What Phase Inverter Is This?

[darkbluemurder]

Guys,

I recently stumbled upon a phase inverter in a schematic which I have not seen before and could not find in any of my books. For easier disussion I attach a file which shows the regular long tail PI (Alternative 1) and the one I need help understanding (Alternative 2).

I understand that it is a differential amplifier since it does not resemble a concertina or a paraphase. The second grid is directly grounded (0VAC and 0VDC). The cathode should be at about 2-3 VDC with that 1.5k resistor, compared to approx. 70VDC in the regular LTPI. Am I correct in thinking that this means 70V more through the tube, hence more headroom and thus more available signal swing to drive the power tubes?

Any input would be greatly appreciated.

Thanks and best regards,
Stephan

[PRR]

Variante 1 is of course a Long Tail.

Variante 2 is a "short tail". For good symmetry the tail impedance has to be much larger than a cathode impedance, or typically >>1K. Because this tail is only 1.5K, it does not want to be symmetric. This designer "cheated" by increasing the back-side plate resistor from ~~50K to ~~300K.

> more headroom and thus more available signal swing

I doubt it. The severe asymmetry means one side or the other is pretty far from "optimum".

I think it is just cheap.

There is a chance that a cheap commercial design was discovered to be "good" in the sense of "interesting overload", and re-used in some newer guitar amp.

[PRR]

Hmmmm... it isn't awful:

[darkbluemurder]

Thanks a lot PRR.

Now what does the graph tell us - it is not too imbalanced, correct? What about the available signal swing?

I know it does not sound too bad because I once had a chance to play the amp that has this set up and at that time it was the best amp in the store by far and very loud, too.

Once again, thanks a lot.

Stephan

[darkbluemurder]

Hmmmm... it isn't awful:

In that particular amp R7 and R8 would be 220k instead of 470k but I guess it will not make a difference to the overall result.

Thanks again,
Stephan

[Merlin]

Just found someone else who has messed with this phase inverter
http://dogstar.dantimax.dk/tubestuf/paraph.htm

(He calls it a cathode-coupled paraphase, which it isn't really, it's a crude LTP.)

[Shrapnel]

Just found someone else who has messed with this phase inverter
http://dogstar.dantimax.dk/tubestuf/paraph.htm

(He calls it a cathode-coupled paraphase, which it isn't really, it's a crude LTP.)

He seems to call ALL the PI configurations paraphase regardless.  Some interesting ideas though for some odd-ball tube usage in a PI.

[darkbluemurder]

Some interesting ideas though for some odd-ball tube usage in a PI.

Indeed. Didn't even know a 6X8 existed. Got to check the Receiving Tube Manual.

Thanks for bringing this up, Merlin.

Cheers Stephan

[PRR]

> R7 and R8 would be 220k instead of 470k

Well, since 220K is less than the 330K plate resistor on one side, we have to wonder.

It looks like it will make a good 50V peak with 5% THD on the differential output.

And the THD is mostly 3rd, not 2nd as we might suspect from the wild asymmetry.

[darkbluemurder]

I meanwhile rebuilt one of my amps and used this PI. Since I have not used the same preamp with another PI I do not have a direct comparison on how the PI sounds, unfortunately. However I like the amp's tone.

My B+ on the PI is 305VDC leading to PI plate voltages of 240VDC respectively 170VDC. I suppose with these voltages there will be even more asymmetric clipping in the PI.

And PRR, your assumption was correct that this design was used in a newer guitar amp (from the 90s).

Cheers Stephan