lm317 cant changing adjust pin voltage

[qwertyqwq]

hi all

i have a problem about lm317 output voltage changing. here the scheme of my circuit. this is just a prototype but i mentioned something weird.
i have 28v dc voltage in input and that input also connected to V+ pin of lm324. In scheme all voltage sources explained and they are controlling with pots.

here is the problem;
when i try to increase the output voltage of lm317 with voltage controll pot it is doing sometihng weird. especially when adjust pin of lm317 getting high something like 7v its responding normal. when its about 10 v and still tring to increase that voltage with pot , after 10.125v the adjust pin's voltage of lm317 is getting contstant 8V and the output voltage is getting stuck at 8v also . i checked the signal with scope and its getting perfectly pure 8v dc voltage. after that when i try to increase voltage it doenst move and when i try to decrease the voltage via pot , after getting down from 8v , its getting low until 1.3v which is normally working area of lm317.
also i must tell that i have also checked to 2n2222's gate pin wiht scope. this is also weird cause ,after voltage is getting higher than 8v , the gate pin of the 2n2222 is acting weird and doing some triangle wave. after output voltage stuck at 8 v the gate pin has something around 600mv pure dc voltage. i have also tried to tie gate to the ground and this time output is getting higher about 10 v and behaving same as before.

i tried to tell my problem as much as i can. hope you guys can help me .

 

[rjenkinsgb]

If you are testing with an actual build of the circuit, I suspect the 2N2222 could be wrongly connected; possibly emitter & collector reversed?
(A reversed base-emitter junction typically has a rather low breakdown voltage).

You may find that you need to add a low value capacitor from ADJ to ground/common for stability as well, otherwise the opamp controlling it may oscillate or be somewhat unstable, especially with load changes.

 

[qwertyqwq]

A reversed base-emitter junction typically has a rather low breakdown voltage

thought that before and now i checked polarity of 2n2222. its looks like true .

You may find that you need to add a low value capacitor from ADJ to ground/common for stability as well, otherwise the opamp controlling it may oscillate or be somewhat unstable, especially with load changes.

thanks. i placed a 100nF cap. but still its acting same.

may that thing happining cause of the lm317's adjust terminal current???? or am i missing something on datasheet ?!?!?!?!?!??
thanks again.

 

[qwertyqwq]

other thing is when it reach through (for example) 10 v of noniverting pin of opamp , the inverting pin is still has 8 v . whitout any feedback of opamp , the output must be positive rail voltage. all that time , im displaying signal on scope but adjust pin of the lm317 is not changing. not even oscillating !!
isnt that weird ?

 

[rjenkinsgb]

There should not be any voltage on the base of the transistor; the fact that it has 0.6V is strange..

I'd try completely removing the transistor temporarily, then see if the voltage control works? It could just be a faulty transistor.

 

[alec_t]

You should have a resistor between the Out and Adj pin of the 317. The IC tries to maintain ~1.3V across that resistor.

 

[rjenkinsgb]

You should have a resistor between the Out and Adj pin of the 317. The IC tries to maintain ~1.3V across that resistor.

That is for using a potential divider to set the ref voltage; with an active control, that should not be needed and could upset the opamp circuit?

 

[Les Jones]

My understanding of the circuit is as follows.
Condition 1 V2 is greater than V3 (So Q1 is not conducting.) The circuit attempts to maintain the regulator output at V1 +1.25 volts. The lowest value the output of the regulator can go is limited by the fact that the output of U2 does not have a negative supply so its output can't go below zero volts (Possibly not even that low. I have not looked at the data sheet on the op amp.) so the adjust terminal current passing through the two 1K resistors will be added to 1.25 volts

Condition 2 V2 is less than V3 (So Q1 is conducting.) The output voltage of the regulator be 1.25 volts plus the voltage across the two 1K resistors due to the adjustment terminal current plus the saturation voltage of Q1

Is my understand the intended way the circuit is designed to function ?

Les.

 

[schmitt trigger]

I've seen many adjustable 317 configurations over the years, but this is a very weird one.

My understanding is that, as Alec_t has already mentioned, that a resistor between Out and Adjust terminals is mandatory. I never seen a *proven* configuration without one. If anybody has seen one that is proven to work, please let me know.

In National's original datasheet, there was a typical application for a 4 amp adjustable regulator, which consisted of a master and two slave LM317s.
The master had the resistor between Out and Adj, and the slaves had their Adj terminals connected to the master Adj terminal. Since all Outs were wired together via very low value resistors, in an indirect way the slaves also had a feedback resistance from Out and Adj.

And the overall feedback was controlled with a PNP and an opamp to adjust the voltage.

 

[rjenkinsgb]

The resistor is only required to set the current through the divider chain, when used with setting resistors.
The reg also needs a minimum load current as the internal circuits use the output as negative power.
(That's covered by the 470 Ohm load on the overall circuit).

An extract from an early data sheet that clarifies the point:

 

[alec_t]

That is for using a potential divider to set the ref voltage; with an active control, that should not be needed and could upset the opamp circuit?

Good point. Simulation confirms the output should be adjustable over the expected range without the Out-Adj resistor, provided the minimum load current is drawn somehow.

 

[qwertyqwq]

There should not be any voltage on the base of the transistor; the fact that it has 0.6V is strange..

I'd try completely removing the transistor temporarily, then see if the voltage control works? It could just be a faulty transistor.

Yeah after i change the npn to another common npn which is bc237 , it started to work. (I ll never buy ic from Aliexpress ).
Now another issue happening. I can change output to 1.3v to 23.5 v. Same issue hapoenin on top voltage. I can see on my meter around 24v , but after some incrementing of adj voltage , adjust pin and output voltage suddenly falling down to constant value of 23.5 !!
I have tried circuit with many output load but it behaving similarly. Of course there are lots of (8 or even 10v ) voltage drop on output with load ( resistor ). At this time its getting constant on 15v or 13v or etc. Sticking voltage is changing via output load. But the same issue happening as mentioned before.

Why?

 

[rjenkinsgb]

Try reducing R3; it should be OK down to eg. 10 Ohms.

 

[qwertyqwq]

Try reducing R3; it should be OK down to eg. 10 Ohms.

when Q1 is closed , adj pin of lm317 will go ground potential. After that noninverting pin of opamp will also go to ground potential, so output will be positive supply rail. Opamp source current will flow through 10 ohms resistor to ground , which is approximately 28V / 10 ohms = 2.8Amps ??
Am i thinking wrong ? Can you please tell me what we expecting after reducing R3 ??
Many thanks.

 

[rjenkinsgb]

OK, I was basing it on the transistor still being out of circuit.

Take R1 down to 10 Ohms & reduce R3 as much as is reasonable; the 324 can stand moderate shorts to ground on its output as it has internal current limiting.

390 Ohms should be OK for R3; it should force the output down by no more than 15V if the transistor is on, which is within the device ratings.

Getting the series resistance from the opamp output to the reg ref terminal as low as practical should improve the control range.

 

[qwertyqwq]

Nop. Nothing changed. Rrrrr getting crazy.
R1 is 10 ohm now. R3 is 470 ohm. Doesnt change anything. Not even a volt !
Anyway. I think im missing a little detail about One of those components. I must read more.
Thanks for all advices. See ya next treat

 

[AnalogKid]

There is no feedback from the 317 output pin to its adjust pin.

ak

 

[crutschow]

There is no feedback from the 317 output pin to its adjust pin.

Yes, he's using the LM317 as a follower in response to the control input voltage, so the LM317 output is 1.25V above the control voltage.

 

[crutschow]

Is goal to have an LM317 with output voltage controlled by a control voltage with a current limit circuit?
If so there are easier ways to do that.

 

[qwertyqwq]

Yes, he's using the LM317 as a follower in response to the control input voltage, so the LM317 output is 1.25V above the control voltage.

The circuit based on youtuber which is eevblog. The name of Lab Power Supply. Jones was making that power supply with lt3080. I desided to build that one. One problem becomes, there is no lt3080 in my country. Because of that i desided to buil with lm317.