Phase inverter distortion comparison

[TooManyProjects]

I've built a number of amps using both cathodyne (always with the Paul C mod) and long-tailed pair phase inverters, but I've never built two similar amps so that I could compare the sound of the two circuits. Does anyone have any experience with this? Do they sound different when they distort? Since the first triode of the cathodyne is just a standard gain stage driving the "actual" phase inverter triode, and the second triode has no gain of its own, I suppose it's reasonable to assume that cathodyne distortion sounds like any other triode (more or less). A long-tailed pair has all of that negative feedback so I would assume that it is more compressed sounding when it distorts. Thoughts?

[Phil_S]

I don't think you can get there from here. These are two very different topologies.
The LTP is a voltage driven circuit. The voltage is more or less symetrical on each triode.
The cathodyne is a current driven circuit. The current on each half (plate and cathode) are about the same. I'm sure you must have noticed that the plate voltage and the cathode voltage on the cathodyne inverter are significantly different.

[sluckey]

The cathodyne is a current driven circuit. The current on each half (plate and cathode) are about the same. I'm sure you must have noticed that the plate voltage and the cathode voltage on the cathodyne inverter are significantly different.

There's only one current path through the tube so the cathode current and plate current are one and the same. Exactly equal.

Sure there's a big difference in the plate and cathode voltages. That's because the plate voltage is the SUM of the voltage drop across the tube PLUS the voltage drop across the cathode resistor(s).

[TooManyProjects]

I understand that. I understated their differences. But none of that answers this question: How do they sound different when they distort? Does one give us creamy peanut butter and the other crunchy? Does one produce greater second order harmonics than the other? Will my tone paint sound distinctly more purple if I use a cathodyne?

[pdf64]

I think that for most amps, the power tube control grids will clip at a somewhat lower signal levels than that at which the phase would. So that’s what the listener will hear.
Same as putting clipping diodes after a DCCF; any asymmetric waveform rounding introduced by the dccf will most likely get chopped off by the diodes it feeds.

[katopan]

Yes they sound different when they distort. As already mentioned in most amps the power stage will clip first, but as you drive harder, unless the power stage is so overdriven by then it doesn't matter anymore (eg. Marshall 18 Watt) then you'll hear the change in distorted tone when the PI starts to clip. Problem is when you think of each type, classic amps come to mind that have other differences as well. So it's hard to separate out the tonal difference contributed by the PI type vs. other differences.

As for the cathodyne, there's an interesting character to the 5E3 Tweed Deluxe's overdriven tone. Some people love it and it's certainly a character of that amp. Some people hate it and implement "fixes" to make the PI overdrive in a way that is better behaved (ie. doesn't have the peak inversion on one side with extreme clipping). These fixes usually include a high value resistor on the cathodyne stage input grid to stop big positive swings on the input dragging up the plate output signal back in a positive direction. So a 5E3 with or without these mods sounds different. A Vox AC sounds different to a Marshall 18 Watt, and although there are a number of factors to that difference, one of them is the different LTP PI and how in the Vox the PI contribution to the amp's overdriven tone is audible, where in the 18 Watt it basically isn't.

[TooManyProjects]

Okay. Thanks. Sounds like I should just leave it the way it is then.

[katopan]

Okay. Thanks. Sounds like I should just leave it the way it is then.

I think you should experiment, test it out and let us know what you find.

[teemuk]

As rough generalisation: Differential (PI) stage typically clips rather symmetrically and "soft" (low order odd harmonic distortion) while a common cathode stage (this would be the typical "driver" voltage amplifier stage for cathodyne PI) clips rather asymmetrically and "hard" (higher order odd and even order distortion). If differential is driven very hard its output will drift towards more overall asymmetry.
Yes, as noted the cathodyne PI may as well overdrive (in addition to its driver stage) and then you get weird effects like phase inversion of one of the peaks. Generally distortion generated by the CC+Cathodyne PI is "harsher" than that of LTP.
These are, of course, mere generalisations and the overall circuit architectures - in each case-specific design - may be tweaked to provide preferred characteristics. As usual.

Also... Essentially BOTH phase inverter schemes can feature global negative feedback. Only difference is that the feedback is inserted to driver (voltage amp) stage in cathodyne scheme. If negative feedback is featured it - as usual - will extend range of linear amplification with cost of clipping becoming more abrupt, and will cause the associated voltage amplifier stage (whether LTP or common cathode driver) to introduce distortion that counteracts overall distortion of the output stage. These are typical traits to all amplifiers with feedback regardless of PI circuit type.

As said, a great chance is that power tube stage distortion outweighs contributions of PI distortion. Phase inverters typically have high headroom but (very likely) there will be interactions at the moment of powr stage overdrive due to grid clipping (possible PI clamping) and global negative feedback causing "error correcting" distortion in the voltage amp driver or voltage amp PI.

[jbrew73]

As a general statement I think the p.i.gives an amp a certain character as much or more so than the output tubes and possibly the preamp. i experimented with a vox, Marshall and orange pi fed by various preamps into various output tubes configurations and was surprised to find the pi appeared to give the amp it overall character.

[printer2]

So what happens if you change the values in a PI? Surely it changes the nature of that design. I am guessing that you can not just compare one PI with the other without seeing what changes to the values do to that PI. Well you can, but then you are only getting half the story.

[roberto]

Even the values wouldn't tell the whole story, because then you need to know the voltage of the power supply, then the impedance of the output, the way it is driven, the power tubes it is connected to, if they are directly connected or through a driver stage, how the feedback is connected, blablabla...

But I would say teemuk's resume is accurate and fits my experience too.

[Mr. Lime]

How about the Paraphase PI?
We can theoretically take a cold clipper or a center biased gain stage as driver. Also feedback can be applied to the cathode.

Is there any disadvantage compared to the cathodyne I miss?
I guess the output swing is the widest of those 3.
Did Fender moved away from the paraphase because of economy (less parts )?
I like the LTPI for channel mixing alà Tweed Mixer for amps without NFB and would guess that the cathodyne is good for amps with big bottles that generate the distortion early in the preamp and need a clean PA.


Did anyone compared the overdrive of Paraphase with LTPI?