SE distortion

Shoog · 2013-12-24 12:59 pm

H E Pennypacker said:
It does sound like a fun amplifier though.

What is a good example of a popular class A PP that people are building?

Lynn Olsen has a few good designs.
Gary Pimm has his Tabor - but its a rather special case since it uses lots of local feedback and all pentodes.
Allan Wrights PP amps are other good examples - but I don't think they are pure class A - just almost.

Shoog

H E Pennypacker · 2013-12-24 1:12 pm

Thanks I will look into those for a future project.

mr2racer · 2013-12-24 1:12 pm

Shoog,
For class A PP do you have to use a special output transformer? Or is it just manipulation of the anode current? And are PP class A amplifiers run in ultralinear as well or just in pentode?

Shoog · 2013-12-24 1:28 pm

mr2racer said:
Shoog,
For class A PP do you have to use a special output transformer? Or is it just manipulation of the anode current? And are PP class A amplifiers run in ultralinear as well or just in pentode?

You can build a Class A PP in what ever flavour you like, Triode, Pentode and UL. The thing is, if you don't want to use gNFB then you are restricted to triode, UL or using Schade type feedback for pentode.

The requirements for the OT are the same as any PP amp. I prefer to run mine at Constant Current so that I can use a zero gapped toroidal OT which minimizes hysteresis in the OT and helps to preserve micro-detail. Its largely up to you what compromises you are prepared to accept.

Shoog

mr2racer · 2013-12-24 1:33 pm

This may be an oversimplification. But if you simply up the quiescent current in a standard PP amp do you force it to stay in class A?

Shoog · 2013-12-24 1:45 pm

mr2racer said:
This may be an oversimplification. But if you simply up the quiescent current in a standard PP amp do you force it to stay in class A?

Any PP amp could stray into AB if there is enough drive voltage to send it there. This is equally true of SE.
You have to work out your load line in such a way that neither valve ever reaches zero g1 voltage point. Effectively that requires upping the quietscent current as a consequence.

Shoog

Shoog · 2013-12-24 2:37 pm

For a fair analysis of the benefits and pitfalls of Class A PP try this;

Class-A Amplifiers explained

Shoog

mr2racer · 2013-12-24 8:00 pm

So the essence of class A is that it is a constant current output stage?

Shoog · 2013-12-24 9:16 pm

The essence of Class A is that it never crosses into positive g1 territory. This effectively means that the bias current is roughly half way between zero current and 2x bias current. This is roughly the same as saying that it has constant average current.

Shoog

20to20 · 2013-12-24 10:30 pm

mr2racer said:
So the essence of class A is that it is a constant current output stage?

Class A is simply that the drive signal never drives the tube outside of the voltage "envelope", between 0v on the grid and the (-) cutoff voltage of the tube operating point. The positive swing cannot be higher than the (-) bias voltage and the negative swing never takes the tube farther than just short of the cutoff level. The tube always conducts. If the tube is biased too (-) then it will go into the cutoff range too quickly leaving the other tube to do the work during that period. Or doing nothing if it's SE. This normally occurs at high volume drive levels.

Miles Prower · 2013-12-25 4:54 am

H E Pennypacker said:
What is a good example of a popular class A PP that people are building?

The Oddwatt. This design uses a self-splitting final in the form of an LTP with active tail loading. The active tail load enforces Class A operation. Of course, this does cost you some output.

Miles Prower · 2013-12-25 5:02 am

Shoog said:
The essence of Class A is that it never crosses into positive g1 territory. This effectively means that the bias current is roughly half way between zero current and 2x bias current. This is roughly the same as saying that it has constant average current.

Shoog

That's not right. The essence of Class A is that the final(s) never cross into plate current cutoff. You can most surely operate something like the 811 in Class A by using a positive V_GK to set a Q-point plate current, then drive it so that the plate current never goes into cutoff.

Of course, you'll need a stiff grid driver that can supply the required grid current. Something like a source follower.

The other thing to keep in mind is that the OPT will need to match the speaker load to 2X (P-2-P) R_L / phase, and not the 4X P-2-P you'd need for Class AB (or B or C) finals.

Shoog · 2013-12-25 8:27 am

Miles Prower said:
That's not right. The essence of Class A is that the final(s) never cross into plate current cutoff. You can most surely operate something like the 811 in Class A by using a positive V_GK to set a Q-point plate current, then drive it so that the plate current never goes into cutoff.

Of course, you'll need a stiff grid driver that can supply the required grid current. Something like a source follower.

The other thing to keep in mind is that the OPT will need to match the speaker load to 2X (P-2-P) R_L / phase, and not the 4X P-2-P you'd need for Class AB (or B or C) finals.

Thanks for clarifying.

Shoog

dady · 2013-12-25 11:58 am

A2 Class.
Please Re-Clarify the concept. You get the proper words for do it.
Happy Christmas again.

Evenharmonics · 2013-12-25 4:30 pm

Shoog said:
These are absolutely none issues in a PP amp designed to run in Class A. A PP amp can be built as two back to back SE amps and will deliver all of the benefits of SE along with the benefits of PP. In short it can give more than an SE amp.

Are there any FFT measurements of those?

Shoog · 2013-12-26 10:04 am

This is the distortion measurement of Allan Wrights PP-1C

http://www.vacuumstate.com/images_upload/gross/pp_1c_thd_s.gif

I think you will find very little difference between the distortion profile of this PP amp and a typical SE amp (apart from been generally lower with the PP).

The Amity follows exactly the predicted distortion profile anticipated from a good class A PP amp - with low overall distortion and a reduced 2nd harmonic component due to cancellation;

http://usr.audioasylum.com/images/2/24819/distortion_1.jpg

http://www.audioasylum.com/forums/tubediy/messages/13/131515.html

Shoog

zigzagflux · 2013-12-26 1:00 pm

Or the 3 stage Karna, at 1W, 4W, and noise floor. Pretty impressive IMO. As noted, looks fairly similar to a SE, but with lower distortion.

mr2racer · 2013-12-26 3:07 pm

Am I correct in assuming that this is at a 1k signal? And the second peak at 2k is 2nd order?...

Evenharmonics · 2013-12-26 4:30 pm

Shoog said:
This is the distortion measurement of Allan Wrights PP-1C

http://www.vacuumstate.com/images_upload/gross/pp_1c_thd_s.gif

At what output level?

The Amity follows exactly the predicted distortion profile anticipated from a good class A PP amp - with low overall distortion and a reduced 2nd harmonic component due to cancellation;

http://usr.audioasylum.com/images/2/24819/distortion_1.jpg

Some measurements on the Amity amplifier - arend-jan - Tube DIY Asylum

Those look like typical class AB amp measurement.

Shoog · 2013-12-26 7:27 pm

Evenharmonics said:
At what output level?

Those look like typical class AB amp measurement.

Zero global feedback amps rarely look anything like that. These are impressive figures for a zero feedback PP amp - equaling and bettering anything a zero feedback SE can achieve.

Shoog

SE distortion

Member

Banned

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Banned

Member

Member

Attachments

Member

Banned

Member