Influence of cathodyne vs LTP PI in a Vox

[roseblood11]

Can anybody compare the influence of a cathodyne phase inverter versus the long tailed pair in a Vox circuit? (with everything else identical...) For example the Pacemaker V2 and the Cambridge Reverb... I'm planning a small practise amp with top boost channel and a ECC99 push/pull power amp, and I don't know which PI to use. Cathodyne would be simple, as I'd only need two 12ax7. LTP would mean three 12ax7, but with the option to add the simple tremolo circuit from the Pacemaker later... But I guess that the PI plays an imporant role in the Vox overdrive sound? The cathodyne would get a 1M grid resistor (see: Merlin...) off course.

regards, Immo

[eschertron]

Since you said "small" practice amp, I'm guessing the ECC99 tube doesn't need 70vpp or more to drive it to clipping. Match the PI design to the power tubes' requirement for drive. I'd recommend the cathodyne, and follow Merlin's suggestions for a high grid stopper value on the cathodyne. Is that the grid resistor you mentioned?

[PeanutNore]

My experience with ECC99s is that you'll need about 24v p2p to fully drive them. If you can get that out of your preamp prior to the PI, you could certainly use a cathodyne with no driver stage. If you don't get that out of the preamp, you won't be able to get much power tube OD out of your amp unless you use a phase inverter with some gain, like an LTP or a cathodyne with a driver stage in front of it. In this case I'd recommend the LTP if you want it to still sound like a Vox. Then you can always find something to do with the extra triode, like a switchable boost in front, or throw in a cathode follower somewhere.

[Chuck H]

It's a tricky question because, as eschertron noted, the drive requirements of the power tubes will likely be different than any VOX amp. My vote is actually for the LTP. Only because I interpret that your building a lower wattage amp so that it can be pushed into clipping on occasion and the LTP has better clipping character (to my sensibilities) than the cathodyne. And PI clipping is a big part of the final overdrive sound. The LTP also provides some gain so you're not usually forced to design the stage preceding the PI to provide the required power tube drive voltage. In this sense the cathodyne IS actually a two triode PI IMHE.

[PeanutNore]

If you want to hear an example of an overdriven low wattage dual triode output with a cathodyne phase splitter, here's a demo of the 12AU7 based Orange Pics-Only clone that I recently finished. It can also be run with an ECC99, and they both bias around the same voltage so the drive requirements are the same, the 12AU7 just doesn't get quite as loud or deep.

Soundcloud link: https://soundcloud.com/sam-brown-225...2-sm57-dean-vx

[Chuck H]

My experience with ECC99s is that you'll need about 24v p2p to fully drive them. If you can get that out of your preamp prior to the PI, you could certainly use a cathodyne with no driver stage. If you don't get that out of the preamp, you won't be able to get much power tube OD out of your amp unless you use a phase inverter with some gain, like an LTP or a cathodyne with a driver stage in front of it. In this case I'd recommend the LTP if you want it to still sound like a Vox. Then you can always find something to do with the extra triode, like a switchable boost in front, or throw in a cathode follower somewhere.

A "proper" top boost circuit has a cathode follower anyway. I think. I'm not expert on all the VOX circuits, but that's how I did my TB amp design because that's what the schem I saw showed.

[eschertron]

The LTP also provides some gain so you're not usually forced to design the stage preceding the PI to provide the required power tube drive voltage. In this sense the cathodyne IS actually a two triode PI IMHE.

I absolutely agree with this. I went back and forth between cathodyne and LTP on a 2xEL84 build I did, and ended up with the cathodyne preceded by a recovery stage. The recovery stage had some local NFB simply to get the gain where I wanted for that particular project (preamp distortion, mostly). The LTP was too much for that amp. Your design is, of course, your design. Based on the numbers provided by PeanutNore above, and a thumbrule of 25x gain out of each side of a LTP, the LTP will get ECC99s to clip with about 1vpp feeding the PI. Maybe one engineering reference point to design around.