Smoked thermistor - inrush current limiting

[yitbos1]

First a notice of warning - I am not very knowledgeable in electronics, so go easy.

I am attempting to use a thermistor to limit the inrush current to 2 Meanwell switching power supplies in parallel (models HLG-240-24A and HLG-120H-12A, specifically). I am powering these from a regular 20A 120VAC circuit. According to the spec sheets I can expect over 100A inrush, so I need some sort of limiting. I attempted to use what I thought was the correctly-sized thermistor, but I only let the magic smoke out.

I used a 10Ω thermistor (120V/20A=8.5Ω) rated to 8A (power supplies total 5.4A steady state input) and able to accommodate roughly 90J/6000uF on the input side (Meanwell said there is only 150uF on the 24V model and 82uF on the 12V one. 170Vpeak & 232uF in the J=½CV² equation = 3.4J). What am I missing?

 

[ronv]

I doubt you need the inrush current limiting because the duration will be so short. But having said that what is the part number of the thermistor?

 

[yitbos1]

I doubt you need the inrush current limiting because the duration will be so short. But having said that what is the part number of the thermistor?

I tried it without the thermistor and tripped the breaker.

The thermistor I used was an Amatherm SL22-10008

 

[aljamri]

I think for inrush current, VDR has to be used, Thermister is used in temperature compansation, or rough temperature measurements.

EDIT:

google "thermister" = https://www.google.com.bh/search?q=Thermister&hl=ar&source=hp&rlz=1R2GGLR_en&aq=f&aqi=g-sL10&aql=&oq=

wiki " VDR " = https://en.wikipedia.org/wiki/Voltage_dependent_resistor

 

[yitbos1]

I doubt you need the inrush current limiting because the duration will be so short. But having said that what is the part number of the thermistor?

I replied with a link to the Newark Electronics page showing the thermistor I used, but since I have < 3 posts, a mod has to approve the post

 

[ronv]

How about a part number & I can look it up?

 

[yitbos1]

I think for inrush current, VDR has to be used, Thermister is used in temperature compansation, or rough temperature measurements.

EDIT:

google "thermister" = https://www.google.com.bh/search?q=Thermister&hl=ar&source=hp&rlz=1R2GGLR_en&aq=f&aqi=g-sL10&aql=&oq=

wiki " VDR " = https://en.wikipedia.org/wiki/Voltage_dependent_resistor

An NTC thermistor is used quite often in my application, according to what I've read. It has resistance at room temperature and at steady state current, resistance drops to < 1Ω due to it heating up. Correct me if I'm wrong, but a varistor protects against voltage spikes, not current spikes.

 

[yitbos1]

How about a part number & I can look it up?

Amatherm SL22-10008. I tried without it and tripped the breaker.

 

[ronv]

Sounds like you have a short somewhere on the input to the module.

 

[yitbos1]

Sounds like you have a short somewhere on the input to the module.

That's what I was wondering. I took a multimeter and read the resistance across the input while the power supplies were still in parallel. I saw a resistance in the 100's of ohms that increased over time (I'm assuming the multimeter was charging the caps itself) and not a short. Is there a better way to test that?

 

[crutschow]

I think for inrush current, VDR has to be used, Thermister is used in temperature compansation, or rough temperature measurements.

It is indeed a negative-temperature-coefficient (NTC) thermistor that is used to limit inrush current.

 

[crutschow]

That's what I was wondering. I took a multimeter and read the resistance across the input while the power supplies were still in parallel. I saw a resistance in the 100's of ohms that increased over time (I'm assuming the multimeter was charging the caps itself) and not a short. Is there a better way to test that?

Try testing them one at a time with no output load. Connect a large incandescent light bulb (say 100W) in series with the input. With no load they should be able to power up with the bulb in series. If one does and one does not, then you can suspect that the one that does not may have a problem.

 

[ronv]

What a great idea.

 

[yitbos1]

It was a great idea! I finally got around to trying it with a heat gun instead of a light bulb (it was more readily available). Unfortunately, BOTH power supplies did not read an output. I must be missing something. I wouldn't expect both power supplies to be defective, especially since they were new. Any ideas?

Here's the spec sheet for one of them: **broken link removed**

EDIT: It is the "A" type.

 

[ronv]

Did the heat gun get hot?
Did you try them one at a time?

 

[yitbos1]

Did the heat gun get hot?
Did you try them one at a time?

Yes and yes.

 

[ronv]

The fact that the heat gun got hot suggests that there is a short. Could it be you have the input and output reversed?

 

[yitbos1]

The fact that the heat gun got hot suggests that there is a short. Could it be you have the input and output reversed?

The inputs and outputs are easily distinguishable from each other, so I didn't reverse them. I would agree that it would seem I have a short, but since both acted the same way, I have to think that I'm missing something.

 

[ronv]

I'm about out of ideas. Do you have any load on it? I wouldn't think the over voltage would trip the breaker but who knows?

 

[yitbos1]

Following up on this issue, I was able to determine what the problem was. I'm almost to embarrassed to say, but the inputs and outputs were in fact switched. No more problems!