Detecting Motor Rotation when Motor is not Energized

[mull982]

We have a 350hp motor connected to a 480V VFD. We recently have found that due to the location of the motor a fan upstream is pulling a draft and causing the load on this 350hp motor to spin and thus causing the motor to spin. This motor and load spinning is causing mechanical problems (suprisingly no electrical problems yet) when this motor is spinning and then started while spinning.

We are trying to determine a way to detect if this load and motor is rotating (due to upstream draft) prior to starting the motor with the vfd. The only thing that I can think of is putting a shaft encoder on the shaft of the motor? Witout a shaft encoder, and with the motor not energized is there a way for the drive to be able to detect that the motor is turning? For instance is there enough residual magnetisim remaining in the motor when it is de-energized for the drive to detect a forward or reverse rotation? Maybe with some sort of sensorless vector setup??

What happens if this load is spinning forwards due to the draft and is then started while spinning in the direction of rotation? Will the current drawn be less since the motor is already running at a higher rpm point on the motor speed/torque curve?

What about if the motor was spinning in the reverse direction when started? I have heard that the motor would draw more than LRC due to the fact that it would be operating with a negative slip. Would the current seen be much larger than LRC? This is what I would think you would see with a DOL start but have no idea what you would expect to see with a VFD start.

Is there a function on the drive similar to DC braking that would lock the motor in place when not running.

 

[mcclary's electrical]

We have a 350hp motor connected to a 480V VFD. We recently have found that due to the location of the motor a fan upstream is pulling a draft and causing the load on this 350hp motor to spin and thus causing the motor to spin. This motor and load spinning is causing mechanical problems (suprisingly no electrical problems yet) when this motor is spinning and then started while spinning.

We are trying to determine a way to detect if this load and motor is rotating (due to upstream draft) prior to starting the motor with the vfd. The only thing that I can think of is putting a shaft encoder on the shaft of the motor? Witout a shaft encoder, and with the motor not energized is there a way for the drive to be able to detect that the motor is turning? For instance is there enough residual magnetisim remaining in the motor when it is de-energized for the drive to detect a forward or reverse rotation? Maybe with some sort of sensorless vector setup??

What happens if this load is spinning forwards due to the draft and is then started while spinning in the direction of rotation? Will the current drawn be less since the motor is already running at a higher rpm point on the motor speed/torque curve?

What about if the motor was spinning in the reverse direction when started? I have heard that the motor would draw more than LRC due to the fact that it would be operating with a negative slip. Would the current seen be much larger than LRC? This is what I would think you would see with a DOL start but have no idea what you would expect to see with a VFD start.

Is there a function on the drive similar to DC braking that would lock the motor in place when not running.

Starting while spinning in the same direction will decrease load. Starting while spinning the opposite will increase load, but will not hurt anything if the ramp times are set appropriately. If it were a vector drive it could hold the load steady, but you would still need input from encoder. Do you have the option of mechanical braking when stopped? An encoder would recognize the motor is spinning, but it's not fixing the problem. It's just "seeing" it. With proper ramp times, I don't think you would have a problem. Most ramp times for 3-4 hundred HP are about 15 seconds-60 seconds. How fast are you ramping? If you're taking out a contactor and installing drive to stop this problem, I would just go with slow ramp time, and it wouldn't matter if it's spinning backward or not.

 

[Cow]

So if this motor is spinning already from the draft of the fan in front of it, how are you starting it now? It's sounds like you need some sort of electric brake or you have to start them both from the beginning at the same time.

If that's the case I'd just use a CT relay on the other fan to break the start signal to this motor.

 

[mull982]

Starting while spinning in the same direction will decrease load. Starting while spinning the opposite will increase load, but will not hurt anything if the ramp times are set appropriately. If it were a vector drive it could hold the load steady, but you would still need input from encoder. Do you have the option of mechanical braking when stopped? An encoder would recognize the motor is spinning, but it's not fixing the problem. It's just "seeing" it. With proper ramp times, I don't think you would have a problem. Most ramp times for 3-4 hundred HP are about 15 seconds-60 seconds. How fast are you ramping? If you're taking out a contactor and installing drive to stop this problem, I would just go with slow ramp time, and it wouldn't matter if it's spinning backward or not.

The ramp time that is currently set up on this drive is 3 minutes (very long). The ramp down time is the same. I belive these long ramp times were set due to prevent any regeneration onto the DC bus since this is a very high inertia spinning load.

Can you explain how a vector drive could hold the load steady with the use of an encoder?

We do not have an option of mechanical braking at this point.

So if this motor is spinning already from the draft of the fan in front of it, how are you starting it now? It's sounds like you need some sort of electric brake or you have to start them both from the beginning at the same time.

We just discovered this problem today so we were previously trying to start this motor while the fan upstream was still spinning. Can you explain what type of electric brake can be used in this application? We have a braking resistor on this drive that is controlled by a chopper when the DC bus reaches a certain level.

 

[brian john]

They make phase rotation meters that will work for this application connect manually spin the motor and. that's it.

 

[mull982]

They make phase rotation meters that will work for this application connect manually spin the motor and. that's it.

Doesnt there have to be a magnetic field in the motor for it to produce the voltage rotation required for these phase rotation meters to pick up? Is there enough residual magnetisim left in the motor after its de-energized to produce this voltage? Or does the phase rotation meter supply the magnetic field to the motor required to produce the voltage?

I'm assuming I could install this phase rotation meter inside the drive on the T-leads to the motor.

 

[mcclary's electrical]

controls need to do the following.
1. Supply enough energy to the system to run full load at rated speed.
2. Dissipate or take out of the system the energy of the fan being drafted
3.Provide over speed protection.
4. Provide means of braking for the load.
Items 3 and 4 are the most important requirements.
With a Load Brake a standard open loop V/Hz VFD will handle most applications.
With a Load Brake type , the mechanical load brake itself handles items 2, and 3 above
along with being the second brake while the motor does the work.
VFD then only needs to
run the motor and control the electric brake.
Without Load Brake a closed loop VFD (flux vector) with encoder is required.
No Load Brake systems require more from the VFD control system. The drive needs to be able
to dump full power into properly sized regeneration resistors. This regenerative braking torque
provides the second means of braking and takes the energy out of the system while slowind down the
load. The over speed protection is then handled with encoder feedback to the VFD. Motor
speed & direction information (encoder feedback) is a very important safety component
without mechanical load brakes. It dates back to static step less controls. When VFD’s are
looked at for control , the VFD choice should be based
on designs that are
available with closed loop control. The choice was (and is) a closed loop Flux Vector Drive
AC drives for many
applications becoming very popular. If there is no mechanical load brake
in the system, power and safety issues must be addressed. Since the motor is used to hold
the load, it is very important that the drive has more motor information than just motor amps.
Using motor speed & direction feedback along with motor current gives the drive a true
indication of what the motor is doing. AC drive type used for this application is a closed loop,
flux vector drive.
For No Load brake applications, there are three requirements:
• Flux Vector Drive
• Dynamic Braking
• Encoder Feedback
FLUX VECTOR DRIVE – The most common type of AC drive that can take an encoder
feedback is a flux vector drive. Where the true flux vector technology is not required from a
safety perspective, it does offer superior performance and allows features like floating the load
and 1000:1 speed range. PEI’s flux vector drive is the Multi-Vector? Series. PEI can provide
the drive module or full built up systems are available. See the Multi-Vector? section following
this page for more information.
DYNAMIC BRAKING -- Without a mechanical load brake, the motor is used to keep the load
from falling. The regenerative energy must be dissipated. This is done with a dynamic braking
(DB) system consists of a DB transistor and DB resistor. The DB transistor is built into every
PEI drive. A separate DB resistor package must be utilized for No Load Brake Hoist
applications. PEI will supply the needed resistor.
ENCODER FEEDBACK – To give the motor speed feedback to the drive requires an optical
encoder. It is a simple device that sends pulses back to the drive indicating both speed and
direction. The most common encoders are 1024 pulses per revolution, with “quaderature” (A &
B channels -- to derive direction), and include “complements” (apposing channels that goes low
when its counterpart goes high – for increased noise immunity). The best location to mount the
encoder is on the motor shaft. Generally a stub shaft is mounted on the back of the motor shaft
to add an encoder. Using the Multi-Vector? drive, it does not have to be mounted on the motor.
It can be mounted on any of the rotating device such as the drum or gearbox. Do note you will
not get the micro-speed performance and the loss of some special vector features if you do not
mount it on the motor. You also have to address backlash issues of the gearing as well. The
cable assembly between the encoder and drive should be shielded and one continuous cable with
no terminations except at the motor and drive and the shield grounded at the drive end only.

 

[brian john]

http://www.megger.com/us/products/ProductDetails.php?ID=217&Description

I have 3 of them care to buy one.

 

[mcclary's electrical]

Brian and the OP are talking about two different things

 

[mivey]

Brian and the OP are talking about two different things

The OP asked if you could detect the rotation direction without power. As Brian explained: yes you can.

 

[brian john]

YEAH I AM WAY OFF ON ThIS ONE....Not sure what I was thinking but they have my meds and bed ready at the local nut hut...MY APOLOGIES>>>>

 

[mull982]

http://www.megger.com/us/products/ProductDetails.php?ID=217&Description

I have 3 of them care to buy one.

I would be looking for something more permenant that would be mounted somewhere and would have contacts to provide the rotation information to our DCS.

It sounds like an encoder is probably the best solution.

I am however very interested in electric braking techniques that can be implemented when the motor is not running?

I am also interested in hearing if the drive can detect rotation without an encoder when the motor is not running? Similar to sensorless vector by calculating motor parameters?

 

[Smart $]

I would be looking for something more permenant

Consider a solenoid-operated pneumatic brake. These are used on steel mill motors of comparable size which operate the cooling bed cams. Of course they use a DCS feedback loop via an encoder to determine exact positioning... but in your case it sounds like you don't need to know the position, you only need it to stop the system when the motor is deenergized.

 

[benaround]

Why can't a damper be installed in the duct ?

 

[topgone]

Why can't a damper be installed in the duct ?

Yup, that got me wondering too! Most fans I have tinkered with have IGV's (inlet guide vanes as opposed to dampers on the discharge ducts)) which is interlocked with the starting controller. When the fan is off, the IGV's are closed automatically so you don't have to brake or monitor whether the fan is free-wheeling!

 

[AdrianWint]

Mull,

Some of the drives have a function called 'flycatching' whereby the drive can detect that the load is free spinning. When a start command is received, the drive will automatically ramp its output frequency, at low power, until it 'catches' the spinning rotor & then either accelerate it or decelerate & reverse its direction as needed. Check the documentation of your device to see it it implements this. Most of the bigger Eurotherm drives do.

Adrian

 

[chaterpilar]

Installing sensor ( rpm based with digital ouput) on the Fan shaft end and interlocking it with the Logic input (Run command of the VSd) will be economical.

Please see the link below..

http://www.electro-sensors.com/index.php/cPath/9




Cheers.

 

[Besoeker]

We have a 350hp motor connected to a 480V VFD. We recently have found that due to the location of the motor a fan upstream is pulling a draft and causing the load on this 350hp motor to spin and thus causing the motor to spin.

We often use an inductive proximity detector for that kind of thing.
Assuming you just need to detect that the motor is moving (rather than knowing it's speed) a fairly simple arrangement would do.
You can pick up on anything metalic that moves in and out of the proximity detector field.
Obviously an encoder would work but it seems like a sledgehapper to crack a not.

Pepperl+Fuchs make a range of sensors that we have found to be robust and reliable.

 

[mull982]

We discussed this issue more this morning and came to the conclusion that others stated saying we need to start this motors with the fan dampers closed. We will put an interlock in our DCS that the fan dampers must be closed prior to starting this motor.

Mull,

Some of the drives have a function called 'flycatching' whereby the drive can detect that the load is free spinning. When a start command is received, the drive will automatically ramp its output frequency, at low power, until it 'catches' the spinning rotor & then either accelerate it or decelerate & reverse its direction as needed. Check the documentation of your device to see it it implements this. Most of the bigger Eurotherm drives do.

Adrian

This sounds like something I am interested in researching futher, either as a backup percaution on this motor or if nothing else for my own knowledge. Can anyone explain to me how this technology works or point me to any information on this subject? The drive we are using is an Rockwell Powerflex 700S and I'm not sure if this drive has this capability although I will look.

 

[mull982]

Ok I spend some time searching for this "flycatching" topic and found information on what it is otherwise called as "flying start enable". It appears that if a motor is already spinning the drive will be able to detect this spinning and the frequency that it is spinning at and thus output power to the motor at the frequency matching what the motor is spinning at. This will alleviate any current or mechanical transients associated with starting the drive at a different frequency.

I'd like to know from a theoretical standpoint how this works however. How does the drive know what speed the motor is running at? Does it have to use an encoder or can it be done without an encoder? It looks like I have seen some cases where it can be done without an encoder. Is there enough residual magnetic flux in the drive where the drive can look at a small regenerative voltage coming back from the motor and thus determine frequency?