matched PI?

daniel · Post by **daniel** » Tue Jan 10, 2006 6:48 pm

Do you guys think the argument that you should use driver tube w/the two sides "matched." I really can't hear any difference.

novosibir · Post by **novosibir** » Tue Jan 10, 2006 7:56 pm

Some people - people who are selling tubes - claim, that a guitar amp with a matched driver tube would be more harmonic

But the opposite is the case! A push-pull-poweramp eliminates by its nature k2 (the second harmonic). And unfortunately the k2 is the harmonic, which makes the sound full, rich and more "meaty".

The best possibility to force a push-pull-powerstage producing some k2 is to feed it asymmetrically through the driver stage (AC assymetrical).

Another possibility would be the use of "unmatched" output tubes (DC assymetrical), but this will end up in enhanced hum and would drive the OT assymetrically AND earlier into saturation, what would result in a lack of bass response.

But here (the first possibility) as also in many other cases of the life: A tad is good, some more maybe is better, but too much simply is too much

Larry

Billy Batz · Post by **Billy Batz** » Tue Jan 10, 2006 9:57 pm

Larry, short of using a balance pot between the plate resistors on V3 how could you tweak the drive to be more assymetrical to be more what your talking about? Are there better sets of values to use for the plate resistors?

novosibir · Post by **novosibir** » Tue Jan 10, 2006 10:23 pm

The balance pot with the wiper on B+ and the both ends to the plate resistors is a HiFi heritage and was (is) there used to adjust best balance for a minimum of THD - what would be sounding cold, sterile and lifeless in a guitar amp.

The best solution would be at one plate the usual 100K plate load resistor (fix) and the other side a 100K lin pot, which is connected between the B+ and the plate. From the wiper then you go to the coupling cap.

In this case also the other plate always would "see" a constant load of 100K, but you can adjust the AC signal of one side individual with the voltage divider.

Without a pot I always end up with the usual 82K/100K, because this already IS asymmetrical.

In a Marshall type driver stage a "symmetrical" operation with a "matched" driver tube would occur with 82K/87.4K or with 94.1K/100K

Larry

Billy Batz · Post by **Billy Batz** » Wed Jan 11, 2006 12:02 am

So 82k already is more unbalanced. I guess dropping that to 75k would be 'too much'?

Star*Guitar · Post by **Star*Guitar** » Fri Jan 13, 2006 2:08 pm

I heard the reason for a balanced PI is to help eliminate fizzies...

rjgtr · Post by **rjgtr** » Fri Jan 13, 2006 3:24 pm

A few of the newer amps I've seen use the 92k/100k arrangement. If you're going to go one way or the other I'd say just use 2 100k resistors and measure them. In the end even a "balanced triode" tube will become unbalanced, just as a matched pair of tubes drifts apart from each other with use. I've used balance PIs in the past and don't bother anymore. In practice I don't see a lot of advantage in a guitar amp.

novosibir · Post by **novosibir** » Fri Jan 13, 2006 3:42 pm

Richard is right! What could a balanced PI tube help, when the PI's circuitry isn't? See also my post above!

And...

Star*Guitar wrote:I heard the reason for a balanced PI is to help eliminate fizzies...

A balanced PI doesn't eliminate anything! Don't believe in the rumors, maybe spreed around from people, who are selling tubes

Larry

Star*Guitar · Post by **Star*Guitar** » Fri Jan 13, 2006 3:45 pm

Okay...thanks...I appreciate it..I am learning all the time..not bad for a 52 year old geezer.. :~)

rjgtr · Post by **rjgtr** » Fri Jan 13, 2006 4:24 pm

My favorite tube seller line is that the PI should be changed every time the power tubes are changed. With that logic all tubes should be changed when power tubes are changed. All the preamp tubes, especially in a simple amp, carry signal and wear. Fortunately, 12ax7s can last a lot longer than power tubes. There's no point in changing them unless the amp's sound changes and subbing out the 12ax7 fixes the issue.