[SOLVED] TL431 as Negative Reference

Hi,

I found this method of generating a negative voltage reference from a TL431:

**broken link removed**



It works, both as a -2.49V reference or as whatever you like with a divider, e.g. -5V out.

I tried to understand the detailed schematic, and the functional diagram, and do not understand how this works, as a few NPNs that when used normally as a positive reference now seem to be receiving a low signal, and I'm wondering why there are a few transistors receiving what looks to an untrained eye as 0V power supply. I am not versed in understanding Op Amp architectures at all, so don't really understand the positive use either, but can get some idea of what will be powered and "on" or "off".

Two questions please:

Is it okay to do this or is it Not Recommended for Any Designs?
How does the internal OpAmp still work? is it because the V+ and V- pins still see a difference in potential and the polarity doesn't matter?





Thanks.

- - - Updated - - -

...This makes me wonder if a dual supply Op Amp that never needs the output to swing above 0V could be powered from 0V and -5V (rather than, for example, +2.5V and -2.5V), because all the IC sees/requires is sufficient difference in potential between its power rails to function correctly. Is that so?

The final schematic (negative voltage regulator) is incorrect, it won't work properly at all. The concept of using a constant current source is it lowers the impedance at the shunt regulator and hence lessens the effect of it's dymanic resistance. Feeding voltage to the output pin of an LM317 and expecting it's input pin to be a constant current source is a total nonsense. If it works at all it's only due to leakage through the regulator.

The TL431 schematics are identical, but turned upside down. From the regulators point of view the polarities are the same, it's just the voltage hangs below zero instead of sitting above it.

Brian.

I installed an op amp to serve as a mic pre-amp in a cassette deck. Looking for the best spot to hook up the power supply, I found 0V ground on the board, and then I chose a convenient negative supply. The reason this worked is because I put series capacitors at the input and output, to block DC.

betwixt said:

The final schematic (negative voltage regulator) is incorrect, it won't work properly at all. The concept of using a constant current source is it lowers the impedance at the shunt regulator and hence lessens the effect of it's dymanic resistance. Feeding voltage to the output pin of an LM317 and expecting it's input pin to be a constant current source is a total nonsense. If it works at all it's only due to leakage through the regulator.

The TL431 schematics are identical, but turned upside down. From the regulators point of view the polarities are the same, it's just the voltage hangs below zero instead of sitting above it.

Brian.

Click to expand...

:O (OMG) I hadn't looked closely at the other schematics using the voltage regulator. That's why I asked - "this seems to work" solutions are not what I want to use for anything. I'll just take the verb "hangs" as opposed to "sitting" as not having negative connotations. Thanks, Brian.

The schematic you posted of the TL431 is just fine and is a valid way of making a negative supply.

The way to understand it is to first understand the TL431 when used like a zener diode (or just accept that it works). Then note how a zener diode can also easily be used to make both positive and negative supplies. Then replace the zener's with TL431s and you have it:



It didn't really help that the schematic drew the part upside down. I would have drawn it with ground on top.

The TL431 is one of my favorite parts precisely because of how many different ways it can be used and a negative supply is one of them. Of course it comes with the disadvantages of a shunt regulator like poor efficiency. But sometimes that doesn't matter. It's cheap, especially if you already have it on your BOM or kit.