Constant current constant voltage power supply for low power application

[Adhith]

Hello there,
I'm in search of a precise and small constant current constant voltage power supply as a part of my college project. I'm looking around 2-3 volts and a current of 0.1 to 0.5 milli amps. could I do this with LM317 or something else. If anyone could give me some suggestions on this, it would be very helpful. Thank you.

[srb1954]

Hello there,
I'm in search of a precise and small constant current constant voltage power supply as a part of my college project. I'm looking around 2-3 volts and a current of 0.1 to 0.5 milli amps. could I do this with LM317 or something else. If anyone could give me some suggestions on this, it would be very helpful. Thank you.

An LM317 or LM317L will have too high a minimum current load to be usable down to the 0.1 to 0.5mA range.

If you are looking for a minimal component solution take a look at the LM334 adjustable current source.

[Circlotron]

Try this -> https://artofelectronics.net/bad-circuits/bckt-op-amp_current_source_2/

Edit - that's only constant current. Oops.

[magic]

That doesn't even work for CC, it's a transimpedance amplifier, what are they smoking

But the principle is sound, ignore regulator ICs and coble together an opamp circuit.

[james_s]

0.5mA should be easily supplied directly by most op-amps, you won't even need an output driver for such a low current.

[Jwillis]

How integrated are you allowed ? Are you only allowed jelly bean parts ? Can  it be for linear or SMPS applications ?   
Take a look at this SEA05

[geggi1]

You can build a transistor current source.
Two resisistors, one zener diode and one transistor.
This is cheep simple and simple to modify if you need it for some other project.

https://en.wikipedia.org/wiki/Current_source

[exe]

Try this -> https://artofelectronics.net/bad-circuits/bckt-op-amp_current_source_2/

Edit - that's only constant current. Oops.

Man, look at the url It's from their collection of bad circuits.

As of LM334, I think it's very imprecise. I'd advice not to use it due to poor temperature stability.

I'd try the first circuit from this post: https://www.eevblog.com/forum/projects/is-550uf-too-big-for-a-power-supply-that-has-cc-limit/msg2161846/#msg2161846 . It's a circuit from lt1010 datasheet.

[exe]

Don't be put off by the apparent complexity, it's not hard to analyse. You can use this topology for microamps to hundreds of amps with the right components (and power sources).

What is the advantage of this circuit comparing to a typical mosfet+opamp current sink? Grounded load? It also seems is doesn't provide voltage regulation.

[Terry Bites]

Nah- that's just crap circuit. If you put the load in the loop instead of to ground it will work- badly.

[Adhith]

An LM317 or LM317L will have too high a minimum current load to be usable down to the 0.1 to 0.5mA range.

If you are looking for a minimal component solution take a look at the LM334 adjustable current source.

Thank you for your suggestion. I looked on the web and got to know about it. But I also came to know that it is very sensitive to temperature. How bad it would be in my case, where the voltage would be 2-3 volts and current around 0.1 to 0.5 milli amps

[Adhith]

That doesn't even work for CC, it's a transimpedance amplifier, what are they smoking

But the principle is sound, ignore regulator ICs and coble together an opamp circuit.

okay. I'm planing to go with LM334. I have a wall adapter of 9V which would be a constant voltage supply, correct me if I'm wrong. Then I add a resistor to drop the voltage to my range of 2-3V. Then I just add the constant current circuit in series with that to get a CC/CV power supply. It would work right?

[Adhith]

How integrated are you allowed ? Are you only allowed jelly bean parts ? Can  it be for linear or SMPS applications ?   
Take a look at this SEA05

Thank you very much. We are now just looking for a basic setup with breadboard for the first trial. The power supply is actually used to charge a particular type of cell (metal air cell) along with other electronic components. In the coming months we would be looking into much more efficient circuits. I'll definitely look into your suggestion at that time.

[Adhith]

You can build a transistor current source.
Two resisistors, one zener diode and one transistor.
This is cheep simple and simple to modify if you need it for some other project.

https://en.wikipedia.org/wiki/Current_source

Thank you. I have seen this option while searching over the web. Have seen a circuit which uses two transistors as well. Whats your thought on LM334 Vs the transistor circuit. The pros & cons in using these?

[Adhith]

As of LM334, I think it's very imprecise. I'd advice not to use it due to poor temperature stability.

I'd try the first circuit from this post: https://www.eevblog.com/forum/projects/is-550uf-too-big-for-a-power-supply-that-has-cc-limit/msg2161846/#msg2161846 . It's a circuit from lt1010 datasheet.

I saw that LM334 have poor temp stability. I'm interested to know how much it would fluctuate for 2-3V and 0.1 to 0.5 mill A usage case.
I will look into the circuit that you have mentioned. Looks great. will tryout this one as well. Thank you

[Adhith]

You probably want a classic CCS. Avoid Howland CC circuits- it's so hard to get any real precision out of them. Looking at  clip 1, the current can be set either by varying either Vref or R1. Io =Vref/R1.

You may have seen this circuit without R2, R3 and C1. But the circuit will often oscillate unpredictably without them. Generating Vref can be done with a variety of ways but it is referenced to the positive supply rail posing some nuisance issues if you want to vary it from a ground referenced source like an ADC. So if it's a static set and forget application use a shunt reference voltage ic between positive supply and the non inverting terminal with a bias resistor to ground and tweak R1. If you want to control it from another system, then replace Vref with a resistor and draw a known current through it. Use low power mosfets or you will have off state leakage currents. You can probably make better choices for the op amps and the transistors than I have. My example circuit, creates 100uA/V. The lower the op amp offset, bias and drift, the more accurate and stable it will be. The odd resistor tweak can be used to sort any scaling errors. Don't be put off by the apparent complexity, it's not hard to analyse. You can use this topology for microamps to hundreds of amps with the right components (and power sources).

Now with available with BETTER COMPONENT CHOICES for operation at 5V

Great! Thank you for you detailed explanation. I'll look into this circuit and get back to you.

[Gyro]

As of LM334, I think it's very imprecise. I'd advice not to use it due to poor temperature stability.

I'd try the first circuit from this post: https://www.eevblog.com/forum/projects/is-550uf-too-big-for-a-power-supply-that-has-cc-limit/msg2161846/#msg2161846 . It's a circuit from lt1010 datasheet.

I saw that LM334 have poor temp stability. I'm interested to know how much it would fluctuate for 2-3V and 0.1 to 0.5 mill A usage case.
I will look into the circuit that you have mentioned. Looks great. will tryout this one as well. Thank you

The LM334 has a temperature coefficient of 0.33% / 'C.  The temperature effect is relevant to the amount of power you are dissipating as well as ambient temperature. You can convert it to a zero temperature coefficient current source by the addition of an extra resistor and a diode. See the "Zero Tempco Current Source" application in the datasheet.

It's always important to study the datasheet.

[exe]

Probably, it's possible to do temperature compensation for lm334, but is it worth the trouble? (not to say it won't be perfect due to imperfect thermal coupling between lm334 and the diode). I found lm334 kind of pointless.

If one needs a current source, it could be done with a couple of bjts. Or with a jfet (but that would require trimming the resistor). Or, if dropout is not a problem, one can make a current source from, say, tl431. This part is easy to get. For small currents it's probably possible to arrange that with just tl431 and a few resistors (I'll try to breadboard that). Or, if one wants to go "fancy", just use a standard schematic from google that can be found by searching "tl431 current source". There is tlv431 with lower voltage reference.

Or, simply use "lm317 current source". Just find a three-terminal regulator with low quiescent current to reduce the error.

[Gyro]

Pulling a random 'standard' TL431 circuit off Google seems to be the exact opposite of studying and understanding the datasheet. 

 Looking at the TL431 datasheet, it doesn't appear to be possible to design a current source for less than its 1mA quiescent current - without implementing a separate bias supply anyway. As was already pointed out at the beginning of the thread, the quiescent current of an LM317 buts it right out of the ballpark for the OP's requirements.

The OP is looking for currents in the range of 100uA to 500uA. The LM334 can do a 10k:1 current adjustment range - 1uA to 10mA, and a 40V compliance.

The LM334 isn't pointless - it's just that you either haven't read the datasheet.... or haven't understood the point.


P.S. For the OP's current requirements, and with a reasonably low voltage drop, die heating won't be an issue anyway.

[exe]

Never implied "not reading datasheets". Although, I feel OP is lost and doesn't understand the suggestions. Hope he'll figure it out.

Looking at the TL431 datasheet, it doesn't appear to be possible to design a current source for less than its 1mA quiescent current - without implementing a separate bias supply anyway.

Well, I'm attaching a picture showing sourcing 24uA without a separate bias supply. Compliance voltage is ~1.6 till all the way to maximum allowed voltage (afaik 18V or so). I'll upload the video showing the performance if anyone is interested. Also tested up to 500uA, worked fine, except inputs needs to be around 3V or (but Idk how much was dropping on the DMM, so it might be burden voltage of dmm didn't let it go lower, spent literally 10mins on this, including 5mins of fighting with bad wiring)

As was already pointed out at the beginning of the thread, the quiescent current of an LM317 buts it right out of the ballpark for the OP's requirements.

Yeah, that's why propose cmos regulators with low quiscent currents. Like HT7318 with declared quiscent current of 4uA (typ).

The LM334 isn't pointless - it's just that you either haven't read the datasheet.... or haven't understood the point.

Well, let's not create drama and blame each other in lack of understanding. I'm proposing working solutions of similar or lower complexity that do what the OP asks (*) and without the need for any corrections.

(*) Actually, I don't really understand why people suggesting current sources where OP asked for CV/CC power supply. Where's CV part?

EDIT: yeah, burden voltage of my DMM kicks in at low current range.

[Gyro]

Well, I'm attaching a picture showing sourcing 24uA without a separate bias supply. Compliance voltage is ~1.6 till all the way to maximum allowed voltage (afaik 18V or so). I'll upload the video showing the performance if anyone is interested. Also tested up to 500uA, worked fine, except inputs needs to be around 3V or (but Idk how much was dropping on the DMM, so it might be burden voltage of dmm didn't let it go lower, spent literally 10mins on this, including 5mins of fighting with bad wiring)

Impressive that you got that low, though I'd question whether you were actually in regulation at that point. The datasheet says 'Minimum cathode current for regulation' is 1mA (450uA typ). Off state cathode current is 1uA (50nA typ), so I think you were probably right on the individual device specific bleeding edge of device turn-on.

Well, let's not create drama and blame each other in lack of understanding. I'm proposing working solutions of similar or lower complexity that do what the OP asks (*) and without the need for any corrections.

No argument there, I wasn't intending drama even if the wording wasn't ideal. Pending any sort of accuracy requirement from the OP, die heating would be insignificant in this case. Without that, I'm proposing one active device plus one resistor. 

(*) Actually, I don't really understand why people suggesting current sources where OP asked for CV/CC power supply. Where's CV part?

Yes, I was wondering that too. The OP doesn't make the requirement clear, or required accuracy of voltage or current. If it is current that's important with 2-3V of compliance then we have it covered. If he wants a 2-3V supply which is only capable of sourcing a maximum 100-500uA into a short, then a Vreg with a simple series resistor would meet the brief.

There's no problem with putting one before the other of course. On the one alternative, the current source would need to accommodate the quiescent current of the Vreg - which as you say, can get very low. The other way round, the Vreg output voltage would need to accommodate the voltage drop of the current source, which for an LM334 can be as low as 800mV up to 100uA (0.5V at 1uA).

[Adhith]

Thought to give you a clarification from my end, since you guys made a really relevant point.

Yes, I'm a bit lost here! especially because I'm a beginner in electronics. What I'm looking for is a simple CC/CV module for 2-3 volts and 1-5 milli A of current.
I do like to see the other options out there as well. Howe ever, I'm not able to grasp the technically terms or findings that you are discussing with each other. I'm also learning and reading these but it will take time from my end.

Yes, the discussion is going in the direction of just the constant current part. The input is a 9V DC wall adapter and I'm planning to reduce to a constant voltage of 2-3V using resistor voltage divider. That would work. right?

For now, I would like to try few simple circuits first and gradually progress to the complex ones that you have mentioned, if needed. In that way, I could learn things step by step.

[Adhith]

The LM334 has a temperature coefficient of 0.33% / 'C.  The temperature effect is relevant to the amount of power you are dissipating as well as ambient temperature. You can convert it to a zero temperature coefficient current source by the addition of an extra resistor and a diode. See the "Zero Tempco Current Source" application in the datasheet.

It's always important to study the datasheet.

Great. Thank you!. I'll try the circuit with and without zero tempCO settings and see how things are performing for my application.

[Adhith]

Probably, it's possible to do temperature compensation for lm334, but is it worth the trouble? (not to say it won't be perfect due to imperfect thermal coupling between lm334 and the diode). I found lm334 kind of pointless.

If one needs a current source, it could be done with a couple of bjts. Or with a jfet (but that would require trimming the resistor). Or, if dropout is not a problem, one can make a current source from, say, tl431. This part is easy to get. For small currents it's probably possible to arrange that with just tl431 and a few resistors (I'll try to breadboard that). Or, if one wants to go "fancy", just use a standard schematic from google that can be found by searching "tl431 current source". There is tlv431 with lower voltage reference.

Or, simply use "lm317 current source". Just find a three-terminal regulator with low quiescent current to reduce the error.

I was under the assumption that, lm334 performs a bit more than the CC circuit using BJTs. seems like I'm wrong. I'll try the BJTs as well. Since I'm already purchasing tl431 for another application maybe I look into that as well.

[exe]

Impressive that you got that low, though I'd question whether you were actually in regulation at that point. The datasheet says 'Minimum cathode current for regulation' is 1mA (450uA typ). Off state cathode current is 1uA (50nA typ), so I think you were probably right on the individual device specific bleeding edge of device turn-on.

That's because the cathode current was 1mA+ The load was not grounded. Here is the schematic I used (see attachment). Basically, I made a shunt current source. The load is in series with the shunt, and tlv431 connected in a way that if voltage drop on the shunt is more than 1.24V, it will "steal" the excessive current. The adj-pin bias current is 10uA max, so there is a potential to go even lower. I didn't check stability boundaries, but seemed stable to me under variety of voltages and currents and on different dmm ranges (i.e., different load resistance). At least my dmm didn't show any significant AC current (I used u61e).

I also checked ht7333-a connect as a current source (claimed to have 5uA typ quiscent current). Seemed to work well too, without any capacitors (a bit of grey area here, as, afaik, datasheet doesn't discuss stability). Checked down to 65uA using 50k trimpot.

My next candidate is LM317L which has quiscent current of 100uA max. This IC, afaik, quite stable, even without external capacitors thanks to darlington emitter folower architecture. I'll check and report it here. I used this little device in the past and I was impressed with its performance. It seemd to me much better than lm317, at least reference voltage so much more stable. I evaluated devices from TI, on semi, and, afaik, ST, all showed good performance. I'll also evaluate min dropout voltage, as it seems the major drowback with lm317* devices. That's because of darlington output stage (with, afaik, three transistors) plus drop on current sense resistor.