Current limiter

Go To Last Post
14 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello,

I need to design a current limiter to limit the "inrush current" to a device. The device can consume 30A at 12 V for the first 10-20 seconds if the voltage is kept at 12 V.

I have a MOSFET to control the power to the device. The reason I need the current limiter is that the MOSFET only handles 10A. After the 10-20 seconds, the current will drop down to around 1-2A and therefore I do not want to just get a 30-40A MOSFET.

If I use a resistor to limit the current it will have to burn a lot of power, ~360W and that is not preferable. I want the solution to consume as little power as possible when the limiter is not activated, and not just burn the energy when it is activated.

Are there any standard solutions for this where I won't need a large heatsink?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Are there any standard solutions for this where I won't need a large heatsink?

Why make life so difficult: use a more powerfull fet, and/or parallel a few.

I can't think of a simpler solution than that. A SMPS is way more costly.

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tessa and Tina, You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The reason is also because of safety. The device is closely connected to humans that I do not want to burn things off from.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well unless the device you're powering is going to get hot, using a stronger FET capable of delivering that amount of current isn't going to make the FET that much hotter. (Especially if you choose one with a very low RDSon)

Your other option is to PWM the FET to keep the average current in range you want, but to do this you may still need a stronger FET.

Writing code is like having sex.... make one little mistake, and you're supporting it for life.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Current limiters at low currents are not so hard. At 30A, much harder.

How about adding an inductor in series? Problem is that it would have to be BIG for the time and current levels you are talking about.

If you limit the current to, lets say 5A, will it still only take 10-20 seconds? Or, will turn-on get longer? It almost sounds like a capacitor charging up. If it is anything like that, then limiting the current will just make it slower to turn on. So, at 5A, maybe it will take 40-60 seconds? From 12V, that means 60W for over a minute.

If it were me, I would rethink what you are trying to do!

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Isn't there a trouble with a series inductor that it's hard to turn off, and you will get nasty voltage spikes across the FET?

Sean.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Indeed. And to prevent that, OP needs to add a schottky diode. He is getting close to a SMPS.

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tessa and Tina, You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Agree. At that current level, a switch-mode supply may be simpler (though, certainly not inexpensive). If human safety is at stake here, I would not go cheap.

To use a PWM switch, you would really need to sense the current. That brings in, at the very least, an ADC with a current sense resistor, and maybe an op-amp. Once you have all this, you might as well use a current limited SMPS.

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A solution might be setting a resistor in the negative line,lets say 0.1 ohm.If the voltage across the resistor amplified x10 is the same like you had an 1 ohm resistor.This voltage then applied to the inverted input of a comparator.To the non inverted input apply a reference voltage,when this voltage is higher than the voltage that measured across resistor multiplied by 10,the output of comparator goes high and drives the gate of N channel FET.The reference voltage could be created by an avr using fast PWM and RC filter and can vary by changing the duty cycle in the output compare pin of the avr.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You are powering the device through the FET, if I'm reading this correctly. To me this means that the VCC goes through the FET drain source channel.

Could you not have some sort of gate bias adjustment to keep the FET from turning on fully for the first several seconds. You may still need a larger FET than the one that you have now because it will be operating in the class A region, but it might work.

You could also use a series resistor that you switched out of the circuit at some point. You could do this with 2 FETs and some miscellaneous parts.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

NTC Thermistor perhaps?

oddbudman

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If the FET is not turned on completely, it will have to burn all the extra power. Lets say 5A 8V, thats 40W and quite hard to cool. NTC will work the same way.

The solution is to have the FET on a small part of the time and off the other, fully. And then have an inductor or so to stop the current from starting to get too large.

I am currently trying to find a solution with a step down DC/DC converter and then just set the output voltage at say 2V to start with. That would keep the current down. And then increase the voltage as the current decreases. Or rather, the current stays the same since I higher the voltage.

I just haven't found a good step down regulator that handles 10A, 12V and where the output voltage can be set in an easy way from a uC in the span of around 2 volts up to 12.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A simple solution that sometimes works:- Use a series R to limit the current to a safe maximum and then after a suitable delay time when the current peak has passed, shunt (turn on) the FET which is connected across the Resistor. The pro is the FET doesn't need to take the full current. The cons are 1. the FET will need to have a low Rdson as it will be in series with the supply.
2. The resistor will have to be man enough to handle the full surge current.
If the switch on surge is only going to occur infrequently, then you may be able to reduce the power rating of the resistor accordingly, but if you overdo it, the thing will fry.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What is usually done in these cases where you need a lot of current form a SMPS is to build it with multiple phases. A 3-phases controller will let you use 3x more of the same components you have now for 3x the current.