Unexpected behaviour of tl431 based regulator

[Gasboss775]

I built a simple voltage regulator, see attached, using a tl431 fed by a jfet wired as a current source and an npn buffer transistor.

I carried measurements of output resistance using the formula; Rout = ∆V/∆I. With loads between 0 and 125mA the output resistance was in fact negative, that is the output voltage actually increased ( albeit marginally ) as more current was drawn. Above 125mA it started to behave as would've been expected, that is a drop in output voltage under increasing current draw.
Can anyone explain this behaviour? I'm concerned that negative resistance could cause spurious oscillations.

 

[ronsimpson]

How much did the voltage change?
You should have very close to 5 volts output.
What is the input voltage?

Probably you don't have good grounds. Are you using a bread board? Can you send a picture? Show where you measure the voltage.
Did you use a current meter?

 

[zahwi]

The datasheet recommends 10nF between cathode and ref, for this type of circuit.

 

[Gasboss775]

How much did the voltage change?
You should have very close to 5 volts output.
What is the input voltage?

Probably you don't have good grounds. Are you using a bread board? Can you send a picture? Show where you measure the voltage.
Did you use a current meter?

The problem was due to spurious oscillations, this was fixed with a 220pF cap across the cathode and adjust terminals of the TL431.

I decided to try it with an nmosfet. I used a 555 based voltage doubler for the higher drive voltage required. The circuit outperformed the original in most respects. Very low dropout. Output resistance of 14 milliohms and good regulation. The only downside was a little extra noise from the voltage doubler.

See attached schematic

 

[Tony Stewart]

TL431 requires up to 1mA minimum load to remain in regulation, and this requirement is not met . You d not need a voltage doubler and there are single component LDO's with better characteristics.

Since increasing transistor load causes Zener bias to be diminished, you might want to add 5k to the Zener output. This may prevent the oscillation as well in the 1st case.

 

[ronsimpson]

I love the little TL431. It can do many things.

 

[Gasboss775]

I love the little TL431. It can do many things.

Yeah, I've seen it used as a small audio power amplifier. I'm sure there are other possibilities too.

 

[ronsimpson]

It is like having a 2.5V reference and an op-amp in one package.

 

[spec]

I built a simple voltage regulator, see attached, using a tl431 fed by a jfet wired as a current source and an npn buffer transistor.

I carried measurements of output resistance using the formula; Rout = ∆V/∆I. With loads between 0 and 125mA the output resistance was in fact negative, that is the output voltage actually increased ( albeit marginally ) as more current was drawn. Above 125mA it started to behave as would've been expected, that is a drop in output voltage under increasing current draw.
Can anyone explain this behaviour? I'm concerned that negative resistance could cause spurious oscillations.

Hi Gordon,

Is there an error in Q1 part number on your schematic, because I'm pretty sure a 2N5858 is an old high voltage bipolar juction transistor?

 

[Gasboss775]

Yeah, sorry the jfet was a 2n5458.

I have amended this error in the schematic on post #4

 

[spec]

Yeah, sorry the jfet was a 2n5458.

I have amended this error in the schematic on post #4

No need to appologise Gordon.
Can I say that your circuit is a very good principle, but there are practical problems. I hadn't posted them because it may seem like I am a smart arse and finding fault with your good work.
But if you would like to hear my deliberations, I will post. Might take me some time though. Right now I'm up to my eyeballs with another problem on ETO.

spec

 

[ronsimpson]

Right now I'm up to my eyeballs with another problem on ETO.

Being a "smart arse" some where else. LOL Now I am the one being an arse.

Many of the members use google translate, How does "arse" translate?

 

[spec]

Being a "smart arse" some where else. LOL Now I am the one being an arse.

Many of the members use google translate, How does "arse" translate?

Nothing too bad I hope

How dare you imply I'm being a smart arse somewhere else- more like dumb arse

 

[ronsimpson]

more like dumb arse

Google translate: ass or arse = donkey.
It is hard to insult someone that is using Google Translate to read/write this forum.
You are a back end of a ass.=You're butt backend.
You're a dumb ass.=You're a dumb ass. (that works)

 

[spec]

Google translate does some stange things. I thought I would be a smart arse, sorry rearend of a donkey, and translateed Hello, good day to you when PM ing Elerion (three tran saw tooth generator thread) and he said it meant hello hope you have a nice day. I spoke to a Spanish room cleaner, in Spanish, at Tenerife once . She imediately pulled the bottom of her skirt down and ran off babling. To this day I have no idea what I had said- still I don't understand women any way!

Interesting to learn that many of the ETO members translate. It always amazes me how bright non native Enlish speaks are handling some of the complex issues on ETO. I have enough trouble and I'm a native speaker.

 

[ronsimpson]

I spoke to a Spanish room cleaner, in Spanish, at Tenerife once . She imediately pulled the bottom of her skirt down and ran off babling.

You probably had $10.00 in you hand.
.....and you thought you would give her a good tip for cleaning the bed. Good thing you checked out before her boyfriend came back.

 

[Gasboss775]

No need to appologise Gordon.
Can I say that your circuit is a very good principle, but there are practical problems. I hadn't posted them because it may seem like I am a smart arse and finding fault with your good work.
But if you would like to hear my deliberations, I will post. Might take me some time though. Right now I'm up to my eyeballs with another problem on ETO.

spec

It was just a quick experiment, not actually intended for anything specific. I was going for a sort of minimum number of parts to do the job concept, without actually using a ready made regulator like an lm317, etc, but feel free to offer suggestions on how it could be improved.

 

[spec]

You probably had $10.00 in you hand.
.....and you thought you would give her a good tip for cleaning the bed. Good thing you checked out before her boyfriend came back.

... more like 1$

(Sorry- missed your and Gordon's posts- alerts not working well for me)

 

[spec]

It was just a quick experiment, not actually intended for anything specific. I was going for a sort of minimum number of parts to do the job concept, without actually using a ready made regulator like an lm317, etc, but feel free to offer suggestions on how it could be improved.

Hi Gordon,

Apologies for not replying sooner, but I didn't get an alert for your and ron's posts.

OK, will have a detailed look at your circuit. My initial feeling is that the device operating points are a bit difficult and that the circuit is too good. I think it has massive voltage gain which would be quite difficult to tame. The normal symptom with that situation is spurious oscillations typically between 4 and 11 Mhz. I have much experience with this and had a bin of dead components to prove it.

 

[Gasboss775]

Hi Gordon,

Apologies for not replying sooner, but I didn't get an alert for your and ron's posts.

OK, will have a detailed look at your circuit. My initial feeling is that the device operating points are a bit difficult and that the circuit is too good. I think it has massive voltage gain which would be quite difficult to tame. The normal symptom with that situation is spurious oscillations typically between 4 and 11 Mhz. I have much experience with this and had a bin of dead components to prove it.

Spec did you ever cast an eye over this circuit again? I'm not in a hurry to get it up and running, it was just an idea. Perhaps it could be used as a linear post regulator along with some of my / our SMPS circuits.