Kilovac Contactor running too hot?

[hummmingbear]

I have a Kilovac 48V Contactor/Relay LEV200A6NAF that my BMS controls to allow or cut power from my battery system. It takes a constant ~48V to keep the connection (closed circuit), and when the leads receive no power from the BMS the circuit opens and it cuts power.

I recently set it up and connected it into my system. I noticed today it is very hot. Just about too hot to hold my hand on it. My guess is it's about 50C right now. Is this normal for these devices? My ambient temperature has been around 25-35C during the day lately. I am feeding it ~53VDC which is well within the nominal voltage it accepts.

I have shut it off completely until I understand what is normal. Any help appreciated.

 

[Maitake]

That's going to be 15-20 watts in the coil. That unit has no economizer.
I would trade it for one with a coil economizer, or find a way to use the aux contacts to lower the coil current.

 

[hummmingbear]

That's going to be 15-20 watts in the coil. That unit has no economizer.
I would trade it for one with a coil economizer, or find a way to use the aux contacts to lower the coil current.

Damn, I remember learning about economizers for these devices and just figured mine had one...for the price it should.

Off-hand do you know which Contactors have economizers? Or is it possible to build my own? I'm not sure how economizers work under the hood.

 

[efficientPV]

Relays typically pull in at 80% of rated voltage and drop out at 20%. Full voltage is only needed to have enough magnetic strength to get the core to pull in. Once that gap is eliminated it takes far less energy. This is a quick and dirty way. Find the resistance of the coil. Put a resistor in series with the coil which is that resistance or slightly higher. Place a 2,000uF electrolytic capacitor in parallel with that resistor, observe polarity. The capacitor looks like a short long enough for the relay to pull in. Relay sees half voltage and half current = 1/4 heat.

 

[hummmingbear]

Relays typically pull in at 80% of rated voltage and drop out at 20%. Full voltage is only needed to have enough magnetic strength to get the core to pull in. Once that gap is eliminated it takes far less energy. This is a quick and dirty way. Find the resistance of the coil. Put a resistor in series with the coil which is that resistance or slightly higher. Place a 2,000uF electrolytic capacitor in parallel with that resistor, observe polarity. The capacitor looks like a short long enough for the relay to pull in. Relay sees half voltage and half current = 1/4 heat.

Thank you! That I can do.

Question with that setup: Currently it's pulling about 13W. If I add the resistor and capacitor as you described, wouldn't that same amount of power be used up in the resistor now instead of the contactor?

 

[efficientPV]

You wanted low tech. In steady state the relay would dissipate the same power as the resistor. Total power of both will be reduced in half. 1/4 of original heat will be in the relay and 1/4 in the resistor. An economizer uses PWM to pulse the relay to reduce the power. I always PWM relays in my microprocessor circuits.

 

[hummmingbear]

You wanted low tech. In steady state the relay would dissipate the same power as the resistor. Total power of both will be reduced in half. 1/4 of original heat will be in the relay and 1/4 in the resistor. An economizer uses PWM to pulse the relay to reduce the power. I always PWM relays in my microprocessor circuits.

Ah! By far that was no critique, and sorry if it came across as one. I just wanted to make sure I understood how it was working.

I actually have PWM boards I can control, would that be a more efficient way of managing it? Would I just cycle the PWM until the relay opens and know that just above that is my "resting" rate once it's closed?

 

[efficientPV]

It could also be done with a cheap buck converter board like the LM2596. Pin 3 is ground connect - of 100uF capacitor to that and + to pin 4, the feedback pin. That would make it think output was too low for a short period of time and send out full power. Then decay down to 10-12V the board is set to.

If done with onboard PWM, the relay would need a high speed diode across it and PWM rate would need to be about 200Hz with a duty cycle of 40-50%.

You want to make sure the contacts have enough pressure on them so don't cut it too close. Current also drops as coil heats up and has higher resistance.

 

[hummmingbear]

the relay would need a high speed diode across it

Everything makes sense for me here except this. Can you expand on what the high-speed diode does? Does it literally just need to be connected to the + & - terminals of the contactor relay?

 

[efficientPV]

Yes it would connect there. All diodes are generally fast turn on. Diodes have a slower turn off speed. If a diode has not finished conducting, it will conduct when power is then applied to it, acting as a short for a short period of time. I don't know what your PWM speed is or diode you intend to use. If just a couple hundred Hz you could likely get by with anything. At higher speeds a common SR5100 or SB5100 is suitable. I went looking and the data sheets ffor several diodes and they wouldn't even mention recovery time. Everyone just knows the classification of diode for the application.

 

[hummmingbear]

At higher speeds a common SR5100 or SB5100 is suitable.

Would this also be suitable for lower speeds as well (i.e. safe to use in any case) or no? Sorry, this part of diodes exceed my electrical engineering knowledge so I appreciate your help.

I'm using an ESP to control a PWM chip, AFAIK I should have no problem setting it to 200Hz with a 40-50% duty cycle

 

[efficientPV]

good for all speeds. On my NANO I would just count down the PWM value each loop till it reached the desired value. Good to see more people using these micros. That would make a nice water heater control.