Crucial to match tubes when running PSE?

[G]

Hi All,

If I wanted to build a amplifier with a PSE output stage would I have to use "matched" tubes? What are the consequences if I don't? Thanks.

G

 

[fdegrove]

Hi,

If I wanted to build a amplifier with a PSE output stage would I have to use "matched" tubes?

What's the bias arrangement you're using?

Cheers,😉

 

[G]

Hi Frank,

I would be using one pair of tubes per side with one cathode resistor for each pair of tubes.

G

 

[fdegrove]

Hi Gavin,

I would be using one pair of tubes per side with one cathode resistor for each pair of tubes.

As long as the output tubes aren't grossly mismatched the common cathode bias resistor will tend to equalise the current between the pair.

If possible I'd still try not to mismatch the output tubes for some current hogging will still occur making the tubes ages unequally.

Cheers,😉

 

[dhaen]

Sorry, I have to disagree

Frank,

I've been looking at your last post for a while, wondering what's going on.

With parallelled valves, the greedier valve will starve the lesser one of current, because it sets the cathode voltage, and hence the grid bias conditions.
I think there are less problems in push-pull, but this is PSE.

Cheers,

 

[fdegrove]

Re:Sorry, I have to disagree

Hi John,

With parallelled valves, the greedier valve will starve the lesser one of current, because it sets the cathode voltage, and hence the grid bias conditions.

Thing is, I DO agree with you which is why I advised against mismatched tubes.

Maybe I should have stated that more clearly...😉

 

[dhaen]

We agree

Hi,

Well Frank, what do you think of the suggestion that Gavin uses individual cathode resistors?
I believe it should be sympathetic to a greater valve mis-match than with a single resistor.

Cheers,

 

[fdegrove]

Hi,

I believe it should be sympathetic to a greater valve mis-match than with a single resistor

And make a part of the individual resistors adjustable?

Cheers,😉

 

[dhaen]

Hi,



And make a part of the individual resistors adjustable?

Cheers,😉

Now you're taking the
Still, it is the best in the world (and not probably)

Cheers,

For G:
It is possible to get a match between 2 not-too-dissimilar devices by matching their curves at 2 points. To do this, you have to tweak the bias point for one end, and the drive for the other. It's a lot of ********, and you're best to find valves that are a reasonable match.
If you have a TV with a CRT, this is how the 3 electron guns are matched.

 

[Colt45]

yeah, a seperate Rk for each lamp should have better current sharing. with a common Rk, the greedier tube can be greedier.

 

[Bas Horneman]

I've tried this in the past...works pretty well..

http://www.svetlana.com/graphics/TB/No.26fig1.jpg

12Watts by Ohmite..i used

 

[dhaen]

A fine application note....

Bas,

That's a good idea, but I think Frank patented it earlier 🙂

Cheers

 

[Bas Horneman]

Was there a date on my schematic?🙄

You mean Frank beat Cyrus to it?😱

 

[G]

This is a rushed idea (busy day) but I think that this would work also. Or is this just a variation on Frank's idea?

 

[Bas Horneman]

That's a good idea, but I think Frank patented it earlier

Hi John,

I was not trying to catch the limelight here. Merely trying to make it graphic. I did not know there was a competition running on who gives the most brilliant anwers fastest?

Cheers,
Bas

 

[NickC]

how about changing the cathode resistor to a ccs too. Work better for me but do leave the 12 ohm ohmite in to match current in 2 tubes

 

[Bas Horneman]

Well the one thing about the 10 ohm resistors in Cyrus's schematics is that you are able to measure accurately the current being pulled by one tube or the other...

So I reckon they are needed? Maybe the geniuses around here can give the definitive answer..

 

[fdegrove]

Hi,

Gavin, using individual cathode resistors and rheostats gives you even more control over unbalance.

The idea was patented by Sir D'haen....😉

Cheers,😉

 

[fdegrove]

Hi,

Well the one thing about the 10 ohm resistors in Cyrus's schematics is that you are able to measure accurately the current being pulled by one tube or the other...

Not really needed but useful nonetheless.
If you look at the drawing you notice the testpoints close to # 2 and #5.

1 Ohm resistors would have worked just the same, BTW.

Cheers,😉

 

[Bas Horneman]

1 ohm resistors work just as well..

It is just a little harder to find matching 1 ohm resistors? or not?