Do tubes do anything well in circuits?

[snaimpally]

Around here, my impression is that tubes are considered distortion generators and little else. In my time with tube amplifiers I remember reading that there were certain things that tubes do better than solid state components. I know nothing about electronics but am interested in whether there is any truth to that, and if so, could whatever that is be incorporated in circuit design.

The principles of amplification don't change going from tubes to transistors. Generally, transistor based circuits can provide more gain. If designed correctly, both tube amps and transistor amps should be relatively distortion free - neither tubes nor transistors should be generating audible distortion. One important difference between tubes and transistors is the nature of the distortion. Tubes produce even order distortion, which is subjectively more pleasing to human ears. In contrast, transistors produce odd order harmonics, which are NOT pleasent sounding. This is where you get people talking about tubes sounding "warm" and transistors sounding "cold".

 

[dfuller]

tubes produce even order distortion, which is subjectively more pleasing to human ears. In contrast, transistors produce odd order harmonics, which are NOT pleasant sounding. This is where you get people talking about tubes sounding "warm" and transistors sounding "cold".

Not true at all. Tubes and transistors both produce even and odd order harmonic distortion. In fact push-pull circuits effectively cancel a lot of even order harmonic distortion.

 

[MakeMineVinyl]

Industrial apps: variable ac source to drive syncro/gyro devices for aircraft/missle guidance systems....

You'd need a Saturn missile to get that amplifier off the ground.

 

[DanielT]

Cool:

NASA develops speedy vacuum tube prototype for computers in space
Researchers have developed a vacuum tube prototype that is both fast and tough against radiation, making it ideal for computers in space.

NASA develops speedy vacuum tube prototype for computers in space

Researchers have developed a vacuum tube prototype that is both fast and tough against radiation, making it ideal for computers in space.

www.theverge.com

 

[SIY]

Tubes produce even order distortion, which is subjectively more pleasing to human ears. In contrast, transistors produce odd order harmonics, which are NOT pleasent sounding. This is where you get people talking about tubes sounding "warm" and transistors sounding "cold".

So often asserted. And, of course, totally wrong. Harmonic composition is a function of topology, not active device.

 

[DanielT]

So often asserted. And, of course, totally wrong. Harmonic composition is a function of topology, not active device.

In addition, there is probably no type of good distortion. As far as I know.

If a transistor-based amplifier is well designed and has an inaudible distortion, the case tube amplifier vs transistor amplifier is closed.I think at least.

Although you can do tubes because it's fun. For the sake of sound history. Because they shine cozy in the autumn darkness.

Yes! This positive side of tube and distortion exists! But then you have deliberately created an art expression that you want to sound in a certain way and that is a completely different thing.

 

[FrantzM]

Ha! I remember Arny Kruger making the same declaration.

As someone who enjoys using tube amps, I'm happy that there are numerous tube amp manufacturers who don't share that view

(Which isn't meant to deny much of the technical claims in this thread about tubes/SS aren't true).

Well... the manufacturers know there is no need. They, however have the clear cognition that there is a demand for tubes, in spite of high and lack of performance

 

[DonH56]

Technically, IIRC, the Maclaurin expansion for a tube is factorial, and exponential for a bipolar transistor. That means tubes could have intrinsically lower distortion than a BJT. As @SIY et. al. have said it is all about the implementation. Ages ago I measured my ARC SP3a1 and its distortion was on par or better than the SS preamps we had on hand, which included Mark Levinson and a few others highly-regarded in their day.

As for small size, there were some little "bead" tubes about the size of a Nuvistor or smaller I remember from long ago. We all called them bead tubes because they looked like the beads on a necklace (and about the same size). They were used in an RF receiver I had to work on at one point.

Purportedly the smallest tube per Google (though I seem to recall the little bead tubes were even smaller -- they were rounder):

 

[DonH56]

Cool.
"The MA-3's power supply has a built-in AC line processor that regulates the line voltage coming into the amp! As a result, line voltage has no effect on the MA-3; it will make full power even if the AC line level has dropped to only 100 Volts."

Ralph (the guy behind Atma-Sphere) is a sharp guy. Nice to see OTL survived after (perhaps in spite of?) Futterman.

 

[Tom C]

Ralph (the guy behind Atma-Sphere) is a sharp guy. Nice to see OTL survived after (perhaps in spite of?) Futterman

I always thought the output transformer was a desirable thing, useful for impedance matching. Are there any advantages to OTL? I thought that was just BS.

 

[DonH56]

I always thought the output transformer was a desirable thing, useful for impedance matching. Are there any advantages to OTL? I thought that was just BS.

A never-ending debate. Transformers have disadvantages like nonlinearity, hysteresis, limited bandwidth, size and weight, etc. Matching impedance makes more sense when the impedance is fairly flat, which does not describe most speakers' impedance (though great for matching tube plates to "near" speaker impedances). Etc.

OTL has its own set of pros and cons, natch. You can get much lower output impedance, which is desirable for many speakers, but the design is more complex (at least mine were) for stability and reliability. I played with OTL amps in my youth with mixed results. I had one that sounded decent, using an opamp to control the output common-mode to reduce DC offset, but when the opamp failed all heck broke loose and it took out amp and speaker. I rebuilt it but, other than testing, never really used it again.

 

[MRC01]

I always thought the output transformer was a desirable thing, useful for impedance matching. Are there any advantages to OTL? I thought that was just BS.

If you're driving a high impedance load, like a headphone amp driving Sennheiser HD-580/600/650/660 which are 300+ ohm, the OT may not be necessary. Over 20 years ago I had an amp like that, Pete Millet's Wheatfield HA-2.

 

[DanielT]

Guys, what do you think about this?

Okay now I came up with a question!
When was it possible to create transistor-based amplifiers with lower distortion than tube amplifiers?

Historical expose. Development of better performing transistors for hifi.

I was intrigued and did some quick "bathroom" reading on Luxor. I had never heard of them. It seems they had a long history of building products many of which they designed themselves, though I expect some under license as well. They did rebadge Luxman receivers, but the one in the link does...

audiosciencereview.com

 

[MakeMineVinyl]

There's also the NuTube from Korg. I get the feeling though that low distortion is not the point of this device; they say it has 'the same rich overtones' (in other words distortion) as traditional tubes. Never mind that traditional tubes do NOT have 'rich overtones' unless implemented to do so. At any rate, I think its more of an effects device rather than a serious amplifying device.

 

[SIY]

A never-ending debate. Transformers have disadvantages like nonlinearity, hysteresis, limited bandwidth, size and weight, etc. Matching impedance makes more sense when the impedance is fairly flat, which does not describe most speakers' impedance (though great for matching tube plates to "near" speaker impedances). Etc.

OTL has its own set of pros and cons, natch. You can get much lower output impedance, which is desirable for many speakers, but the design is more complex (at least mine were) for stability and reliability. I played with OTL amps in my youth with mixed results. I had one that sounded decent, using an opamp to control the output common-mode to reduce DC offset, but when the opamp failed all heck broke loose and it took out amp and speaker. I rebuilt it but, other than testing, never really used it again.

The main thing getting rid of the transformer gets you is the ability to use much more negative feedback. The Futtermans used some stupidly high amount to get their distortion down. Ralph Karsten's designs are much better.

 

[DonH56]

The main thing getting rid of the transformer gets you is the ability to use much more negative feedback. The Futtermans used some stupidly high amount to get their distortion down. Ralph Karsten's designs are much better.

Thanks, was hoping you would chime in! I should have said that explicitly instead of implying it with lower output impedance (which is also due to a bunch of cathode followers in parallel).

I struggled with my long-ago designs, partly just inexperience (still in college), but a large part was because mine were still pretty load-dependent despite higher feedback, and I was trying to drive ESLs, so getting decent stability was a PITA. I made a lot of nice oscillators...

I don't really know Ralph (nor he me) but we've exchanged messages on another forum and he seems a really sharp guy. Also not prone to the usual tube b.s. marketing, rather a solid engineering kinda' guy.

 

[egellings]

Maybe it's "the little engine that could" appeal of tube amps that makes them stick their little chests out at the muscular S.S. competition.

 

[Cbdb2]

A never-ending debate. Transformers have disadvantages like nonlinearity, hysteresis, limited bandwidth, size and weight, etc. Matching impedance makes more sense when the impedance is fairly flat, which does not describe most speakers' impedance (though great for matching tube plates to "near" speaker impedances). Etc.

OTL has its own set of pros and cons, natch. You can get much lower output impedance, which is desirable for many speakers, but the design is more complex (at least mine were) for stability and reliability. I played with OTL amps in my youth with mixed results. I had one that sounded decent, using an opamp to control the output common-mode to reduce DC offset, but when the opamp failed all heck broke loose and it took out amp and speaker. I rebuilt it but, other than testing, never really used it again.

I think that the transformer disadvantages are part of the "tube sound" in a lot of amps.

 

[mhardy6647]

The amp has a handy bias meter and switching to measure the individual output tubes. The amp is very stable and rarely requires adjusment. Unlike most power amplifiers of the period, these amps could and were run at full rated output continuously in industrial applications.
View attachment 165764

Yup.

Tubes produce even order distortion, which is subjectively more pleasing to human ears. In contrast, transistors produce odd order harmonics, which are NOT pleasent sounding. This is where you get people talking about tubes sounding "warm" and transistors sounding "cold".

Nope.

So often asserted. And, of course, totally wrong. Harmonic composition is a function of topology, not active device.

Yup.

 

[Doodski]

Does anybody know what the component circled in red is/does?