TL431 as an audio amplifier.

[schmitt trigger]

You have read correctly: TL431.

At first, I thought it was an April Fool's joke.

But apparently there are a significant amount of these circuits in the web.
For instance:
http://vu2ppp.com/_simple_audio_amplifier_using_tl431

One day, I might breadboard one, just out of curiosity.

[Raj]

Amature radio folks...Perfecting the art of using components in surprising ways since 1920s

[Kleinstein]

I have not heard of the TL431 used for this before, but I have seen the LM317 used for this.

[schmitt trigger]

Thoughtfully, they concede that it is not an audiophile grade amplifier.

[amyk]

Amature radio folks...Perfecting the art of using components in surprising ways since 1920s

Any nonlinear component can detect radio signals. Any active component can have gain.

[T3sl4co1l]

Why not?  It's not an "adjustable zener", whatever that means; it's a proper op-amp.

Albeit a somewhat quirky one: whopping (yet suspiciously stable) 2.50V input offset voltage, too few pins (+in is tied with VEE), self powered through the output -- which is an open-collector output, too?

It's rated for 36V and 100mA so quite capable of real power, limited by package dissipation more than anything.  The OC output limits operation to class A, of course.

Another way to think of it: an ideal BJT, if a slow one (fT ~ low MHz).  Precision Vbe, hFE > 25k (practically infinite).  Acts more like a precision Darlington actually, as "Vce(sat)" isn't very low (about 1.9V).  Effectively, internal Baker clamped.  (Don't know offhand what saturation recovery and rise/fall time are like, aside from multiple ~us.)

Tim

[floobydust]

This 0-15V 5A power supply uses four TL431's instead of any op-amps:
https://www.eevblog.com/forum/beginners/cheap-chinese-adjustable-lab-power-supply-kit-on-ebay-(-3ag1-xiaolin-)/msg1412511/#msg1412511

[johnwa]

You have read correctly: TL431.

At first, I thought it was an April Fool's joke.

But apparently there are a significant amount of these circuits in the web.
For instance:
http://vu2ppp.com/_simple_audio_amplifier_using_tl431

One day, I might breadboard one, just out of curiosity.

Yeah, I saw Charles Wenzel's "one transistor" radio , meant to give it a go. I guess you could use a ZN414 or equivalent, and make a "two transistor" radio that should go pretty well.

[aidan.walton]

In fact I have used a 431 exactly in this mode within a conversion I did to an older guitar amplifier. The problem I faced and the reason for using the 431 I think is quite interesting. I will share......
The original power output stage of the amplifier was a class AB 'cooker' I was converting the unit to run on batteries, so I replaced the class AB with a class D output stage for improved efficiency. However the pre-amplifier stages all ran off a split rail supply of +-15V. Unfortunalty the class D IC required a single rail supply and to obtain the power out of the IC it required the full 30V of the supply and I had to span the class D IC across the entire split rail. This had the unfortunate consequence that the input stage to the power amp was now signal referenced to the -15V rail. The pre-amps where signal referenced to the 0V midpoint. At switch on, capacitivly coupling the pre-amp to the power amp caused (or would have caused) enormous thumps and prolonged DC offsets on the output from the power amp as the 2 stages setteled into their respective reference levels. The power amp input impedance being quite high meant these setteling times would have been enormous. To get around this I used a TL431 and drove its reference input coupled from the pre-amp stage and its anode on the -15V rail and its cathode (pulled up to the positive rail through a suitable resistor) acting as the drive stage into the power amp. In this way as the power rails came up and diverged, the TL431 rapidly pulled the power amps input signal down to the correct reference level (well much closer to it) as the power amp had a quiescent input voltage of ~3V. So now I only needed to capacitively couple across of 1V or so difference. Goodbye problem.

My experience is that when driven carefully the TL431 is a very reasonable amplifier and a prefectly good op-amp.

[David Hess]

self powered through the output -- which is an open-collector output, too?

See?  There is nothing unusual about configuring an operational amplifier with the output as an input and each power lead as an output.  Right?  RIGHT?

[SiliconWizard]

Well, you can even make linear amps out of CMOS gates. Not extremely linear either, but that's fun too.

[Refrigerator]

I tend to think of TL432 as a perfect MOSFET.

[StillTrying]

Here's a 555 radio.

http://electronics-diy.com/am-radio-lm555.php

[daqq]

Cool!

...so... is this component abuse? Do we call PETA or something?

[magic]

Thoughtfully, they concede that it is not an audiophile grade amplifier.

Class A, predominantly 2nd harmonic distortion if any, what's not to like?

[exe]

I was surprised by such circuits too until I realized it has ~1MHz bandwidth, more than enough for audio or a power supply. TI's datasheet draws the part a zener, hence, probably confusion. Other datasheets draw it as an opamp with reference and output transistor, which should  accepting such circuit . E.g., look here: https://www.onsemi.com/pub/Collateral/TL431-D.PDF .

I'm not sure how practical such circuits because there are plenty opamps that are cheap and designed for audio (e.g., tl071 or ne5532). Nonetheless, I don't mind listening such amplifier, I don't think I'll hear a difference.

...so... is this component abuse? Do we call PETA or something?

Not according to me . Who's PETA?

[AVGresponding]

I was surprised by such circuits too until I realized it has ~1MHz bandwidth, more than enough for audio or a power supply. TI's datasheet draws the part a zener, hence, probably confusion. Other datasheets draw it as an opamp with reference and output transistor, which should  accepting such circuit . E.g., look here: https://www.onsemi.com/pub/Collateral/TL431-D.PDF .

I'm not sure how practical such circuits because there are plenty opamps that are cheap and designed for audio (e.g., tl071 or ne5532). Nonetheless, I don't mind listening such amplifier, I don't think I'll hear a difference.

...so... is this component abuse? Do we call PETA or something?

Not according to me . Who's PETA?

PETA is a charitable organisation, People for the Ethical Treatment of Animals.

Following this reasoning we should have PETE, People for the Ethical Treatment of Electronic.

[schmitt trigger]

I am sure most of us in this forum would be guilty.

Who here hasn’t overstressed a component until the magic smoke comes out?
Sometimes on purpose 

[tszaboo]

Wow, with all these circuits posted here, I wonder if anyone can build an audio amplifier with the NE5532.

[Refrigerator]

I am sure most of us in this forum would be guilty.

Who here hasn’t overstressed a component until the magic smoke comes out?
Sometimes on purpose 

One time when i couldn't be bothered to fetch a BJT i used TL432's instead because they were on hand. 

[David Hess]

Wow, with all these circuits posted here, I wonder if anyone can build an audio amplifier with the NE5532.

It has been done several times:

https://eecaldas.wordpress.com/2017/10/02/ne5532-power-audio-amplifier/

[tszaboo]

Wow, with all these circuits posted here, I wonder if anyone can build an audio amplifier with the NE5532.

It has been done several times:

https://eecaldas.wordpress.com/2017/10/02/ne5532-power-audio-amplifier/

I guess the irony didn't really came through.
"Using an audio amplifier to amplify audio? Unheard of!"

[exe]

I guess the irony didn't really came through.

I think it did .

[magic]

https://eecaldas.wordpress.com/2017/10/02/ne5532-power-audio-amplifier/

These puppies are RC4558, though


AliBay is the final mythbust to all audiphoolery

[harerod]

Cool!

...so... is this component abuse? Do we call PETA or something?

Any LTspice users here? The appropriate organization would be the SPCS.