Phase Inverter Preference & Why?

[The Ballzz]

Hey Folks,
Hoping to get some opinions and possibly start a discussion of the ins/out of the two main types of phase inverters, Cathodyne & Long Tailed Pair (LTP), we can ignore the Paraphase, for now. I've read many articles and watched many videos on the subject and they still leave me with the same thoughts/questions:

A) What is the actual/perceived tonal & response difference between the LTP and cathodyne, with all else being generally equal?

B) Many dissertations state the the LTP is superior because it provides some gain , while the cathodyne does not AND that the cathodyne requires a "driver" stage to make up for this. Either way, we're still using two triodes in the process!

C) What is actually "gained", improved and/or changed by doing it one way or the other?

Through study, I've noted that many of the "truly" revered vintage amps turn out to use a cathodyne inverter. Especially many E-series Tweed Deluxe, Super, Pro and Twins, as well as most, if not all, push/pull Princetons. These amps all seem to share a certain desired and unique character that I'm wondering if it is caused by the cathodyne inverter, or simply a coincidence that they all share that inverter style? I also notice that they all use either no NFB or NFB/presence variously in different ways.

Thanks In Advance,
Gene

[Leo_Gnardo]

There's also the rarely seen common-cathode "lazy" inverter, no actual inverter tube, you send signal directly to one output tube, and its cathode is wired to the other output tube in the push pull pair. Grid of the second output tube receives no signal, just has a leak resistor. to ground. Self/resistor biased of course and no bypass cap across the cathode resistor.

I've seen some written about it, saying that the tubes will be asymmetrically driven & maybe not sound so good, BUT I had a very old Gibson here a couple months ago with this setup and it sounded awesomely good for a little old amp from late 40's - early 50's. Pair of 6V6, low B+ maybe 350V, got only 7W or so, and on the scope the output looked just as nice as you'd want, no noticeable imbalance. I think it's a very overlooked scheme and would like to build an ankle biter amp or two or three, see how far I can push this idea up the power scale, also drive larger speakers than that little ol' Gibson.

Back to the comparison between the other methods, had to put my tuppence worth in.

[Ken Moon]

This doesn't have to do with perceived tonal differences, but a cathodyne will give you a much more limited voltage swing on the signals going into the power tubes, so you may have trouble overdriving bigger tubes like 6L6s, EL34s, KT77s, KT88s, etc.

This can be overcome somewhat by having a higher B+ (350V seems to be a good starting point) on the phase inverter node.

If heavily overdriving the power tubes is not a design goal, a cathodyne can work well with 6L6s, like in the tweed 5E5 Pro Amp.

[Ken Moon]

I forgot to mention that there are also issues with balancing cathodyne inverters - look at how the Tweedle Dee Deluxe uses a trimmer to balace the cathodyne.

[John_P_WI]

IMO, the imbalance of the cathodyne is what gives it the magic. As Ken said they can be somewhat anemic driving large tubes. Additionally, they can sound absolutely horrible when overdriven - a large grid stop can help this.

[Phil_S]

The more limited voltage swing of the cathodyne seems to work better for pushing triode power tubes.

Also, I think it's worth noting that the cathodyne is a design that is meant to send equal current to each half, whereas the LTP sends equal voltage. There can be balancing issues with both kinds, so "equal" isn't really that.

[brewdude]

I was under the impression that a 12au7 as a cathodyne was able to work well with the larger power tubes.

[The Ballzz]

Wow, Just WOW!

Amazing and fast responses! I had completely forgotten about the 'common-cathode "lazy" inverter' and that comes even closer to explaing an even deeper, underlying question that I will get into a bit later in this discussion!

The responses, so far, go a long ways into overviews of what my studying has already revealed, yet still don't really get into the "nitty-gritty" of the specific questions originally posed. I am looking for much deeper understanding!

Thanks To All & Please Keep The Thoughts Coming,
Gene

[Phil_S]

I was under the impression that a 12au7 as a cathodyne was able to work well with the larger power tubes.

Yes. It is rated for much more current than a 12AX7. Since you only need one triode, it opens up for choice a number of inexpensive single triode tubes.

[Garthhog]

I use a 12AU7 in a cathodyne circuit in my HayMakers, and I have 0 problems saturating an EL34. There are others ways to get a power tube to saturate without jamming giant signals into the grid.

[jjman]

A LTPI will have more drive potential than a cathodine at the same supply voltage and input voltage. But supply and input voltage are both “adjustable” by design as alluded to. So a “strong” drive to the output tubes should be achievable via either approach when considering all the “adjustable” variables.

Seems to me that many/most schematics from classic hifi amps, at least the ones I see posted online sometimes, have cathodine. So I no longer accept the internet guitar-amp wisdom that cathodine is automatically unbalanced, at least before clipping. During clipping, there are detailed analyses online and techniques to mitigate the expected bad stuff. Giant grid stoppers and various implementations of diode clampers are used there.

It would be nice to see and HERE a true comparison of the 2 approaches where both are driving the outputs with the same strength signals, with and without output clipping and with and without the cathodine clipping helpers.

[tubeswell]

In terms of the bang for buck you get out of your dual triode, an inverting driver stage and a cathodyne splitter together have more gain than an LTP.

An LTP is a good phase splitter for a guitar amp because it offers a decent amount of output signal swing to drive those bigger output tubes harder, but you probably want an additional triode in your pre-amp (in front of the LTP) to really get a strong signal swing (if you have a spare triode handy. An LTP is compromised between gain and balance. You can balance a LTP more evenly with a bigger tail resistance and an even plate load, but you progressively lose more and more gain the bigger you make the tail. You can add another driver tube after the LTP (but you need another tube for this).

A paraphase inverter has excellent gain but is impossible to balance precisely, although there are various ways or getting good approximation of balance. But on the whole it can be quite good for a guitar amp where increasing distortion is called for with increasing gain. There's lots of good even-order harmonics from a paraphrase inverter.

The cathodyne's peak output signal is limited to 1/4 to 1/3 (depending on how you set up the stage) of the cathodyne's HT voltage. This limits how hard you can drive an output tube. Furthermore, cathode-biasing the cathodyne can get a really high input impedance, but this puts the stage out of balance, which increases the output impedance. This is all problematic for driving output tubes, which have low input impedance, and a big bias voltage. One 'solution' is to put a buffer/driver stage between the splitter and the output tubes. (But then you are using more tubes). You can also use fixed-biasing on a cathodyne, but this doesn't permit as-high-an-input-impedance as bootstrapped cathode biasing.

[Phil_S]

While we're at it, can anyone offer commentary on using an interstage transformer instead of a tube splitter? In my limited research, these seem to have a turns ratio of something between 1:2 and 1:3. They are generally small transformers and not all that costly (though I suppose some of the hi-fi ones might get up there in price.)

[Stevem]

Well speaking from my work with 60s SS amps that make use of a transformer for a PI I can tell you that tone wise there high end roll off is a welcome thing to tame the bad nature of a SS amp clipping the preamp section.
Amazingly Sunn amps used this PI format into 1971 when other top name SS amp makers ditched it in 66!

[jon]

While we're at it, can anyone offer commentary on using an interstage transformer instead of a tube splitter? In my limited research, these seem to have a turns ratio of something between 1:2 and 1:3. They are generally small transformers and not all that costly (though I suppose some of the hi-fi ones might get up there in price.)

Please elaborate I am very curious about inter-stage transformers.