Power supply current limiting

[AxleD]

Hi,

I have put together a simple regulated power supply consisting of a current regulator (one LM317 with pot in series with the output of the 317) and a voltage regulator (another LM317 in the standard configuration). A few silly questions but I unfortunately don't have a power supply (and have never used one) so i need to confirm:

1. Is it normal for the output voltage of the current regulator to decline as i dial the current down?
2. I presume that as soon as the voltage starts declining I am then effectively in current limiting mode?

Thanks

[T3sl4co1l]

The current limiter should be upstream, not downstream.  The one on the load side always takes priority; so you've made, more properly speaking: a voltage limited current source.

Tim

[AxleD]

Sorry absolute noob here. What does upstream refer too?
 

[T3sl4co1l]

In this case, on the rectifier-filter side, not the load side.

Tim

[AxleD]

Hi Tim,

Thanks.
I have included the setup i have in the attached (straight from the datasheet). I believe the current limiter is on the rectifier side.

When dialing down the current, the output voltage of the first 317 falls, causing the output voltage of the second 317 to fall. Not sure if that's how its supposed to work.

[Zero999]

Yes, that's what you should expect to happen.

The circuit can only regulate either the current or voltage due to Ohm's law. It can't regulate both current and voltage simultaneously.

[T3sl4co1l]

Most importantly, it should do it in a relatively sharp manner.  If you use a large number of different load resistances, and graph the output (V, I), you should find a square, or something near that.  This method won't give a perfectly square corner (where it goes straight from voltage to current limit), but it's pretty good for most purposes.

Tim

[Zero999]

A big disadvantage of this design is the current limiter has a large drop-out voltage: 1.25V, plus voltage drop of the LM317 at full load, then there's the drop-out voltage of the second 1M317.

[AxleD]

Thank you all. That's exactly what i was looking for.
With that working i am allot closer to finishing off this project. From my limited knowledge, the design does seem inefficient but it will have to do for now.

If i can be so bold and ask another related question...
I was thinking of adding a current limit led that lights up whenever the supply is in current limit mode.

I was thinking of using two sense resistors - 1 after the current limiter 317 and one after the load and comparing the voltage drop across both
The attachment show's the circuit
Will this work? Any ideas on how best this could be achieved?

[mikerj]

I was thinking of using two sense resistors - 1 after the current limiter 317 and one after the load and comparing the voltage drop across both
The attachment show's the circuit
Will this work? Any ideas on how best this could be achieved?

No it won't work. Remember the current in a series connected circuit is the same through every element. The current coming out of the second LM317 will be equal the the current going into it, plus the quiescent current of the LM317 itself.

You could get a crude indication just by looking at the difference between the input and output voltage of the first LM317.  Under normal conditions, this voltage drop will just be the normal dropout voltage, but when it's current limiting the output voltage will quickly drop.  However, it will be difficult to get anything like an accurate indication because the dropout voltage is not constant, it increases with current, there is a temperature dependance and there will likely be small differences between parts.

[Zero999]

There's no need for two sense resistors. Looking at the voltage across Iset and turning on the LED when it exceeds 1.2V would work.

[Rupunzell]

Constant current or constant voltage. Most bench power supplies work this way by a switch over between current control or voltage control (regulation).

It is possible to make a limited current voltage regulator or a limited voltage current regulator, these are often specialized for a specific application.

Suggest have a look at the service manuals for a typical bench supply to see how they work before proceeding. Other examples can be found in data sheets like the uA723 or LM723. These older IC regulators have more user driven configurations, but the user needs to know something about how power supplies in general work.


Three terminal regulators appear easy to apply for a given requirement until the requirements exceed their original intent, then it can get messy real fast.


Bernice

[AxleD]

Thanks!

I will check the voltage drop across the first LM317 as suggested.

Appreciate all the help and the suggestions regarding further reading.

Cheers