Powerflex 525 Pulse Input Response

BlueAltezza

Member
Join Date
Nov 2012
Location
Cleveland, OH
Posts
243
I set up a Powerflex 525 drive to follow a signal from a master conveyor's encoder. The follower motor has an encoder wired back to the 525 encoder card.
Speed reference is set to Pulse, Feedback is set to Quad Check, Pulse Scaling and PPR values are set to match. Accel/Decel set to .1, basically followed the documentation on the knowledgebase to get a master/follower set up.

The follower starts and stops with the master conveyor belt and does follow the speed properly, but the response/reaction to the changes in speed of the master conveyor is way too slow.

Is this normal for the setup, are there other parameters that need adjusting, or should I be trying to set this up differently (PID maybe)?
 
...but the response/reaction to the changes in speed of the master conveyor is way too slow.

Could you be more specific when defining 'too slow'? Are there any controllers involved within?

Also, what causes the 'master' to change speeds? Reference or load?

PF525s are decent VFDs, however, one cannot expect the features and accuracy of the PF750s.
 
Basically the slave drive/motor combination has to match the performance of the master as closely as possible. So start there. Does your slave drive keep up with the master in terms of acceleration and deceleration? If not, fix that first by either upsizing your slave, or slowing down your master.

Your accel/decel in the slave drive should be set to zero if possible, since it has to match the master's accel/decel rate. But .1 sec is pretty close to zero...

I don't see that using a PID would help, but then again I don't know much about your application. If it's on a marriage line, then a PID loop is probably going to cause more trouble than it's worth (including when it decides to run backwards). If the two conveyors aren't tightly coupled, and all that you really want it to do is to stop in a position relative to the master, then maybe...
 
Before diving too deep into it I'd first quickly set up an old fashioned 'Analog Follower' setup between a Master VFD and the Follower one. You might be surprised how accurate old-school implementations still are.

The Master will generate a 0-10VDC on-board Analog Output signal proportional to its 'Speed Feedback' signal which will be connected to the Analog Input of the Follower programmed to use its on-board Analog Input as Speed Reference.
 
The PowerFlex 525 has four different Accel Time and four different Decel Time parameters. In general, the default is for it to use Accel Time 1 and Decel Time 1, but it's worth checking to be sure that it is doing so and that you aren't inadvertently selecting one of the other Accel Time or Decel Time values with discrete inputs or bits in the networked Command Word.

Is the drive connected to an Ethernet or other network connection ? If so, make sure that you aren't sending an Accel Time 1 or Decel Time 1 value to the drive as a DataLink.

How many pulses per revolution is the encoder ?

And what is the gear reduction between the Conveyor Motor and the Encoder ?

A VFD can follow a high-PPR (like 1024 or more) pulse train from a motor tailshaft relatively easily, but if the pulses are coming at a slower rate because of a reduction in the drivetrain speed, that affects the ability of the velocity loop to give good performance.

Also, RA Knowledgebase Article ID 628200 (Access Level: Everyone) recommends to set the Accel Time 1 and Decel Time 1 to zero, and to set the Jerk (rate of accel change) option to zero.
 
Could you be more specific when defining 'too slow'? Are there any controllers involved within?

Also, what causes the 'master' to change speeds? Reference or load?

PF525s are decent VFDs, however, one cannot expect the features and accuracy of the PF750s.

Too slow to me is over a second for the follower to match the master's speed. I always expect some minor reaction delay, but when it comes to seconds it's just too much. The master will change speeds based on the customer's line status - Downstream or Upstream conditions may slow the machine down and return it to normal speed. There's also a delay during the start/stop phase, so even if the machine didn't changes speeds while running the start/stop part is still a problem.

Basically the slave drive/motor combination has to match the performance of the master as closely as possible. So start there. Does your slave drive keep up with the master in terms of acceleration and deceleration? If not, fix that first by either upsizing your slave, or slowing down your master.

Your accel/decel in the slave drive should be set to zero if possible, since it has to match the master's accel/decel rate. But .1 sec is pretty close to zero...

I don't see that using a PID would help, but then again I don't know much about your application. If it's on a marriage line, then a PID loop is probably going to cause more trouble than it's worth (including when it decides to run backwards). If the two conveyors aren't tightly coupled, and all that you really want it to do is to stop in a position relative to the master, then maybe...

If I run them independently they are close when starting and stopping, dialing in the ramp values gets it pretty close. I just want tighter tolerance between the two - more detail in my next reply

Before diving too deep into it I'd first quickly set up an old fashioned 'Analog Follower' setup between a Master VFD and the Follower one. You might be surprised how accurate old-school implementations still are.

The Master will generate a 0-10VDC on-board Analog Output signal proportional to its 'Speed Feedback' signal which will be connected to the Analog Input of the Follower programmed to use its on-board Analog Input as Speed Reference.

I usually set them up that way (and prefer it due to ease) and for most of my applications it works well. In this particular type of machine that setup hasn't worked well in the past as products that are between the master and follower would rotate or start to lean due to the speed difference. I did a master/follower setup with different drives on my last application like this one and it actually made the customer and my techs happy. I did write that on the calendar in case it never happened again! This customer spec'd 525s so I'm trying to mimic the functionality of what I did last time.

The PowerFlex 525 has four different Accel Time and four different Decel Time parameters. In general, the default is for it to use Accel Time 1 and Decel Time 1, but it's worth checking to be sure that it is doing so and that you aren't inadvertently selecting one of the other Accel Time or Decel Time values with discrete inputs or bits in the networked Command Word.

Is the drive connected to an Ethernet or other network connection ? If so, make sure that you aren't sending an Accel Time 1 or Decel Time 1 value to the drive as a DataLink.

How many pulses per revolution is the encoder ?

And what is the gear reduction between the Conveyor Motor and the Encoder ?

A VFD can follow a high-PPR (like 1024 or more) pulse train from a motor tailshaft relatively easily, but if the pulses are coming at a slower rate because of a reduction in the drivetrain speed, that affects the ability of the velocity loop to give good performance.

Also, RA Knowledgebase Article ID 628200 (Access Level: Everyone) recommends to set the Accel Time 1 and Decel Time 1 to zero, and to set the Jerk (rate of accel change) option to zero.

I'm controlling the drive via Ethernet (L43) and don't have the ramp values being set from any controller tags. I'll set the other ramp values to .1 to make sure those aren't the issue. The encoder on the conveyor motor is 1024 (scaled to 310 in A537), encoder on the follower's motor is 120. I'll check the other ramp values and jerk rate.
Thanks!
 
I think Ken pointed out the culprit: Master pulse rate and Follower feedback pulse rate.
A VFD can follow a high-PPR (like 1024 or more) pulse train from a motor tailshaft relatively easily, but if the pulses are coming at a slower rate because of a reduction in the drivetrain speed, that affects the ability of the velocity loop to give good performance.

You have a 33% reduction in Slave Reference from the Master generated Pulse (1024:310) and then another 85% for the Slave's own feedback (1024:120); that's quite excessive and a one second 'response' for a PF525 is pretty 'normal'.

This seems to be a pretty critical speed follower application; again, PF525 are not quite suitable for this kind of 'accuracy', however, since it is a customer specification, you might want to look into higher PPR encoders on both the Master and the Slave sides.

The velocity loop response 'speed' is directly proportional to the pulses' rates (Reference and Feedback); a higher pulse rate (even reduced by the transmissions' ratios) will increase the response proportionally, hopefully to the level where it will comply with the application's requirements.
 
Good info, thanks! I don't have any higher resolution encoders to slap on the machine so for now I have to go old school and use the analog out from the master.
 

Similar Topics

Are the N files in the 525 the same as the 40p for ethernet? (E-Comm card) I have used 40P with Micrologix 1400 Messaging but don't see the N...
Replies
0
Views
54
Hey all, I am currently working on a depalletizer for a customer and we are doing a hoist upgrade. This is a SLC500 processor and the drives are...
Replies
6
Views
183
Hi I got a powerFlex 525 AC Drive (used) i changed parameter t053 to 2 to do factory reset and now i have both parameters 46 and 47 on 1 but cant...
Replies
2
Views
130
Got this new to me powered on today and have no main air and the only problem I have found is the f059 on the vfd. S1 has 24v, none on S2, and 48v...
Replies
10
Views
333
So our Powerflex 525's are set up to allow REMOTE operation from PLC through network. But if network is lost we want to be able to run in LOCAL...
Replies
5
Views
483
Back
Top Bottom