VFD Not Consistently Ramping down

JankyPLC

Member
Join Date
Sep 2016
Location
WI
Posts
46
I've got a Yaskawa GPD315 running a belt conv and am having issues getting it to stop in a consistent time. System has a PE that when hit, stops the run command to the drive which is hardwired.

Best guess right now is as it gets warmer in the plant the belts have a little more friction causing the belts to stop quicker because in the morning when it's cool, and on days it doesnt get hot, we have fairly consistent stop positioning.

Things that have been checked are:
N020 - Decel Time 1 = 1.0 (second)
N023 - S-Curve Selection = 0 (Disabled)
N005 - Stopping Method = 0 (Decel (ramp) to stop)
N024 - Speed1 =5.00
N025 - Speed2 =63.6 (speed I'm deceling from in above question)
N026 - Speed3 =21.2
I'm not getting OV(overvoltage faults).

Any ideas?
 
Braking resistors installed?


Maybe heat is effecting one or more. If cracked to where when they get hot they open or a short piece of exposed wire overheats to where it creates too much resistance. This would drastically reduce the dynamic braking.



Also, does the motor or drive system have a mechanical brake that could be effected by heat?


EDIT: just read the part where it appears to be stopping quicker when hot. For this the mechanical brake grabbing more when hot might be relevant.


EDIT2: Or a bearing or other mechanical problem that drags when hot.
 
Last edited:
Braking resistors installed?

This would be my first suggestion, if not already in place and programmed to enable the braking transistor. Otherwise, the Bus Voltage Regulation may well be kicking in and extending the decel time. Having a DB resistor in place and enabled should allow the drive to follow the set ramp time, or kick out on OV (if not strong enough).

You may be able to simply extend the decel time, so the inertia plays a smaller part than the friction, thus the motor is being powered to a stop, rather than the motor over-powering the drive.
 
You may be able to simply extend the decel time, so the inertia plays a smaller part than the friction, thus the motor is being powered to a stop, rather than the motor over-powering the drive.
Extending the decel time ( distance ) just gives the friction more time or distance to act on the same inertia/momentum.
I don't see how this will make the deceleration more consistent UNLESS there is some maximum deceleration rate that is being exceeded.
 
If the drive is not capable of giving the load any negative torque (due to the bus voltage), you can extend the decel time beyond the coasting time, and thus some positive torque would be required to follow the ramp.

It may not be feasible due to bandwidth/cycle time/distance, but it will be more consistent...
 
Extending the decel time ( distance ) just gives the friction more time or distance to act on the same inertia/momentum.
I don't see how this will make the deceleration more consistent UNLESS there is some maximum deceleration rate that is being exceeded.


The OP seems to state it stops FASTER when hot, too soon for where the product should be.


For this a Decel time would help, instead of relying on it coasting for a specific distance.
 
A few things to think about on this
I presume that you have a breaking resister correctly installed on the VFD
Check the wiring for a lose connecting that could be heating up as the day goes along
The fact that you are not getting an Overvoltage fault means that you are dumping at least some of the energy through the breaking resister or you are not breaking with the VFD .
Have you actually measured the deceleration time is the deceleration time actually getting longer or re you just overshooting the position.
It looks like you are trying to do positioning using just the deceleration time alone that’s not the best way to do positioning to many variables
Was this system working correctly for some time and starting to have problems or has it had this problem from the start.
How are you holding position when the VFD is turned off, things can drift.
The product entering the belt conveyor can push the belt a little as the part makes the transition from one belt to another
Dos this conveyor have a motor brake. If so the motor break may actually be stopping the conveyor and not the VFD ( that would explain why you are not getting the OV fault ) on a motor the break friction goes down are the temperature goes up
Here are many possibilities without actually seeing the entire system we are all just guessing, but it should give you a few things to look at.
You should consider converting this to newer vector drive with an encoder feedback. Use the VFD to hold position at stop that way you eliminate a lot of the variables. This way the distance from the PE to the final stop position will always be consistent.

Let us know what you find
 
I agree positioning with no feedback is not ideal. Process change means we need to stop parts in a 6" window where before position never mattered. Ideally this conv would have an encoder like the one its feeding which this conv should be following the same command profile of. If the problem is unsolvable without encoder, it will unfortunately be the last step due to age of machine.

No braking resistor involved. Anyone know if Yaskawa drives extend decel vs OV fault? I didn't see mention of it in the manual and we do not have faults.

No physical brake on motor, this is a horizontal vbelt style conv so gravity is not fighting us thankfully.

Part is entirely on this one conv section with nothing else entering to push conv. Belt speeds are matched upstream and downstream.

Interestingly this morning - cool temps in plant, conditions where we see the most distance traveled on decel - I shortened the decel param from 1 to .5sec and it certainly stopped quicker which to me says the ramp isn't extending past 1sec due to bus overvoltage. Not sure why stop is so much shorter when hot unless I'm seeing sub-1sec ramp down when hot, in which case lowering decel might make cold stop time more consistent with hot stop time.

I will verify all connections and belt tightness today too to eliminate those variables.

Thanks for the help thus far everyone!
 
disclaimer:
I have not read this entire thread.
I was trying to remember the reason why we didn't go w/ the Yaskawa. I am not bashing Yaskawa. I've heard nothing but great things about the drive. I was like, was it the lack of the rj-45 plug? no, that is the panel builder's problem (the panel builder is the seller of the drive). was it the lack of dc injection motor shaft fixing during stop? No, I fixed that w/ some programming tweaks and a jmpr. Oh, that was it, I couldn't get a consistent stop ramp.

Yes, it is probably possible. I was just too stupid.
We use positioning w/o feedback.
 
No breaking resister if that’s true then they have to be using DC Injection breaking
That would explain a lot
Check the vfd parameters
The thing with DC injection breaking
At best you can only get about 30% motor torque
The drive output transistors are turned off on stop command so the buss voltage will not rise and trip the OV fault
You have to engage it at some motor speed after the stop command and the motor slows some
As the motor speed reduces the breaking torque is reduced so at near zero speed you have 0 breaking torque at no point will you ever be able to hold position with DC Injection

On option would be to install the correct breaking resister and reconfigure the VFD to use it.
If using the internal breaking transistor you still have limitations 10% duty cycle
To get more than that it requires an external breaking module and resister or VFD capable of full line regeneration
I think you may be able to get where you want by adding the breaking resister and turn off the DC Injection breaking
Check with you local distributer for the correct resister
 
No breaking resister if that’s true then they have to be using DC Injection breaking
That would explain a lot
Check the vfd parameters
The thing with DC injection breaking
At best you can only get about 30% motor torque
The drive output transistors are turned off on stop command so the buss voltage will not rise and trip the OV fault
You have to engage it at some motor speed after the stop command and the motor slows some
As the motor speed reduces the breaking torque is reduced so at near zero speed you have 0 breaking torque at no point will you ever be able to hold position with DC Injection

On option would be to install the correct breaking resister and reconfigure the VFD to use it.
If using the internal breaking transistor you still have limitations 10% duty cycle
To get more than that it requires an external breaking module and resister or VFD capable of full line regeneration
I think you may be able to get where you want by adding the breaking resister and turn off the DC Injection breaking
Check with you local distributer for the correct resister

I stopped reading at "at no point will you ever be able to hold position with DC Injection " as that is 100% FALSY FALSE false.
You inject DC into an AC motor it will lock the eff up.
Learned that in Nuke school.
 
I also have not read the entire thread. If you believe DC injection can hold position , I would like to know your definition of "hold position", say in a asuspended load application (crane).
 
Ganutenator

WOW
I have seen you post a lot of bad or misinformation and just let it go with no reply because most people can see for themselves and understand and there is no point in staring the pot but when you point the misinformation directly to me I feel need to post a response
As a reference I was doing drives before most on this site were even born. From pure DC systems Vacuum tube rectified DC motor controls, SCR’s drive systems AC VFD’s to the current VFD used today and I am still learning. I have machine manufactures using my control designs after I redesigned there controls for my clients. Factory service for multiple drive manufactures
I try to provide information to help others solve a problem they can use it or not it’s of no matter to me
And I have seen quite a bit of bad information posted here as well as a lot of good information it up each to take away what they want. Everybody has something to contribute
I like to think everybody that views or post here wants to learn or help somebody else
Posting that something is 100% false to me is on the verge of slander particularly when there is documented proof otherwise. It’s clear to me that you just don’t understand as much as you think you do.
It’s a well established and documented fact that DC Injection is limited to 30% of the motor rated torque. Starting at rated speed and falls off as the motor speed falls off. At zero speed you have zero toque. So holding a load at zero speed as needed for positioning is not possible
Take a minute and do the calculations, zero speed = zero torque no matter how much current / power you apply to the motor , Zero Torque means no power to hold the load. These are facts that don’t change. Try to suspend a load on a hoist at zero speed using DC Injection breaking. Make sure you stand well clear because I guarantee you will drop the load. I have suspended a load on a hoist motor many times using just a Flux Vector VFD and motor, and even them the motor shaft will obsolete for/rev while the load is held in position. But the motor shaft will be moving. The motor feedback tells the vfd where the motor shaft is and it calculates what’s needed to bring it back.

When stopping any motor you need to get rid of the energy in the motor
There are only a few ways to that
Mechanical Break Converting the motor energy to heat through friction in the brake
Dynamic Breaking Converting the motor energy to heat by passing the motor current / torque to a resister through the buss loader transistor in the VFD or using an external breaking module and resister

A quick Note here I like to use the term Buss Loader Breaking over Dynamic Breaking when talking about VFD’s While the way they work is different the end result is the same. Dynamic Breaking is an old term from the pure DC motors.

Common Buss configuration Multiple VFD’s shearing a common DC Buss So as one VFD is absorbing energy from its motor that energy is used by other vfd- motors sharing the common buss

Then there is Line Regeneration the motor energy is passed back into the AC Supply line where it can be used by other motors.
Then you have DC Injection breaking Injecting a DC current into 1 or 2 of the motor windings

Each method has their own advantages and disadvantages the system designer has to weight out all these and chose what’s best for their application sometimes they get it wrong.
Over the years I have designed and installed each of these systems without any problems.



As for you statement “ injecting DC into an AC motor will lock it up “ is false it will rapidly reduce the motor speed to near zero speed, within the limits described above it will not completely stop the motor ( friction on the motor will complete the final stop just like coast to stop) and if the DC is left on it will overheat the motor and burn up the windings that why the VFD limits the time and current for DC Injection Braking . Also some vfd’s allow you to use the DC Injection as a motor winding heater to keep the moisture out of the windings, again within limits

I noticed that you don’t like Yaskawa drives to each their own, But at least be honest about it.
You said NO RJ45 connector, everything in the last 20 or so years has one ( Keypad Port) I use it all the time in fact they also have a built in RS485 multidrop networking connection
So it would appear you are condemning something without having knowledge
They are the largest manufacture of drives in the world. Their drive technology is about 5 years ahead of the others. Tech support is next to none. There are many other reasons I recommend them.
I prefer not to use Rockwell if I have a choice. I have my reasons and will not discuss them here.

As for this application either they have a heavy friction load sopping the motor or they are using DC injection breaking you have ruled out everything else
I would still try the breaking resister I think it will give a consistent repeatable ramp
I have had conversations with tech support and the tell me the breaking is about the biggest support call they get because people don’t understand how it works and how to set it up.

If you want to continue this PM me
 
Verified belts are tight. They are very rubbery thick belts though so I could see temp/humidity affecting them.

Shortening accel and decel params seems to have helped. It's not perfect but I think Gary is right, going to need a Braking Resistor to have as much control as possible in this setup.

Also attached is a snip of the manual explaining when coast to stop is disabled - ramp to stop is enabled (N005=0) how the ramp profiles work with DC braking if anyone finds it interesting.

Thanks for the help everyone.

Screenshot 2022-05-17 110119.png
 
I'm not quite sure how the programming is on that unit. It's been a few years for me since I did much with Yaskawa, but I do remember there was a parameter which you had to change to enable the DB transistor. It was something like 'Stall Prevention during Decel'.

The function was that if it was enabled (default), the drive would decel along the set ramp as long as the bus voltage didn't rise above a threshold. If it reached the threshold, the ramp was disabled (speed held) until the bus voltage dissipated enough to continue (threshold-hysteresis). When disabled, the DB transistor turns on at this threshold, clamping the bus voltage and allowing the decel ramp to continue (or trip on OV).

Hope this helps. I don't want to get in the middle of the squabble above. There are some pretty harsh comments and accusations that I have strong feeling about, but don't want to feed the fire.
 

Similar Topics

Hello, I'm trying to delve a little into rs-485 communications for a couple projects of mine. Until now I've been using a delta vfd and a delta...
Replies
2
Views
18
I have an issue with Power Flex 525 during running processing, the VFD stopped suddenly while the PLC and VFD connection ok, VFD does not have any...
Replies
1
Views
85
Has anyone setup communications with Red Lion 3.0 MODBUS to Baker Hughes Centrilift GCS VFD? Thanks
Replies
0
Views
70
Hello, I've been trying to learn this a while now and still have not found out how this works. I have an Omron CJ2M PLC and an ABB ACS 355 VFD...
Replies
1
Views
190
Looking for some assistance. I am very familiar with Yaskawa VFDs, but not so much with AB VFDs. I am working on some hardwired AB PowerFlex 700...
Replies
2
Views
86
Back
Top Bottom