AudioRoundTable.com: Thermionic Emissions

Home » Audio » Thermionic Emissions » Why SE in SET amps?

Show: Today's Messages :: Navigator :: Polls
E-mail to friend

Why SE in SET amps? [message #11382]

Mon, 03 September 2007 23:41

granch
Messages: 118
Registered: May 2009

Viscount

Could someone succinctly explain to an old duffer like me why it has to be "SE" in Triode amps. In my youth that was asking for even order harmonic distortion which a push-pull would cancel. My first big project involved replacing a push-pull parallel 2A3's amp with another to get more power in a high school auditorium. I must have been committing a terrible sacriledge (unintentionally, of course). I was so innocent in those days I even tried to explain negative feedback by going around the loop and got myself lost. Granted even order distortion is less objectional to the ear than odd order. But wouldn't less be even better? Please no cursing, folks.
-Dick

Report message to a moderator

Re: Why SE in SET amps? [message #11383 is a reply to message #11382]

Tue, 04 September 2007 08:46

Wayne Parham
Messages: 18789
Registered: January 2001

Illuminati (33rd Degree)

It's probably not something you'll be able to discuss in terms of math. If you look at the numbers, you're right. Push-pull offers lower distortion. Well-executed solid state even lower distortion. But to a SET purist, push-pull is "flat and uninvolving" and there is no such thing as well executed solid state. They're lured by the warmth of SET magic.

A Taste of Tubes

There are two things that may explain some of the reasons why people prefer SET. For one thing, Class A amplifiers have no crossover distortion. The entire audio signal is handled by one amplifying element. And for another thing, the distortion that is made is largely low-harmonics, mostly second and third. This is mostly due to the fact that the tube conductivity isn't perfectly linear, and doesn't have anything to do with more complex things like feedback nonlinearity. A tube clips gracefully too, not with sharp harsh-sounding square edges but almost sounding like it just won't go any further. So without resorting to romantic adjectives, you can see some things that may be part of the SET magic.

Report message to a moderator

Re: Why SE in SET amps? [message #11384 is a reply to message #11383]

Tue, 04 September 2007 14:58

granch
Messages: 118
Registered: May 2009

Viscount

Thanks a million, Wayne. Just what I thought. The reference was really terrific! A great read and a great refresher. As I recalled, the best sounding amp ought to be a push-pull triode amp operating in class A or AB.
I'm still using my old ultra-linear Wiliamson amp -in my organ I think. I don't think it was a kit and it did use a peerless (altec) output transformer as I recall. For serious power I use solid state but very conservatively rated.
-Dick

Report message to a moderator

Re: Why SE in SET amps? [message #11385 is a reply to message #11384]

Tue, 04 September 2007 17:50

Wayne Parham
Messages: 18789
Registered: January 2001

Illuminati (33rd Degree)

For SET amps, I really like the Audio Note Kit 2. The Kit 1 is nice too, but I'm partial to my Kit 2. Good value at under $2K. For Ultra-Linear, check out the Stoetkit Jr. MkII. It's a real sleeper. Best value for under $1K tube amp kits, in my opinon.

Report message to a moderator

Re: Why SE in SET amps? [message #11386 is a reply to message #11384]

Wed, 05 September 2007 07:12

SteveBrown
Messages: 330
Registered: May 2009

Grand Master

Dick, while it isn't exactly a kit, there is a really nice design called the SETH that you can find info on at the Magnequest website. This is a PP triode design using 2A3's. I built it using PP iron salvaged from a Fisher 500C receiver, 6550's strapped in triode mode, and in a stereo chassis. I have to say, it sounded great. The secret here is the phase splitting transformer, which provides a very elegant way to split the phase, is easier to implement than tube splitters, but costs more. At any rate, others who have heard my PP amp agreed that it was one of the better PP designs they'd heard. I've built lots of PP and SE amps and they each have their charms. The PP stuff definately seems to control the bass better, and the SETh design maintains the integrity of the mids and highs as well - without making them sound strained or dead, as many amps using feedback seem to do. Anyway, hope that helps. The MQ site has extensive info on building one of these, and again, you don't have to be stuck using just MQ iron (though it is very good). I believe Jack at ElectraPrint also makes a splitter like this. Best of luck!

SETH @ MQ

Report message to a moderator

Re: Why SE in SET amps? [message #11387 is a reply to message #11384]

Wed, 05 September 2007 21:53

granch
Messages: 118
Registered: May 2009

Viscount

I want to thank all of you for the helpful suggestions. Actually I am not in the market and was just being curious. When I buy amps in the next year or so (if I do) they will be great big momma's for bellowing my organ concerts to the lake I live on. Like 2500 watts for sub woofers and I am still figuring alternatives for the mids and tweets. Mids will be 4 Altec A-7s, per channel (8 in all) and tweets will be the horns that come with the A-7s - but driven separately with a 3 way electronic crossover. I have been using 3 A-7's per side driven by 200 watts per channel with no subwoofers. I want it louder with less strain on the equipment, but have to be careful because my property is atop a 20 ft sand cliff and there is no earthquake insurance up here in northern MN.
-Dick

Report message to a moderator

Re: Why SE in SET amps? [message #11388 is a reply to message #11382]

Fri, 21 September 2007 22:59

RC Daniel
Messages: 18
Registered: May 2009

Chancellor

Hi Dick

I have not posted over here for a long time; anyhow...

I am thinking that (within reason) some folks may be more sensitive to the ratios of odd : even harmonc distortions, rather than absolute levels. Could a reduction in 2nd harmonic distortion expose odd-order artifacts or somehow make them more obvious / offensive?

Hmmm, perhaps this is analagous to dietary intake of omega6 : omega3 fatty acids or blood HDL : LDL ratios and their effects on health?

Got to run
Raymond

Report message to a moderator

Re: Why SE in SET amps? [message #11389 is a reply to message #11388]

Fri, 21 September 2007 23:35

granch
Messages: 118
Registered: May 2009

Viscount

Howdy!
It could be. I know that even harmonic disrtortion is much more tolerable than odd, so that its presence could possibly mask the odd. I know of no studies on this. But there are other forms of distortion that are very bad news. I haven't seen anyone mention "intermodulation distortion" (IM). That is supposed to be very bad news as it produce all sorts of trash besides harmonics. It occurs when two or more tones are fed through a non-linear (ie. distorting) device (like an amp). It produces sums and differences of the two (or more) tones which vary with the signals. I don't know why it isn't mentioned anymore - perhaps because its easier to sweep under the rug and forget. I believe that its one kind that negative feedback amps are good at suppressing, since the distortion products are generally not in the original signal. There is also FM or doppler distortion caused by the loudspeakers themselves. Think of a cone vibrating at high frequency while it is simultaneously being driven slowly back and forth by a low frequency. Causes the doppler effect or a frequency modulation of the higher freq by the lower. That's one reason why horns sound so good - they minimize the excursion of the diaphragm by loading it heavily and uniformly. Also a good reason for sub woofs to only cover a narrow frequency range. I believe that the common answer to all these problems is to make the amp as absolutely linear as possible. A very well designed amp with feedback run well within its ratings (both bandwidth and power) can be the nearest thing to linear that I know of.

Report message to a moderator

Re: Why SE in SET amps? [message #11390 is a reply to message #11389]

Sun, 23 September 2007 16:35

RC Daniel
Messages: 18
Registered: May 2009

Chancellor

>>>A very well designed amp with feedback run well within its ratings (both bandwidth and power) can be the nearest thing to linear that I know of.

Report message to a moderator

Re: Why SE in SET amps? [message #11391 is a reply to message #11390]

Sun, 23 September 2007 20:49

granch
Messages: 118
Registered: May 2009

Viscount

There is a great amount of truth in what you say. As my sainted mother used to say, "'Every one to their own taste', said the old lady as she kissed the cow." There are also things like the placebo effect. And there is a big variation in the hearing acuity of various people.
Due to a common old service injury, I am now losing my hearing to the point where I have to use hearing aids in both ears to help understand speech. But for music listening I have to remove the aids. And I am still busily designing sound systems.
I used to make a lot of distortion measurements (total distortion is easy to measure). The distortion levels got so low that I couldn't measure them anymore with my dated equipment.
My original question was why only single ended triode amps when push pull triode had lower distortion.
-Dick

Report message to a moderator