Contactor Failure

[methods]

I will write it up once root cause for the entire system is determined.



This was during a bench test shake down.
Parts pedigree
Procedure and pre-flight checkout

We are working on basic precheck
One test is measuring contactor resistance
The contactor is to be instrumented such that during a run-up the contacts can be read like a shunt. GO or NO GO.

We can do a simple Red Light, Green Light for this. Visual.
It can be as complicated as an interlock... but... keep it simple yes?
The complicated version measures contactor resistance and latches the contactor open

Meh... Visual only.
No electronics which are not needed.

Contactor contacts must be measured under load to present.

-methods

 

[bigmoose]

That thing arc under load disconnect? ... then weld under subsequent turn on?

 

[methods]

Investigating.

 

[liveforphysics]

Single non-complete pre-charge make event welds them pretty reliably.

Or like Dave mentioned, any operation that damages the surfaces can lead to welding once it's taken current on damaged contact pads.

 

[methods]

It was strapped with a 2K 25W Precharge resistor - permanent strap no button.

The system was recently reconfigured tho... Perhaps the contactor coil was inadvertently powered during the cable swap, the precharge resistor failed, or some other issue revolving around reconfig.


-methods

 

[fechter]

Hard to see in the picture. Are the contacts damaged? The coil looks toasted. Measuring the voltage across the contacts is a good way to verify contact. If the controller was drawing current, then the precharge resistor would not get the caps up to full charge.

I've been real impressed with the ability of FETs to handle large current spikes, like my battery spot welder. A solid state contactor should be able to just slam on fast without blowing anything up. I've been looking into that as an alternative to the soft-start, constant dv/dt precharge circuit I designed before. I also have a precharge circuit that drives a relay coil once the precharge is completed.

 

[methods]

One of our primary suspicions is a heavy load on the Kelly leading to the precharge not being able to keep up. Major change was Battery voltage from 80V to 100V... Which may have been enough to weld... Then heat.

-methods