Using Inverter with exsiting slip-ring motor for overhead crane 2

[Banphot]

Hello:

In our steel making plant factory, there is an electric overhead crane with 120 ton hoist driven by 220KW slip-ring motor. The motor is started using 4 steps of rotor resistors, and speed is controlled via SCR circuit.

We would like to install vfd inverter to control speed instead of existing SCR, and utilize the existing slip-ring motor by short rotor circuit at its slip-ring. Could anybody suggest the disadvantage or risk of doing this? Is there any experience of doing this with heavy steel mill crane application?

Thanks in advance.

Banphot

 

[Skogsgurra]

Yes, plenty of experience here.

First. Do not use anything but a good vector controlled VFD. And a company that has done it before.

Then, always fit an encoder with at least 1024 PPR on the motor shaft. Nowhere else. You can run into lots of unpleasant surprises if you do it any other way.

And, very important, install du/dt filters on the VFD output. You will probably destroy your old motor insulation with the aggressive PWM waveshapes if you do not do it. A sine filter would have been nice to have, but most vector vontrolled VFDs have problems with sine filters. The autotuning does not expect extra capacitances in the motor path. Also, a sine filter reduces dynamic behaviour - and that you need when handling molten steel.

Short your rotor winding permanently and remove brushes from the sliprings.

If your crane is as important as most steel mill cranes are, do install a back-up system. And switch between the two every week or month. It can take a very long time (hours - days) to fix a problem in a VFD. And having a back-up system is well worth the money the day you need it.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[itsmoked]

do install a back-up system.

Great point. Would be really bad to have a vat of steel stuck and cooling mid way to somewhere.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[niallnz]

Don't forget to size the drive based on its heavy duty rating. Cranes are constant torque applications, and steel mill cranes are fast, hense you need lots of torque. I upgraded a hydro power station crane of 120ton capacity but only used a 55kW motor on the hoist, hoist speed was slow. Do every thing that Gunnar suggests, must have an encoder feed back on hoists, don't care how good an open loop sensorless vector control algorithum the drive has, for slow speed operation and safety closed loop control is the only way to go.

Cheers Niall

 

[roydm]

Skogs
Will the slipring motor with rings shorted develop as much torque as when Rotor resistance = Rotor reactance?
I was told this is when the torque is Max.
Roy

 

[Banphot]

Very thanks for plenty of suggestion.

Banphot

 

[Skogsgurra]

Roy,

Yes. It will now depend on the drive and its ability to deliver current. Thermally, the motor works under same conditions as before. I expect someone to tell me that the losses that were dissipated in the starting resistors will now develop in the rotor. But that is not so, since you do not run with a large slip any more.

You can run the motor at peak torque all the way from zero RPM and up. Starting DOL does not give you that possibility and that's where the rotor resistor is useful to get a high (equal to peak) torque.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[roydm]

Skogs,
I saw a good example of how the starting torque increases with rotor resistance. A large rail mounted crane with a motor on oposite corners. The electricians mistakenly wired the rotor with the same cable as stator.
When it was started the bogie with the longest cable (too much resistance) just about climbed off the tracks.
thanks
Roy

 

[roydm]

Question, why are you advocating an encoder for the VFD, surely varying the frequency will give much better speed control than the original wound rotor which just gave variable torque.
I must be missing something.
Roy

 

[burnt2x]

Close-loop flux vector control requires speed encoder or tacho for feedback to the VFD controls to operate the driven motor at optimal efficiency. With no feedback (open-loop), the VFD performance can only be as good as the guy who sets the parameters in the VFD's control memory.
Take a look at how deep the owner's pocket is and decide what mode you want.

 

[Skogsgurra]

Roy,

There is a very good reason for using an encoder.

All encoderless VFD drives (except one) have a problem around zero speed. They simply cannot hold the load.

The information about motor position is based on induced EMF in the stator winding and since there is no induced EMF at zero speed, the drive has no idea what is going on and loses control. The reason is that, at higher loads, it is very easy to "slip over" the peak torque. And once there, the internal toruqe loop gets its feed-back polarity reversed. Usually this means that the load hits the floor with high speed, if the operator doesn't hit the emergency brake - and it works.

I have witnessed this a few times when being called to investigate accidents with cranes. Every time, the company that did the installation had reasoned along lines like "a VFD is so much better, so we just put it there". What they did not know, or think about, is that a voltage controlled (or DOL) WRIM doesnt lose control at zero speed since it then is working against rotor flux rotation and usually with maximum rotor resistance, which produces the torque needed.

So, good as they are in many respects, VFDs have a few shortcomings as well. And it is good to know about them in applications like cranes and other loads where the load exerts a static load on the motor at standstill.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[itsmoked]

Hi Roy,

Cranes also need to hold loads stopped possibly putting out full torque while stationary. The encoder assures the VFD knows what's happening and doesn't leave it to an algorithm to flawlessly understand what's happening in a nearly stopped or completely stopped motor.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Skogsgurra]

Smoked

That's a very good way of putting it!

"leave it to an algorithm to flawlessly understand what's happening" - it says it all.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[roydm]

Keith, Skogs,
Thanks for the good explanation
"Cranes also need to hold loads stopped possibly putting out full torque while stationary", that's what the electro-mechanical brake is for, I would expect that to apply as soon as the VFD frequency dropped to about the full load slip speed.
I have crane experience but all prior to VFDs, I always thought that would be a great application, smoother control, longer brake life, no more drum controllers, filing contacts etc.
Regards
Roy

 

[Skogsgurra]

Yes, you are right about the mechanical brake. But a crane that makes slow "hovering" possible without using the mechanical brake is something you should try your hands on. Makes locking a spreader to a container a breeze and also setting it down ashore or in hull. I don't think that a ship-to-shore crane can be built without that possibility. And it is very good in all other applications as well.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[edison123]

Good topic and good discussion. Thanks people.

 

[Banphot]

Hi All:

Extensive value recommendation for all.

Then, it seems very important for the encoder reliability. Is it possible to install the back-up encoder to the motor shaft?

Banphot

 

[itsmoked]

Is it possible to install the back-up encoder to the motor shaft?

Good point! Yes they fail occasionally and certainly they can be doubled up.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Skogsgurra]

Yes. It is possible to have double encoders.

But, the extra arrangements usually makes the installation less reliable. So it is not usually done. Most systems (I think all) can be connected parallel to the encoder*. The input impedance is usually in the 10 kohms range and there's no problem feeding two such systems from one encoder.

There's never a long stand-still when an encoder fails. Provided you have a spare one, which is crucial. It takes a lot longer to diagnose and repair a frequency inverter. That's why you need a double system in critical applications.


*There is some superstition about encoder signals. People construct buffer amplifiers to keep signal integrity and do other complicated and unnecessary things. Never found that necessary.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[roydm]

Keith, Skogs,
Thanks for the informative discussion, I'm sure by my asking some stupid questions others will avoid trying without the encoder.
As for encoders I favour the hollow shaft variety that are supported by the motor bearing. This avoids miss-alignment and finicky couplings.
I have also seen problems with the type of output TTL over open collector. TTL is much faster.
Regards
Roy