• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Class AB2, Class B, and Hifi

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
It has been stated before in this forum that 45's operating in class AB2 putting out around 20 watts sounds quite nice. One wonders if some of the traditionally class-B triodes such as the 809 could also sound good if properly driven (mosfet source followers or whatever).
Considering that the data sheet for the 45 at AB2 shows no signal current at 28 mils and full power at 138 mils, one might say that the 45's are running at 20% class A.
Now the sheet for the 809 shows the currents at 70 and 250 mils when running with zero bias. So at 28% class A, this compares favorably with the 45.
Could it be true that the infamous class B may not always be as bad or as close to cutoff as reputed? Or is there something else being overlooked. Perhaps the mu is too high. It is apparent that audiophool tubes are generally low-mu, whatever that does for the sound.
 
If one gets the two phases to combine well (with out any trace of cross-over dist) then class-B is perfectly fine. I would think a sharp cut-off type of tube is better than a soft/remote cut-off in that duty.
Btw, wouldn't 28mA of idle current qualify for the term class-AB?
 
Btw, wouldn't 28mA of idle current qualify for the term class-AB?

Yes, tecnically if there is some idle current, then there is some overlap area where both tubes are conducting. This IS the definition for AB operation. Class B requires exactly a 180 degree conduction angle which is impossible to achieve in reality.

Or is there something else being overlooked. Perhaps the mu is too high. It is apparent that audiophool tubes are generally low-mu, whatever that does for the sound.

A lower mu tube will have a lower plate resistance which can make operation without feedback possible in triode mode. High mu triodes and pentodes generally require some form of feedback to make the amplifier's output impedance compatible with most modern speakers.

One wonders if some of the traditionally class-B triodes such as the 809 could also sound good if properly driven

A pair of 809's in push pull AB2 does indeed sound quite good. I ran them with some Schade style feedback even though the "experts" here have repeatedly explained to me how Schade does not work on triodes.
 
High mu triodes and pentodes generally require some form of feedback to make the amplifier's output impedance compatible with most modern speakers.

Well that explains why audiophiles don't like Class B triodes. I guess they don't really have any advantages over beam tetrodes anyway. Traditionally, zero bias and no screen supply were the advantages, but for hifi, negative bias would still be needed for the drivers and the directly heated filaments would also probably need DC supplies.

Thanks for the comments.
 
tubelab.com said:
I ran them with some Schade style feedback even though the "experts" here have repeatedly explained to me how Schade does not work on triodes.
So-called "Schade feedback" is just the old anode follower circuit, so will work with any active device - pentodes, triodes, even (inverting) op-amps. The problem is that some people think something magic happens when you put feedback around a pentode: somehow the pentode is fooled into thinking that it is a triode. People who think like this would naturally assume that you can't 'Schade' a triode as it is already a triode.
 
I have understood that the reason why AB2 class operation is not "perfect" is that the anode current is cut off part of the drive cycle and this causes known negative effects in the output transformer.
In case of pentode and triode this is obvious.
But in case of UL-connected output stage there is still screen current flowing thru the output transformer when the anode current is already cut off.
Or at least the cut off is smoother and takes place with higher drive.

Opinions about this ?
 
AB2 class operation is not "perfect" is that the anode current is cut off part of the drive cycle and this causes known negative effects in the output transformer.

One tube gets completely cut off in AB1 or AB2. The only difference between AB1 and AB2 is the presence of grid current in A2 or AB2. This used to cause issues in the driver but mosfet drive totally eliminates this concern.

The "known negative effects" are also overblown by some of the audiophile "class A is the way" types. Some even claim to hear the abrupt turn off of one output tube. The truth is that the turn off happens gradually with most music, and the Gm doubling effect also happens gradually. Gm in a tube is not constant, it varies with tube current, so while one tube is losing Gm as it approaches cut off, the other gains it as its' current increases.

But in case of UL-connected output stage there is still screen current flowing thru the output transformer when the anode current is already cut off.

The screen current tends to drop as the tube approaches cut off, and can go to zero in some cases.

Every case is different and a well designed class AB1 or AB2 amplifier can sound very good. True a pentode will have a higher output impedance than a triode and some type of feedback is generally needed to lower the output impedance. I didn't say GNFB, since I tend to avoid it in my designs, but it is possible to build a good sounding amp with GNFB. My Simple P-P is one example.
 
but it is possible to build a good sounding amp with GNFB. My Simple P-P is one example.

yes......in my case, operating in pentode mode, the screens should be well filtered, regulated even or fed from a low impedance source.....this has been admonished in several books about the subject and obeying rules this gives rewardingly good results......
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.