right, the NFB can go directly to the grid and so there's actually one less capacitor in the loop than the standard Fender PI!

That as far as I know, is why the capacitor on the NFB side of a Fender PI is bigger than the one on the input side. It's to spread out the time constants and keep the NFB stable. Getting rid of it altogether is better still, I guess.

As redelephant hinted, you can use a transistor current source to make the tail look exactly like a current source. That fixes the imbalance, but maybe gets a bit avant-garde for your average guitarist