Stop right there.
1) If the unit is powered by a VFD then the AC waveform is rectified and a new AC (3 phase) waveform is synthesized.
This makes it entirely irrelevant what happens -- phase wise -- when the unit transitions from one primary loop to another.
2) This is obviously a fully engineered high powered electrical consumer -- being fed by a double-headed primary Service.
3) I've never seen (disco/ safety switch) fusing placed AFTER a VFD. Such a placement is a sure fire way to destroy the VFD!!!
EVERY time a fuse blows the electronics should be fried. VFDs DO NOT LIKE TO BE shut down without a taper. If such a situation occurs, some reactor/ inductive tank is required to stop "wave form recoil." (for lack of a better term)
I do not intend to get into a long, deep, involved physics lecture on electromagnetic phenomena. Suffice to say that VFDs emit extremely high frequency wave edges on their way to the creation of synthetic alternating current. Because of this, this current has to be smoothed out via inductors/ chokes/ coils/ reactors so that the exterior circuitry is not blown out by insanely high voltage peaks.
At the frequencies involved, these pulses can BOUNCE BACK from the dead end of ordinary conductors -- right back into polarized electronic elements that CAN'T take the reversed polarities.
You've got extensive damage in your VFD as a DIRECT CONSEQUENCE of the layout.
It will REPEAT until this goofball scheme is corrected.
There can never be a situation where the VFD is forced to juice an instant open circuit.
In sum, you've got a ticking bomb on your hands.
You need to study up on VFD practices. It sure sounds like your whole facility was installed incorrectly -- by those totally unaware that VFDs can't (normally) survive load-side blown fuses. (VFD fusing has to be up the line.)
The zany notion that it's your SCRs that are dead is cute.
I'd bet your last dollar that the logical control electronics are toast. They are tiny by comparison. Without them, your SCRs don't fire off. This makes the SCRs look like they're what's fried. Actually, SCRs are inherently robust and can take far more punishment than the control chips.
One last note: VFDs produce what the fuses would regard as circulating current. That is, they suffer the ill effects of the voltage peaks just like old time electrical motors.
Now that the high peak problem is understood, the NEMA crowd has engineered an entire suite of motor designs that is able to tolerate VFDs.
The synthetic waveform has many of the characteristics of a strongly lagging or leading motor. That is it needs more reactive power than real power. This reactive current juices the fuses ABOVE the level of real load. It's enough to cause fuses to blow at loads below the motor protection design threshhold.
Last edited by Tesla; 05/05/14 09:29 PM.
