SMPS Bench Supply -- Current Limiting / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

SMPS Bench Supply -- Current Limiting

Move Thread

LAN_403

Ruben

Thu Nov 18 2010, 03:56AM

Registered Member #3263 Joined: Sat Oct 02 2010, 04:43AM
Location: Sydney, Australia
Posts: 34

Hi All-- I've been lurking for a while, but classes have finished and i can finally start some projects!

I've picked up three 4-20v 40A SMPS (Meanwell,

) as surplus for an absolute steal. I plan on using two in series for a high current bench supply, 8-40v, at up to 40A. (the third is backup for if something dies).

The output voltage is controllable from 20-100% via a 1-5v signal, and the supplies also have a remote sense function, although I hear this is only good for say +- 5% adjustment. The datasheet seems to imply hiccup current limiting at around 45A. The supplies are well used and I don't want to tax them too much, also 40A+ is overkill for many applications -- I'd like to implement some form of adjustable current limiting.

Ideally this would limit the output voltage to provide constant current to the overload, before tripping a relay if still in overload when at minimum voltage (Control signal = 1v). I've got a few ideas for this kicking around, most based around a shunt resistor, and feeding back a signal to the voltage control input:

1) [Simplest] Comparator circuit to just trip a relay as soon as an overload occurs,

2) Microcontroller running an appropriate difference equation to (hopefully) provide constant current limiting. This implementation could also control a relay for soft-starting the supplies (60A inrush, each)

3) Some clever arrangements of op-amps to provide an appropriate feedback signal.

4) Modifying the supplies themselves.

I'm leaning towards 2) as thats where my experience lies. My first port of call would be figuring out what the step response to dropping the control voltage looks like, whilst under a reasonable load. I don't have the knowledge to tackle option 3 or 4, but have a hunch they could be simple and elegant solutions.

Any suggestions, comments or guidance? Anything else i need to consider here?

hboy007

Thu Nov 18 2010, 11:01AM

Registered Member #1667 Joined: Sat Aug 30 2008, 09:57PM
Location:
Posts: 374

I would defininetly recommend getting a SMPS that has a suitable voltage range, then modify the feedback loop. I have used several MeanWell supplies and hiccup limiting mode is not always implemented. Some, e.g. the 24V 30W hat rail mountable one switch to constant current mode and gradually drop to as low as 6-7V if the load exceeds the current limit.

When putting the supplies in Series, don't forget to add 30A Schottky diodes (ps. they come in two in a TO-247 package) across the outputs to prevent voltage inversion for the event of one SMPS deciding to hiccup if you really want the 40V.

modifying the feedback loops would imply simultaneously derating the voltage output for both units to get current limiting. I can't tell how well current the regulation per unit will perform but I expect some trouble with the center point shifting or oscillating.

Oh well and maybe you can synchronize the oscillators. Noise caused by asynchronous switch-mode converters makes feedback loop implementation not-so-fun.

Electra

Thu Nov 18 2010, 11:40PM

Registered Member #816 Joined: Sun Jun 03 2007, 07:29PM
Location:
Posts: 156

If it were one single supply, Iâ€™d be inclined to try and find the resistor that setâ€™s the current limit and connect an external rotary switch with several resistors, to select different values of current limit. If the manufactures wonâ€™t let you have a copy of the circuit diagram, this means hours of trying to trace out parts of the circuit to find out what to alter.

As you say you want to run two supplyâ€™s in series I doubt this is a good idea since one supply is bound to start to limit before the other.

Some simple electronic fuse type circuit, as you say a comparator and a latch to disconnect the load or shutdown both supplyâ€™s. Is easy to make variable, but of course has to be manually reset each time it trips out. The only other disadvantage is that it can be a nuisance if the equipment youâ€™re powering has large electrolytic caps in it. Which cause a surge current when switched on, and so triggering the current limit.

If you just need something to stop your latest project going up in sparks and flames, good old fashioned mechanical circuit breaker, Iâ€™ve even used the house type when powering from lead acid batteries.

Steve Conner

Fri Nov 19 2010, 11:37AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I've modified SMPS before to have adjustable voltage and current like a bench power supply. For the current limit I used a 0.01 ohm sense resistor, a precision op-amp to amplify it and drive a meter, and a second op-amp as an error amplifier to compare the measured current with the current limit knob setting. I connected the output of this op-amp via a diode, to the TL494's feedback pin, where it combined with the other error amps.

I had to slug this feedback loop down a lot for stability. So when I short this power supply I get a quick burst of the supply's existing 8 amp limit, then it falls to the limit set by the new feedback loop. But in practice, it works just fine.

These power supplies are somewhat different though, they have the "PAR" terminals for parallel operation. As far as I know, the PAR terminal gives out a voltage proportional to the supply's output current, but it's bidirectional. If you clamp the voltage on it and stop it rising, then that also limits the output current. That is how the current sharing works: connecting the PAR terminals on several supplies together forces their output currents to be equal.

So maybe a current limiter could be as simple as a shunt regulator between PAR and ground. And maybe you can connect a voltmeter to the PAR terminal too, to read the output current, though it might not be all that accurate.

Ruben

Tue Nov 23 2010, 03:30AM

Registered Member #3263 Joined: Sat Oct 02 2010, 04:43AM
Location: Sydney, Australia
Posts: 34

Thanks for the great suggestions guys;

I've emailed Meanwell asking for a schematic, but I dont like my chances -- inside the supplies are four IC's:

the lower two are a LM324 and LM339 respectively, however the names of the upper two have been sanded off (from all three supplies) -- So I dont really know what Im dealing with (are other things used besides the TL494?)

At any rate I'm exited by the simpicity of using the parallel function, I'll be trying that first!

Failing this, do you think it'd be possible to achieve current regulation by feeding back a signal derived from a sense resistor to the voltage control input?

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob