Maker Pro
Maker Pro

How to Size a Current Inrush Limiter

Gristle McThornbody

Mar 14, 2012
31
Joined
Mar 14, 2012
Messages
31
I've asked a few questions and gotten some great help with a small linear power supply that I am building, and have one more question for the group. Because this is a moderately high current (10 amp) supply and because I need low ripple numbers, I will be using a fair amount of filter capacitance. I'm wondering if it might be a good idea to use a thermistor to limit the inrush current a bit to lower the stress on the transformer, rectifier and the caps themselves.

Is this worth doing?
If so, how do I size it?
Is there a site that would be helpful?

FWIW, this will be a 24 volt, 10 amp supply operating from standard 110 vac.
 

gorgon

Jun 6, 2011
603
Joined
Jun 6, 2011
Messages
603
To size the inrush limiter is easy. Set the max current you want at inrush, and use Ohm's law to compute the resistance needed at startup. Remember that the peak voltage is 155-160V for a 110Vac. The max current will be at peak voltage.

The problem is to get rid of this limiter when running the power in normal mode, since it will have a loss in itself if not removed.

If you make a solid state switch that use a zero voltage crossover start, the voltage will increase from zero and you'll get the minimum startup current possible.

A more advanced power of this size will have its own startup ramp of its output voltage, automatically reducing the inrush current over some ms.

TOK ;)
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,681
Joined
Nov 17, 2011
Messages
13,681
A thermistor is a possible solution. In fact: it is often used. The drawback is that it works only once - at least if you look at short time frames.
Switch on for the first time. Thermistor's resistance is high, inrush current is low. Thermistor heats up, allowing continuous flow of nominal current.
Switch off and on again after a short while (e.g. power outage, power supply unused for a few seconds just to rewire the output etc.). Now the thermistor is still hot and has low resistance. It will not limit the inrush current. More: http://www.epcos.com/web/generator/...operty=Data__en.pdf;/PDF_Applicationnotes.pdf

Here is a treatise on active inrush current limiting: http://www.bonavolta.ch/hobby/files/MotorolaAN1542.pdf


Harald
 

Gristle McThornbody

Mar 14, 2012
31
Joined
Mar 14, 2012
Messages
31
Because I plan to keep this simple, a thermistor seems a good way to go. If the thermistor fails to work for any reason, I will not destroy anything: I'm using this merely to reduce stress on components.

Back to my question, how do you spec out a thermistor? From an earlier response, do I just look at a cold resistance to set the max inrush AC current and use that to determine the max dc current hittting the bridge and filter caps? Can I assume that when the thermistor heats up, it will be at a resistance low enough not to eat up too much power?

Or is there any point in trying to provide this protection in the first place?
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,681
Joined
Nov 17, 2011
Messages
13,681
You have to look at the cold resistance as well as at the hot resistance and the power it takes to heat the NTC. It is all explained in detail in the EPCOS app-note I linked.

And yes, there is a point in limiting inrush current. It not only lessens stress on components but also on the mains wiring/fuses.

Another idea: I assume your capacitors are charged from a mains tzransformer and rectifier circuit. Instead of limiting the current on the secondary side (low voltage), you could place the NTC on the primary side (mains voltage). With this setup you will not only limit the inrush curren to the capacitors, but also the inrush current of the transformer. The latter being a relevant part of the total inrush current, limiting it will protect the whole setup even better.



Harald
 
Last edited:

Gristle McThornbody

Mar 14, 2012
31
Joined
Mar 14, 2012
Messages
31
Okay, I'm still totally lost on how to spec out a thermistor for this application. The supply I'm building operates on 115 vac and will produce about 8 to 10 amps at 24 vdc. I'll use capacitors for filtering, probably around 20 to 30,000 mf. (I want low ripple.)

Based on that, the steady state supply current (primary side of the transformer) at max load should be around 2 amps. This power supply will normally operate between 20 to 50 percent of full load.

So if I choose to limit the inrush (primary) current to 3 or 4 amps, this should leave plenty of room to operate and reduce stress on the various components. Unfortunately, I have no idea how to proceed from here. A little help, please?
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,681
Joined
Nov 17, 2011
Messages
13,681
Find an NTC with a current rating that is at least as high as the nominal current through the thermistor during operation. From the datasheet(s) find the nominal resistance and the fit parameters k,n. Use these parameters and the application guide I linked previously to arrive at the right thermistor.

Example:
Use the datasheet for S364
Take R25=10 Ohm. That would be the initial resistance for limiting the inrush current. You have to evaluate for your application whether this reduces the current enough. For sake of the example let me use this resistor.
From the datasheet take
k=1.55
n=-1.37
From the application note, chapter 1.6, take
Rntc = k*I^n where I=2A nominal current of your application. From this follows:
Rntc=0.6 Ohm
Which generates a voltage drop of 1.2 V during operation.
Using this procedure you can find other NTCs which may fit your application better (higher R25, less Rntc,.. whatever you want to optimize).

Data sheets and also a selection guide are e.g. here: http://www.epcos.com/web/generator/.../NTCInrushCurrentLimiters/Page,locale=en.html for download.

Harald
 
Top