inrush current limiter

[kappa_am]

hi all,
I have an inverter in which in faulty condition, it is probable two capacitor with different voltage get parallel. I want to use inrush limiter to limit the current flow in this condition. I looked at net and have found Ametherm limiters(https://www.ametherm.com/inrush-current/as-series-inrush-current-limiters.html). in datasheet of inrush current limiter, they have indicated a resistance value; I like to know this value is related to normal condition or when it act as current limiter. I also will be so grateful if somebody provide me equivalent circuit of this component according to values indicated in datasheet.

Thank you

- - - Updated - - -

gathered that I can not use inrush current limiter (NTC). because it is already heated in circuit when circuit works properly.

any suggestion to over come the issue I have talked about ( paralleled capacitors) would be appreciated.

Thank you

 

[rudr_p]

Cannot we use resistance in series with the inverter capacitor with a switch like igbt/thyristor?

 

[kappa_am]

no! in that way a lot of losses rise. I like a component which has almost zero resistance in normal operation (for example current below its nominal current), and high resistance (not very high just almost 50-200 ohm) when a high current flow through it.
P.S I can not use inductor.

 

[schmitt trigger]

Taking as an example, the Amatherm AS32 0R536:

Resistance: 0.5 Ω ± 25% This is the Thermistor resistance while cold
Max Steady State Current up to 25°c 30 A This is the maximum allowable current your equipment must consume

Resistance @ 100% Max Current: 0.01 ohm Thermistor resistance at 30 amps
Resistance @ 50% Max Current: 0.03 ohm Thermistor resistance at 15 amps

 

[D.A.(Tony)Stewart]

Inrush current Limiters or ICL are NTC and Overcurrent limiters or Polyfuses are PTC. They are small enough to heat up faster than heatsinked parts being protected. Is this what you want?

 

[kappa_am]

I looked at littlefuse site. its PTCs have very low trip current. I need an 100V PTC with Itrip=7 or 8A . I would be grateful if you introduce me a company or part number.

 

[BradtheRad]

I like a component which has almost zero resistance in normal operation (for example current below its nominal current), and high resistance (not very high just almost 50-200 ohm) when a high current flow through it.

This is similar to an incandescent bulb. Its resistance goes up as the filament gets hot.

However its resistance range is not so wide as you describe.

 

[chuckey]

There is a conflict of long term power dissipation and effectiveness. Take for instance the use of a PTC, in the long term is could dissipate, say .5 W BUT if the incoming mains dips, as the PTC is "hot" it will not reduce the inrush current when the mains returns. Another way is to use a resistor in series with the mains which is shorted out by a relay, again 2W (or more) dissipation but it will need a driving circuit. A way I fancy is to use a resistor in series with the earth return of the resevoir capacitor, this reduces the inrush current and use a FET to short it out after a delay. Needs control circuitry AND could have an effect on the ripple output of the PSU but should use less then 10 mW.
Frank

 

[kappa_am]

I need rapid response, In us range. I think, using control system does not provide that.

There is a conflict of long term power dissipation and effectiveness. Take for instance the use of a PTC, in the long term is could dissipate, say .5 W BUT if the incoming mains dips, as the PTC is "hot" it will not reduce the inrush current when the mains returns. Another way is to use a resistor in series with the mains which is shorted out by a relay, again 2W (or more) dissipation but it will need a driving circuit. A way I fancy is to use a resistor in series with the earth return of the resevoir capacitor, this reduces the inrush current and use a FET to short it out after a delay. Needs control circuitry AND could have an effect on the ripple output of the PSU but should use less then 10 mW.
Frank

 

[FvM]

Reviewing your previous posts, I presume the question background is power switch protection in a multi-level converter. Gate drivers with desaturation detection are the usual way to switch off overcurrents. If the problem is about initial capacitor charging, there should be resistor networks to achieve suitable capacitor bias.

 

[kappa_am]

Hi FvM,
how is it going?
right. but I do not want to switch off, I like to equalize two capacitors' voltages, Those get paralleled. I like just limit the current. I can not use series resistor because it cause losses, and adding a control to insert it in faulty condition is not fast enough.
this is not about initializing, just a new design for fault-telorant MLI
any other suggestion would be appreciated. Thank you.

Reviewing your previous posts, I presume the question background is power switch protection in a multi-level converter. Gate drivers with desaturation detection are the usual way to switch off overcurrents. If the problem is about initial capacitor charging, there should be resistor networks to achieve suitable capacitor bias.

 

[FvM]

To balance the voltage between two capacitors, you can either use a circuit involving inductive energy storage, or a certain amount of energy is turned into losses. It essentially doesn't matter if the energy is burned in a resistor or a semiconductor device operating in saturation, the amount of lost energy is the same.

It's a simple calculation involving E = CU²/2 and Q = CU. When paralleling two capacitors without voltage conversion means (inductor), total charge is maintained but part of the energy is lost.

I think, you should first become clear about the intended operation of your inverter and then search for technical solutions.