Lathe VFD causing over current in AC unit

[Besoeker3]

OK, clearly we have different reference frames for "reduce the deceleration rate". In your experience it's clear that 3-5 seconds would constitute a significant reduction in deceleration. OP didn't mention the braking time, which might be important information.

Unless I missed it, he didn't answer the question if he could reduce the decel rate.
Anyway, a 'scope trace of what's happening to the line voltage would be helpful. As I mentioned before the only electrical connection between the lathe and the AC is the supply.

 

[StarCat]

Carrier Tech Support

Carrier Tech Support

Did you ask their A/C service company to look at it and possibly contact Carrier? Seems to me that would be the first thing to do before calling an electrician.

I can't believe this is the only installation of that unit that has this problem. Most systems use VFDs these days and I have never heard of anything like this.

-Hal

I would for sure get Carrier Tech Support involved in the matter and point them to this thread.

 

[Jraef]

Again, unless the lathe drives are overly sophisticated, nothing having to do with the Decel rates is going to affect the incoming line.

I'm now thinking again about the braking issue being electro-mechanical brakes, not VFD dynamic braking. EM brakes typically create a heckofa voltage spike when they release (inductive kickback), which is how they work; energize to release, de-energize to brake. When a VFD such as the Carrier Unit is running and there is a large ringing voltage transient, meaning a portion of the transient drops BELOW the Forward Conduction Voltage threshold of the bridge diodes, the rectifier bridge will stop conducting into the DC bus for a few cycles, yet the transistors are still feeding energy to the motor, pulling it out of the bus caps. This then can cause a current spike as the capacitors recharge themselves when the transient is over, but because the VFD is ALREADY running, the pre-charge resistor is not in the circuit to prevent that current spike. So the VFD detects it and shuts down. The answer to that is to put a line reactor ahead of the carrier VFD, slow down the rise time of that transient, stretch it out across multiple cycles.

 

[Besoeker3]

So the VFD detects it and shuts down.

I thought it was the AC that trips rather than the lathe circuit.

 

[Jraef]

I thought it was the AC that trips rather than the lathe circuit.

That's what I said, albeit that I referred to it as the "Carrier Unit" (Carrier is the AC mfr).

Transient caused BY the EM brake on the lathe is getting to the carrier VFD (assuming it has no line reactor) via the common incoming line connection and causing it to shut down.

 

[Besoeker3]

Transient caused BY the EM brake on the lathe is getting to the carrier VFD (assuming it has no line reactor) via the common incoming line connection and causing it to shut down.

Do we know that the lathe brake is electro mechanical?
Most, if not all, EM brakes I've come across have been DC operated. These could, if not properly suppressed, cause a spike on de-energisation rather than on application.

 

[synchro]

Most, if not all, EM brakes I've come across have been DC operated. These could, if not properly suppressed, cause a spike on de-energisation rather than on application.

And thereby generate conducted and possibly radiated EMI like an ancient spark transmitter.

It might be possible that the controller part of the AC unit is being interfered with and that is causing it to make errors in controlling its VFD (perhaps causing excess current drain and subsequent shut-down). This mechanism might take a smaller level of interference than it would to directly effect the higher power electronics inside the VFD.

 

[kwired]

And just to be perfectly sure, have you independently confirmed (instead of just being told) that it is an overcurrent fault and not an overvoltage fault?

Seems more likely to be overvoltage doesn't it?

Slower machine = lower productivity.

This is happening when decelerating at shut down. Productivity is intended to stop when shutting down.

 

[gadfly56]

Seems more likely to be overvoltage doesn't it?

This is happening when decelerating at shut down. Productivity is intended to stop when shutting down.

Or it could be for a tool change. The OP referred to a "cycle", not necessarily a complete tool shutdown.

 

[Besoeker3]

Seems more likely to be overvoltage doesn't it?

We need that waveform trace.

 

[S'mise]

The answer to that is to put a line reactor ahead of the carrier VFD, slow down the rise time of that transient, stretch it out across multiple cycles.

I agree.

 

[Hv&Lv]

Something to think about..
call your utility and see if they will come out and put an analyzer on the line. You say the only common connection is supply lines, it should show up.
I have a Dranetz HDPQ that I install for our problems, and customer problems. Even put it on my house for kicks. Using it now to measure angles between two different generation feeders.
Takes 15 minutes to set up and install, take it out after two or three days, you get a full report of what’s going on.

 

[Jraef]

Well, the OP posted his question, then his last activity was 6 minutes later. In other words he never came back to even read the responses... So he either found the problem and solved it, or he was a troll. Either way I think it's a dead issue now.

 

[kwired]

Well, the OP posted his question, then his last activity was 6 minutes later. In other words he never came back to even read the responses... So he either found the problem and solved it, or he was a troll. Either way I think it's a dead issue now.

You saying as moderator you can tell if a member read a post even if they never made any reply?

 

[ptonsparky]

Dang, I always like to hear the solution. Even if it was turn the switch on/off

 

[Jraef]

You saying as moderator you can tell if a member read a post even if they never made any reply?

No, but the stats show when they were last logged in. Compare that to the time of the post and you can see that he has not been back since posting his question. It’s possible that he read responses without logging in, there is no way to tell if that happened. But he can’t respond without logging back in.