Are discrete transistors dead?

[Rudinhoob]

...outside integrated circuits?

Can OpAmp's replace them?

 

[Runei]

I used op-amps today and discrete transistors for controlling a DC motor :) To me they aren't dead. Also I have a magnetic levitation unit containing transistors and op-amps.

 

[Windadct]

They are inanimate - so yes technically... sorry this place can become so literal!

As for your Q -- noooo we will be using them until the day of one-off IC fabrication - even then, still needed.

 

[Rudinhoob]

But their usage is in decrease. The number of transistors in todays circuits is much less than from 80s. The same with vacuum tubes.

 

[Jiggy-Ninja]

But their usage is in decrease. The number of transistors in todays circuits is much less than from 80s. The same with vacuum tubes.

Do you have figures for your assertion? I tend to view such proclamations with skepticism.

Being the yungin that I am who has never even seen a vacuum tube in person, I can only speculate. However, I believe the main reason vacuum tubs went away is because of size, cost, high operating voltages, and heat.

Discrete transistors are still much cheaper and much smaller than op amps. Even if it is true that their usage is declining, they'd still have enough advantages that would make it bizarre to say they're dead.

About a year ago I tore apart a pair of cheap $5 walkie talkies from Toys-R-Us. There wasn't a single IC anywhere in it.

 

[berkeman]

But their usage is in decrease. The number of transistors in todays circuits is much less than from 80s. The same with vacuum tubes.

For cost-reduction reasons, we are indeed sweeping more external discrete transistors into our mixed-signal ASIC designs. Minimizing the number of external components can help to cut down on the cost of the circuitry. But there are usually a few components outside the ASIC that cannot be pulled inside for voltage or power reasons. So there will always be a few discrete transistors outside of the ASICS in many designs.

 

[Rudinhoob]

Even if it is true that their usage is declining, they'd still have enough advantages that would make it bizarre to say they're dead.

I'm not saying they are 100% dead like vacuum tubes, I mean their use is so limited outside the ICs.

 

[berkeman]

I'm not saying they are 100% dead like vacuum tubes, I mean their use is so limited outside the ICs.

Vacuum tubes aren't 100% dead. There are still some important applications where they are used.

Quiz Question for you Rudinhoob -- where are they used now?

 

[Rudinhoob]

Audio amplifiers?

 

[berkeman]

Audio amplifiers?

That would be one, although one can certainly debate whether using vacuum tubes helps the audio quality (some argue that it does, others don't think so).

There is a much more important application... think BIG TUBES!

 

[Redbelly98]

There is a much more important application... think BIG TUBES!

I am intrigued. My first thought was displays for televisions and computer monitors, then realized that no, evacuated tubes are no longer used to make them either.

 

[trollcast]

There is a much more important application... think BIG TUBES!

The magnetron in a microwave is a type of vacuum tube.

 

[Rudinhoob]

I'm now thinking some military applications where vacuum tubes are not damaged by strong electromagnetic waves? What about communication transmitter/receivers?

 

[berkeman]

I'm now thinking some military applications where vacuum tubes are not damaged by strong electromagnetic waves? What about communication transmitter/receivers?

Pretty close! They are used in high-power microwave and radar applications, as well as radio/TV broadcasting equipment. The power levels and high voltages lend themselves to vacuum tube circuitry better than semiconductor circuitry sometimes...

 

[gnurf]

[...] one can certainly debate whether using vacuum tubes helps the audio quality (some argue that it does, others don't think so).

My friend who plays the guitar in a metal band once told me that tube amps were preferred by almost everyone in his circles. Dumbfounded, I looked it up, and it turns out (IIRC) that tubes have a strong distorting third harmonic which "sounds good". Realizing that my exquisite rational skills couldn't probe into the domain of taste, I had to let it go, but I couldn't help thinking that his beloved tube amp did a little more than advertised (i.e., it's not simply amplification).

 

[jim hardy]

My friend who plays the guitar in a metal band once told me that tube amps were preferred by almost everyone in his circles. Dumbfounded, I looked it up, and it turns out (IIRC) that tubes have a strong distorting third harmonic which "sounds good". Realizing that my exquisite rational skills couldn't probe into the domain of taste, I had to let it go, but I couldn't help thinking that his beloved tube amp did a little more than advertised (i.e., it's not simply amplification).

Quite a bit of research went on in the 1990's.

One of the better reports i recall demonstrated the differing harmonic content of tube vs solidstate pre-amplifiers when over-driven into distortion. Which is how the musicians participating in the study used them, to surprise of the investigators..

Solidstate amps tend to provide more symmetric clipping than tubes because of the latter's cutoff and saturation behavior. Solidstate amps tend to be feedback controlled so they'll drive right to the power supply rails instead of approaching them in a gradual, nonlinear fashion. They clip both top and bottom peaks alike, and it begins suddenly.
Recall from your Fourier that odd harmonics affect both peaks symmetrically, even ones don't.

The investigators' Fourier analyzer showed the tube amp output when overdriven to contain more even numbered harmonic content, as a result of asymmetric clipping, than did the solidstate amp with its symmetric clipping..
So, to a musician's ear they produce different "overtones".


It made sense to me...

 

[rbj]

... one can certainly debate whether using vacuum tubes helps the audio quality (some argue that it does, others don't think so).

My friend who plays the guitar in a metal band once told me that tube amps were preferred by almost everyone in his circles.

One of the better reports i recall demonstrated the differing harmonic content of tube vs solidstate pre-amplifiers when over-driven into distortion. Which is how the musicians participating in the study used them, to surprise of the investigators..

investigators in the 90s?! pretty dumb investigators if they didn't know that guitarists were using their guitar amps for distortion.

Solidstate amps tend to provide more symmetric clipping than tubes because of the latter's cutoff and saturation behavior. Solidstate amps tend to be feedback controlled so they'll drive right to the power supply rails instead of approaching them in a gradual, nonlinear fashion.

in other words, hard clipping instead of soft clipping. that's what we call it.

They clip both top and bottom peaks alike, and it begins suddenly.
Recall from your Fourier that odd harmonics affect both peaks symmetrically, even ones don't.

or odd-symmetry distortion on a sinusoid will generate odd-numbered harmonics and even-symmetry distortion on a sinusoid will generate even-numbered harmonics.

solid-state distortion was quite odd in symmetry, so no even-numbered harmonics.

tube distortion (if it wasn't push-pull) was asymmetrical so it was a combination of odd and even symmetry distortion.

The investigators' Fourier analyzer showed the tube amp output when overdriven to contain more even numbered harmonic content, as a result of asymmetric clipping, than did the solidstate amp with its symmetric clipping..
So, to a musician's ear they produce different "overtones".

some overtones sound nice and others less nice.

but the softness to the clipping is also very important.

and there are some other weird non-linear effects with an old tube amp. the poorly-regulated power supply droops when you hit a power chord, so it clips at even a lower clip level.

there is also some reactive components in a tube amp. transformers (with hysteresis) and coupling capacitance and, in the tubes, inter-electrode capacitance. so it's a mix on non-linear components and it's not memoryless. hard to model non-linear systems with reactance or "memory". makes a very subtle difference in distortion compared to a memoryless non-linear system with the same static non-linear curves.

 

[jim hardy]

pretty dumb investigators if they didn't know that guitarists were using their guitar amps for distortion.

Yes, they were not musicians...

Thanks for your clarifications and vocabulary additions ..

This one caught my eye

there is also some reactive components in a tube amp. transformers (with hysteresis) and coupling capacitance...

another trait of inductors is their tendency to maintain current(Lenz), and a power pentode is rather a constant current device. So a tube amp likely is more of a current source(high impedance) than a voltage source(low impedance)... that's why the precaution never let a tube amp operate into open circit it'll wreck the output transformer's insulation...

Amplifiers tend to be tested with resistive load so voltage and current have same shape..
But a speaker has inductance and inertia - it's really a motor-
and a motor responds different to a high impedance current source than it does to a low impedance voltage source. I've never seen a mathematical analysis , but amplifier impedance is a term in speaker enclosure design...

so many projects, so little time...

 

[skeptic2]

Another point of view:

http://electronicdesign.com/analog-...ve-tubes-versus-devices-find-their-origins-sa

http://electronicdesign.com/analog/...ate-audio-amps-last-word-or-house-fire-part-2

 

[cepheid]

I'm confused. Isn't the answer to the question in this thread:

"Only if you bias them wrong.."

 

[rbj]

Thanks for your clarifications and vocabulary additions ..

what were the additions?

("hard" vs. "soft" clipping? "hysteresis"?)

another trait of inductors is their tendency to maintain current(Lenz), and a power pentode is rather a constant current device. So a tube amp likely is more of a current source(high impedance) than a voltage source(low impedance)... that's why the precaution never let a tube amp operate into open circit it'll wreck the output transformer's insulation...

yup, we knew about that.

Amplifiers tend to be tested with resistive load so voltage and current have same shape..
But a speaker has inductance and inertia - it's really a motor-
and a motor responds different to a high impedance current source than it does to a low impedance voltage source. I've never seen a mathematical analysis , but amplifier impedance is a term in speaker enclosure design...

and the electro-acoustic coupling of the box can be reflected back as part of the load attached to the amplifier. i know there have been studies by some folks in the Audio Engineering Society. i'll see what i can find.

 

[the_emi_guy]

...outside integrated circuits?

Can OpAmp's replace them?

Go to www.digikey.com "discrete semiconductor products"

You will see over 100,000 part numbers for sale. There are over 30,000 single FET part numbers alone.

This is not a antique shop, this is where practicing EEs buy parts that they need. There are no electron tubes for sale there.

 

[sophiecentaur]

My friend who plays the guitar in a metal band once told me that tube amps were preferred by almost everyone in his circles. Dumbfounded, I looked it up, and it turns out (IIRC) that tubes have a strong distorting third harmonic which "sounds good". Realizing that my exquisite rational skills couldn't probe into the domain of taste, I had to let it go, but I couldn't help thinking that his beloved tube amp did a little more than advertised (i.e., it's not simply amplification).

Musicians - even those who invent 'new' music all tend to be quaint old fashioned things when it comes to their amplifiers. The Valve Sound is preferred yet it is the most Un-Hifi you could ever imagine. It is common for a band to use old fashioned amps which are played into microphones and that signal goes to the massive (solid state) amplification of a stadium.
I can sympathise with this when I find it very hard to love Vinyl or Tape any more.

 

[rbj]

My friend who plays the guitar in a metal band once told me that tube amps were preferred by almost everyone in his circles. Dumbfounded, I looked it up, and it turns out (IIRC) that tubes have a strong distorting third harmonic which "sounds good".

well, anything that clips both positive and negative peaks will have a strong third harmonic and i can guarantee you that this is not unique to tube amps. solid-state amps have lotsa third harmonic so just having a strong third harmonic doesn't necessarily sound so good unless you want a sort of "solid state" sound. the solid-state sound has only even harmonics (to a sinusoidal input) because it clips at both rails virtually identically (big, sharp, nasty hard limiting).

it's because the tubes clip the positive rail differently than it clips at the negative rail (which, to the biased tube, is ground) that makes the non-linear input/output curve not particularly symmetrical. that means it can be broken into an even-symmetry component (that happens to only generate even harmonics) and an odd-symmetry component (that happens to generate only odd-numbered harmonics).

but other reasons people like the tube sound is because that the non-linearity is "softer". something you would get if the higher order terms in the Maclauren series are gone or have very small coefficients, relative to the first derivative, which is the linear term and does not distort. it's one thing when some of these "spurious" frequency components start appearing gradually as the amplitude of the input increases. but to make these harmonic components appear suddenly as the input amplitude increases, that is indicative of hard limiting and sometimes you don't want that (like for a clean, mellow, nice guitar sound).

but sometimes guitarists will drive the input to a guitar amp to some distorted level for any time they hit a note or a chord "hard enough". so you get a sort of cleaner sound for "playing nice" but it lights up the distortion when you hit it harder. this is kinda like fuzztone, but not quite the same.

with a tube amp, it might be something you would get driving it with a compressor. hit it hard and you get this nice tube distortion, hit it lightly and you get a nice, less grungy sound.

but if you do this with a solid-state amp, you would get much the same effect as what you would have with a simple hard limiter (and i mean clipping, not the audio "limiter" product that is really a compressor or AGC of some sort) with the threshold set to some adjustable level that is placed between two levels of how hard you hit the axe.

but, it's second-harmonic distortion that is the "warm tube" sound and not normally found in a solid-state amp.

Realizing that my exquisite rational skills couldn't probe into the domain of taste, I had to let it go, but I couldn't help thinking that his beloved tube amp did a little more than advertised (i.e., it's not simply amplification).