High-Inertia Load VFD - Dynamic Brake Overheat!

[Steve Bailey]

All those changes in speed are costing you money. Accelerating the load takes energy (kWh) which you're paying for and then wasting as heat when you decelerate the load. Add that cost to the cost of any hardware changes you might have to make to prevent faulting out the drive and you should have a pretty strong case for getting permission to retune the PID.

 

[Peter Nachtwey]

You guys are working for your "Thick as a brick" award.
Tuning the PID will make NO DIFFERENCE if it is in saturation during ramp down or acceleration. Tuning the PID will make a difference when not accelerating or decelerating so much that the current goes into saturation. What about feed forwards? Feed forwards will help with the control when the system isn't in saturation.


It will take about the same amount of energy to go from point A to point B. That isn't the issue. Exceeding what the system is capable of is.
Reduce the speed or lengthen the the ramps so the current doesn't go into saturation.

 

[MaxK]

this is basically a stand-alone system and is only operated when the mill is running in the steady state and the speed setpoint is not really changed during the run.

Actually, why PID controller in this system is needed?

Tuning the PID will make NO DIFFERENCE if it is in saturation during ramp down

Reduce the speed or lengthen the the ramps so the current doesn't go into saturation.

Can you please show me the ramps on #13 curves?

 

[Peter Nachtwey]

Actually, why PID controller in this system is needed?

Closed loop control is required for speed control. However, if the OP doesn't care how the load stops then ramping down in open loop will work too.

Can you please show me the ramps on #13 curves?

[/quote]
There aren't any. So? What does that have to do with over heating or going in to saturation when ramping down?

About the plot.
Does that 5-minute plot look right to you? It doesn't look right to me.
1 I don't trust Rockwell trends for motion because the sample times aren't fast enough and not guaranteed to be at consistent intervals.
2. If this is a high inertia load, why is the speed changing so quickly? Is the speed really changing that quickly?
3. Could it be a sampling and trending problem? Uneven sampling will make it hard to calculate an accurate velocity.

It isn't clear to me if this motor what is generating the green control signal. Is the control signal proportional to speed or acceleration? If it comes from the PLC then it is probably the frequency as stated in #1. Then green line looks like it comes from a PLC register. Why is this value changing so much if it is a frequency? Is this value sent to a drive that actually does the PID control or is the SLC trying to control the speed reference? If so, NOOOOO!

Speed control only requires a PI controller, but it is important that an accurate velocity can be derived from the encoder. This is where many speed control systems fail. If the PLC is trying to do the speed control, how is it computing the speed?

 

[JesperMP]

How was the existing brake resistor dimensioned ?
By going by the VFD manufacturers recommendations ?
My experience with other VFDs, is that the manufacturer just suggest some random size, usually at the lower end for braking low inertia. Also, the manufacturer usually assume that the braking is only at the end of operation - like 1 stop per hour, not for a constant braking action like in this case.
You can possibly go up in size 2x - 3x for the brake resistor. If the problem is that the brake resistor is currently just on the very limit of its operational temperature, then going up in size can be the simple fix.

 

[MaxK]

This topic is the clearest example when participants offer solutions to an issue they have solved previously (I am not exception). Although this is how any discussion works...
Further primitive sarcasm:
drbitboy has already given the most correct answer "LN2 cooling?"
I would suggest better:
1. install an ultra-precise control system
2. build a massive cooling factory next to the mill
3. build a recuperative system
4. .. waiting for proposals...

for a system the requirements for which, data on the functioning of which are not known