Amertherm Inrush Current Limiter

Status
Not open for further replies.

FaradayFF

Senior Member
Location
California
Greetings,

Have any of you guys used inrush current limiter in LED lighting circuits? Where did you install the inrush current limiter?

Thanks,
EE
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
190203-1355 EST

FaradayFF:

No. But it can be a good idea.

I assume you are talking about a Negative Temperature Coefficient thermister.

I think your question should be more directed at --- how do I select a thermister for inrush limiting?

To answer your question as asked you simply put the limiter in series with the load.

A thermister, whether positive or negative, is a resistor where resistance is very sensitive to temperature.

See https://www.ametherm.com/blog/inrus...ush-current-limiter-the-single-part-solution/ about midway down is a resistance symbol and a useful discussion.

https://www.ametherm.com/inrush-current/selecting-an-inrush-current-limiter.html

A data sheet https://www.ametherm.com/inrush-current/inrush-current-limiters-full-line

The resistance at 20 C will give you an idea of what could be peak inrush current. The SSI (Steady State Current) is steady state load current after inrush, and this current puts the thermister at its designed full load operating temperature. Besides the steady state current the energy passing thru the thermister during inrush determines the size, diameter and thickness (thermal mass), of the thermister.

You could use one thermister per LED bulb, or you could parallel many LEDs, and use one thermister for that group of parallel LEDs. These would be two different thermisters in ratings.

Study the Amertherm literature.

If you had 12 10 W LEDs as a load, then that is about 1 A steady state load.
See
https://www.ametherm.com/datasheetspdf/SL0812101.pdf
to access I have to manually copy this to my browser.
This dissipates about 0.8 W at 1 A. Initial 20 C resistance is 120 ohms. Thus, maximum inrush current in a 120 V circuit might be around 1 A. I would want to run an actual experiment to see what happens with turn at 90 deg.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
190294-2219 EST

FaradayFF:

I found a CL-70 NTC thermister that I have used in a power supply for many years. The original company has been bought out by Amphenol. So a search by CL-70 NTC Amphenol should get you basic data. Their website is very poor.

Basic data is at https://www.amphenol-sensors.com/en/thermometrics/current-limiters/858-cl-series
This is listed as 16 ohms at 25 C and rated for steady state load of 4 A.

With a CL-70 in series with a 100 W tungsten bulb and 123 V I could get a peak inrush of about 5 A and steady state about 1 A. RMS of 1 A peak is 0.7 A. Compare this to 17 A I have measured on a 100 W bulb years ago.

Total resistance of the bulb plus CL-70 and current shunt is about 10 + 17 + 1 = 28 ohms. I used a 1 ohm current measuring resistor. Calculated peak possible current is 174/28 = 6.2 A. So I got fairly close to turning on at 90 deg. The bulb alone would have been about 17 A.

If I had a more appropriate current limiter, for example a CL-90, then initial peak would have been about 123 / (120 + 10 + 1) = 0.94 A.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
190204-2401 EST

FaradayFF:

I need to point out that an NTC thermister is not a panacea for LED bulb inrush.

These devices are thermal in nature and have thermal time constants. For something like a CL-70 it may take a minute for it to get back to close to its room temperature value.

I think it could be quite advantageous for increasing the number of LEDs on a single snap switch. But on a dimmer possibly only a small increase in the number of LEDs.

The problem with dimmers is the short time constant for overload of the switching device, SCR or whatever. In a power outage there may be very short duration losses of voltage. These can be short enough that the NTC doesn't change to high resistance before voltage reappears, and high inrush occurs.

Note: a snap switch does not open on a power outage. It may see a high inrush current on return of power, but not while switching.

A better solution is a series resistor and a time delay relay. This can provide instantaneous recovery to its high impedance state. But this is more complex, more costly, bigger, and probably a higher probability of failure.

.
 

FaradayFF

Senior Member
Location
California
190204-2401 EST

FaradayFF:

I need to point out that an NTC thermister is not a panacea for LED bulb inrush.

These devices are thermal in nature and have thermal time constants. For something like a CL-70 it may take a minute for it to get back to close to its room temperature value.

I think it could be quite advantageous for increasing the number of LEDs on a single snap switch. But on a dimmer possibly only a small increase in the number of LEDs.

The problem with dimmers is the short time constant for overload of the switching device, SCR or whatever. In a power outage there may be very short duration losses of voltage. These can be short enough that the NTC doesn't change to high resistance before voltage reappears, and high inrush occurs.

Note: a snap switch does not open on a power outage. It may see a high inrush current on return of power, but not while switching.

A better solution is a series resistor and a time delay relay. This can provide instantaneous recovery to its high impedance state. But this is more complex, more costly, bigger, and probably a higher probability of failure.

.

Thank you for the insightful response! I appreciate your explanation. Good thing I thought about the inrush. The typical data the vendors publish (steady state current) may be a bit misleading.

Best Regards,
EE
 
Status
Not open for further replies.
Top