transistor linearity

[robotmaster]

If transistors weren't linear they couldn't amplify triangle waves, or any wave that well.

But here is a virtual sim of Falstad with a triangle passing perfectly through the transistor,  have I done this right?   What do people mean by transistors not being linear?

Here is the link->

https://www.falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWcMBMcUHYMGZIA4UA2ATmIxAUgoqoQFMBaMMAKADcQUUAWTvPTtyop+UMUOpUp0BCwBOfASIG5eysWHjyQ2NJwSEdezCjExIrDruE8jwiVJASkjmLIDudwVWv7DkFk9CDFNsPF5g0L0A7AwqSK8EtAEIKUCdcM58L2SodN9cgtsYuPAwU3UuNVFUvM9mCpryrwD65oLRX1bFL19VPIAXHUhqlWNRKggGbGgw7mwEbGIwhDB+VOg48LI8SEIw4gRSYmQqABM6ADMAQwBXABsB9MJsuZAEt+633KTsgKG+iMdNhDP1JiAGNxYMQRKsMNxuMR8NwNgQMIRNItdAghKsUnAQBcbg8np5sCDehTPvkqZkXipMl8KepyYYTHVgWyQjluQEODxhKIUAYej4nLQzFAZOlhWyhfKBAEAPbvd68KhCMjUTQoaDqziq7AsMJiABitQlTBAADUlY9rgBzOgsIA

[Andy Chee]

What do people mean by transistors not being linear?

Context is important.  Where did you read that information from?

I'm going to guess it's in relation to transistor hard clipping on audio peaks in comparison to vacuum tubes soft clipping?

Please provide a link for context.

[MarkT]

The emitter current is an exponential function of base-emitter voltage - very non linear.

However the collector current is approximately linear with base current.

So it depends if the signal is a current or a voltage.

[robotmaster]

What do people mean by transistors not being linear?

Context is important.  Where did you read that information from?

I'm going to guess it's in relation to transistor hard clipping on audio peaks in comparison to vacuum tubes soft clipping?

Please provide a link for context.

For example here->
https://forum.allaboutcircuits.com/threads/explanation-of-small-signal-vs-large-signal.105987/

And here->
https://electronics.stackexchange.com/questions/502016/proof-of-linearity-of-a-transistor

[CaptDon]

A transistor may be extremely linear across a portion of its operating curves and very non-linear in other areas. NO blanket statement or assumption can be made. The circuit surrounding the transistor can assist in correcting non-linearities. Sort of why one transistor usually has a bunch of friends nearby!!

[Zero999]

It's nothing but a magic smoke generator. The current will be 83A. It's classic a case of playing around with a simulator, and lack of understanding.

What exactly are you trying to achieve or learn?

[robotmaster]

I'm just shorting 5 volts through the Darlington pair,  that shouldn't kill a transistor, you are allowed to short through from collector to emitter.   Its only that high in amp because it hasnt gone through a load yet, its just a bare wire.

[Ian.M]

I'm just shorting 5 volts through the Darlington pair,  that shouldn't kill a transistor, you are allowed to short through from collector to emitter.   Its only that high in amp because it hasnt gone through a load yet, its just a bare wire.

Err . . .  *NO*.   See BJT SOA.  Go outside the SOA and the transistor dies. The further outside, the quicker it dies.  See the datasheet of your transistor for its SOA.  The SOA is *never* a simple rectangle.  This means that a 50V 5A transistor will pop if you believe the marketing 'specs' on the front page of the datasheet can be applied simultaniously, and try passing 4.9A through it with 49V across it!   

Also the SOA is normally specified at 25°C ambient for small signal devices or heatsink temperature for power devices.  If its hotter you have to derate the power proportionally to how far past 25°C it is towards T_j(max). 

[Zero999]

I'm just shorting 5 volts through the Darlington pair,  that shouldn't kill a transistor, you are allowed to short through from collector to emitter.   Its only that high in amp because it hasnt gone through a load yet, its just a bare wire.

The load is part of the circuit. It won't work like that. Sorry, it's a nonsensical circuit. It isn't even clear how it's all connected up, with the overlapping components and lack of a 0V reference point.

BJTs are only linear over a small current range. Here's an example which demonstrates the output being more linear when the input is a lower amplitude.

[robotmaster]

I just was a bit confused about the issue, I heard transistors were non linear,  but if it were a lot so that way then they wouldn't even be able to amp anything! They would be unusable for anything but binary.

They are approximately linear, because a linear output is the ideal output you want,  so transistors do work for scalar values!  so they work good for analogue computers, its just they are a bit off on the numerical reading by a little,      how much being how far off a perfect triangle amplification it is.

Heres my sim,  I've fixed it with a load this time.
https://www.falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWcMBMcUHYMGZIA4UA2ATmIxAUgoqoQFMBaMMAKADcQUAWKsDQzj3DEUUMULBwxVGAhYAXTt0FVuqobxANc0DJAxdK2LvwTMu2KLshcR2DAhOYMeQkklUAJnQBmAQwBXABt5FgB3FU4EATVOPDwocMiUePBCVVTIJNiUhNy4hKyI-LAuLkjS8qKKsor+ROKhNDym5SyAJwqRZPwxDyz7DLzUktqIGXZwetiwbtiqSw1oJBkrOUHwWtHymekksHS6mI1J2dEc3vnxGWXpNZZO5uTlK-79rdSDocSOSorDq5UJYCVayFgAexAxBAhCqIAAHnhJMxqDBxpwMVRoZYYJBREwwCgIeAILCxIjkRAVrBIGBoih3BjRLwIDiabTSJyudziAkGKQWEA

[ejeffrey]

There are a bunch of ways to measure gain of a transistor, current gain, voltage gain, transconductance gain, and more.  They are all nonlinear to different extends but it doesn't matter.

Transistor amplifiers are almost always used with feedback.  Even simple emitter degeneration is a form of feedback.  Feedback linearized the circuit and so the secret to good linear behavior is lots of feedback.  That requires lots of gain.  So transistors male good linear amplifiers not because they are inherently linear but because they have lots of gain.

[Zero999]

I just was a bit confused about the issue, I heard transistors were non linear,  but if it were a lot so that way then they wouldn't even be able to amp anything! They would be unusable for anything but binary.

They are approximately linear, because a linear output is the ideal output you want,  so transistors do work for scalar values!  so they work good for analogue computers, its just they are a bit off on the numerical reading by a little,      how much being how far off a perfect triangle amplification it is.

Heres my sim,  I've fixed it with a load this time.
https://www.falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWcMBMcUHYMGZIA4UA2ATmIxAUgoqoQFMBaMMAKADcQUAWKsDQzj3DEUUMULBwxVGAhYAXTt0FVuqobxANc0DJAxdK2LvwTMu2KLshcR2DAhOYMeQkklUAJnQBmAQwBXABt5FgB3FU4EATVOPDwocMiUePBCVVTIJNiUhNy4hKyI-LAuLkjS8qKKsor+ROKhNDym5SyAJwqRZPwxDyz7DLzUktqIGXZwetiwbtiqSw1oJBkrOUHwWtHymekksHS6mI1J2dEc3vnxGWXpNZZO5uTlK-79rdSDocSOSorDq5UJYCVayFgAexAxBAhCqIAAHnhJMxqDBxpwMVRoZYYJBREwwCgIeAILCxIjkRAVrBIGBoih3BjRLwIDiabTSJyudziAkGKQWEA

Yet another unintelligible schematic, so I can't comment.

If you can't use the software properly, please draw it on a piece of paper and photograph/scan it.

It still isn't clear what you're trying to do.

Is it supposed to be a practical circuit, or is it just a thought experiment?

Here's another example showing the non-linearity of a common emitter amplifier. V2 is swept from 500mV to 650mV and the output plotted. Note how it's not a straight line, more exponential, relative to V1, and that the transistor saturates, when the output voltage is just above zero. Rser is the equivalent series resistance of V2. It's there to add 100R to the base, the same as adding a base resistor, but quicker, as I just had to set the properties of V2.