Recent Projects

[Les Reed]

Not been around here for a while, but have still been keeping the projects going.

I thought my hairy chested 1000V 845SE with 10Y drivers might be my ultimate and difficult to improve upon beast.

But it's difficult to keep the hobby under control when you have a workshop full of test gear and parts.

A quote I kept reading on another forum - It's difficult to build a bad 2A3 amp - or something like that. So I spent a while designing one and after 2 single ended amps in a row, it had to be push pull.

Now I won't go into the boring detail of the circuit at the moment, but the main feature is the use of a Lundahl interstage to do the phase splitting. Not using a valve for this job, eliminates crossover distortion.

It delivers about 11W into 4ohms, no global negative feedback, and at normal listening levels only registers 0.1% THD around 5W. It's also virtually hum free with AC heating throughout.

This is easily my best sounding amp (including one I built after it). Once again I am a PP fan. lol

So what's left to build? Perhaps a flea power SE using vintage DHT's - but I find it difficult to conquer the hum in the input stage without pricey components added. So that might be it for a while.

[Mike H]

"Crossover distortion" ?

[Les Reed]

Yes, I understood that is the term to describe any distortion introduced as a result of imperfect phase splitting/recombination of the signal. You can see it on a scope display from a fixed bias guitar amp if you put the pots out of adjustment. It might be called something else though. lol

[Nick]

I am not sure just what the official definition is, but normally it seems to be used to describe the non linear part of the transfer curve as the bulk of the signal passes from one half of the output stage to the other, and if thats the case, then the phase splitter isn;t involved.

But as I say, I don't know the real def, but having heard Andrew I's amp I agree that Lundahl interstages do a good job at splitting phase.

[pre65]

What about the way you told us about using the common cathode or cathode coupled method ?

That seems to work without using a valve or transformer.

[Mike H]

Edit
Have I misinterpreted again? Anyway to continue with 'cathodyne' (which I don't think you meant)

Resume
It certainly does, though there's just a couple of things to bear in mind:

It's impedance sensitive, so the input impedances of the follwing stages must be exactly the same for both top and bottom halves.

Also must be exactly the same if you want both top and bottom waveforms to be of the same amplitude. Which you normally do.

Thirdly, if a triode is used you get a Miller effect twixt grid and anode, but not twixt grid and cathode. This may cause asymmetrical outputs at high frequencies.

Using a pentode wired as a triode instead negates this however, due to screen and suppressor grids combatting the Miller effect. From which I now better understand an expression I've read quote "XXX pentode wired as a triode had better linearity than any triode type that was tried."

Pick the bones out of that


Re crossover distortion, only come across that before really re push-pull transistor output stages, which must have a quiescent current for the region of operation between where one side or the other cuts off (stops conducting). Without the quiescent current you get like a flat step in the waveform at the zero crossing point. Called, crossover distortion.

If anyone can describe it better please have a go

[Will]

Switching distortion ?? I think.

[Nick]

It's impedance sensitive, so the input impedances of the follwing stages must be exactly the same for both top and bottom halves.

Also must be exactly the same if you want both top and bottom waveforms to be of the same amplitude. Which you normally do. Very Happy

Thirdly, if a triode is used you get a Miller effect twixt grid and anode, but not twixt grid and cathode. This may cause asymmetrical outputs at high frequencies.

Thats a big can of worms you are getting near, not going there.

[Mike H]

One more thing to try us.

Switching distortion ?? I think.

In all the books I read they called it "crossover distortion"