Phase splitter tubes /matching triodes?

[goraman]

How close do the phase splitter tubes need to be matched (triode to triode?)
My Manley Stingray uses 6414 tubes as phase splitters,how critical is matching them?

[goraman]

What happens if tubes used as phase plitters in a push pull are not closely matched?
Dose anyone know?

[Captain Valve]

Depends how much feedback is applied, if any at all. Also a push pull power stage will largely cancel any second harmonic.

[NRG]

If you think about is further you'll realize that the phase splitter is key in the way a PP amp resolves low level information and ultimately the way it engages you...accuracy in the Phase splitter is key ;)

[SteveTheShadow]

Aye....
The phase splitter is absolutely critical to the performance of a push-pull valve amplifier. It is a make or break aspect of their design.
IME the great classic phase splitters are the transformer, and the differential long tail pair, in that order.
Also IME, the very best phase splitter is no phase splitter; which is why my own push-pull amps are fully balanced and take a balanced input from my DAC straight onto the grids of the first stage.
But then I'm an awkward sod:lol:.

[Radford Revival]

It generally depends on the topology of the phase splitter. Some circuits will be more tolerant to imbalance than others. A concertina style phase splitter relies on heavy degenerative feedback and only uses one active device, but due to the way it works it will generally cope with quite a wide range of tolerance spread in valve performance. A long tailed pair's tolerance of imbalance depends on a few design details, for example if the cathodes are loaded with a constant current source then the AC balance will be more or less perfect (within the tolerance of the anode resistors) but there could, in extreme cases, be a large DC imbalance should one of the valves be very different to the other, which could lead to distortion. Generally if it gets to that point it means one of the triode halves is nearly dead while the other is OK, so that's an extreme scenario.

Sometimes long-tailed pair phase splitters have AC balance pots (not needed if they are CCS loaded) to null out any AC imbalance caused differing triodes

Generally a competent circuit will take the spread of parameters of valves into account (and luckily the circuits themselves tend to be quite forgiving of them anyway) and one doesn't need to worry too much, given decent quality and working valves.

Arguably perfectly matched phase splitting is only ideal when the output valves themselves are matched in gain anyway, so even given a 100% perfect phase splitter, the output valves will introduce their own imbalance.

[goraman]

I got 10 triple mica black plate 6414 computer tubes very carefully matched and replacing them made a wonderful difference.
The G.E. tubes that where in there only had around 3,000 hrs. and the 6414 is rated 10,000 but rumor has been the Raythions just sound so much better.
Well, they do!
Thank you for explaining the importance of this match, I am still learning a lot about valves and valve circuits.