Indexing with VFD. Best practices? Alternatives?

CapinWinky

Member
Join Date
Aug 2011
Location
Virginia
Posts
566
I have a very low cost application (so, no standard servo in the budget) with essentially a star wheel that need to index one position at a time. It does this by detecting a product entering the star wheel to turn on the VFD, and detects the star wheel reaching the new position with a prox to turn off the VFD. This works up to a certain rotation speed, after which, our motor/drive combo can't stop in a repeatable way.

We are currently using a pretty low end Asynchronous motor with a gear box and a medium end drive with jerk turned off and DC injection turned way up.

Stepper doesn't seem like a viable solution for hard accel/decel applications. I'm wondering if keeping the VFD and using a Synchronous motor instead of an induction motor would improve start/stop performance. I'm not sure why we are using a gearbox, but we don't need that much torque. It may be improving stop repeatability by dividing the motor shaft position error, but I'm wondering if removing it will improve things for us.

Ideas?
 
Use a flux vector VFD it will give you a constant acc / dec times
some flux vector vfds have a built is profiles that can help you get repeated positioning
pm me and maybe I can help you

Gary
 
We use different VFDs depending on the customer. I would say PowerFlex 40 is probably our standard, but we have a version using Altivar 12 VFD (we could upgrade to the Altivar 312, but probably not the Altivar 32). We are starting to make them with Micro820s going modbus to a powerflex 4m.

We would be open to using a different VFD, we just have a policy of playing Rockwell and Schneider off each other to get better pricing.
 
There is a crude (relative to servo) positioning function built into the new PowerFlex 525 that is specifically intended to fill the gap between simplistic VFDs and servos. I've used, it's surprisingly good. It needs an encoder on the motor and an encoder card added to the drive, but it is very simple to implement.

Going to a synchronous motor will not really buy you anything.
 
I have been using Yaskawa for years the pricing I get is very completive.
they are years ahead of the others in technologies and they all have built in Modbus and you can get all the popular network options. I would like to hear more about your project pm me with so we can get into more details the is available here
 
Removing the gearbox will probably work against you because the motor will then have to work harder to start and stop, not counting have to operate a lower speed which asynchronous motors are not good at. If you have a fair amount of inertia in the system, an asynchronous motor will struggle to decelerate it. I would first consider braking resistors, then a higher ratio gear box, and lastly a synchronous motor possibly with an encoder.
 
Problem: Simple index system - Must be Very Low Cost

You have a PX saying when to stop and a VFD controlling the motor

Best Practice is to add an encoder and control the VFD using an analog
then use the count remaining to control speed the VFD to a stop position
Usually existing motor and VFD are ok

Simplest with what you have is to use timers and a second output to the VFD so you have two speeds, Run High Speed for a Time then Engage Low speed before stopping on the Prox
ie the Low speed is your highest existing rotation speed that works.

I cannot see you talking about the Decel time/Rate on the VFD ? Is the drive stopping the motor as fast as possible? Do not forget that to stop the drive very quickly you may need a braking resister.
 
Could you add an additional sensor to ramp to a slow speed before reaching the 'in position' sensor? Or, using one sensor, allow the overshoot, then creep in reverse until the sensor turns off (similar to a servo homing routine)?

🍻

-Eric

EDIT: I see Michael beat me to the punch :)
 
We're slamming things pretty hard, and upon further investigation, we're close to maxing out what is physically possible with our motor/gearbox combo. We can operate at 140 indexes a minute, requiring full stop each index with a 1720RPM@60Hz motor, 25:1 gearbox, and 3 position star wheel, driving it with 90Hz max. That's 8.33333 rotations of the motor per index and the 90Hz gives a max RPS of 43. Using triangle moves, each index is 0.388 seconds, giving a little less than a 10% dwell time.

Realistically, our stop tolerance is +/- 5 degrees, so about +/- half a motor rev before things start jamming up. I'm slightly impressed our current setup can hit that at 140 indexes/minute, since I would not expect it to be able to ramp up to 90Hz and back in under 0.4 seconds. I think this has gone from an exercise in making the system faster to talking about up-selling customers to a servo system.
 
I did a very similar system not too long ago and had the exact same problem. We were only given a couple days to make it work, so the decision was to put a brake/clutch in line with the motor. The motor ran continuously and the clutch would engage until the index position switch was made, then the clutch was released which in turn engaged the brake. It was very hard mechanically on the system, but it did work.
 
The Kollmorgen AKD drive does servo indexing and uses ModbusTCP and aren't terrible expensive, so we're going to check them out. I'll look into clutch/brake combos as well, we could probably drop the VFD for a real economy solution. We may also scrap our PowerFlex 4M version regardless since they aren't vector control. It seems issues with those is the whole driving factor for this effort.

Thanks for the ideas everyone.
 
make a ratchet system, that way enough time to stop.
your way demands a lot of the motor due to its mass.
if you can put a synchrowormwheel in front, you canhave a constant speed. like on a bottling machine.
 
140 start/stop cycles per minute, an induction motor, and a 25/1 gear reducer, is way over the practical limits of the system regardless of the VFD technology or control system. You clearly have gotten it to work for you after a fashion but this is no system for reliable production purposes. A lab or test cell maybe, but not production.

A small budget unfortunately does not change the laws of physics or the reliability of components. Performance on this level need a corresponding level of spending to achieve production-level results and reliability.

As much as I like VFD's, this is a servo job and a basic servo system should be used.
 
Further on my earlier post, there isn't an induction motor made that is designed for 140 start/stops per minute. That's probably the main inadequacy of the system. The gearbox also is slamming it's gear lash back and forth 140 times per minute as well which if far beyond any normal design criterion for gearboxes. There are some out there that are non-servo but with very little lash but you spend more for them. Finally, the VFD itself has a relatively easy time of it. It can dance back and forth 140 time per minute with no problem but it highly unlikely that it can manage a motor properly at that rate, even if it is a flux vector drive (with encoder).

It just too much of a reach. Unless you are an OEM with no concern for your customer's ultimate dissatisfaction with the machine. But, over the long run, that doesn't do an OEM's reputation any good either.

Bottom line: performance at a certain level requires spending at a corresponding level--even if you don't want to!
 

Similar Topics

Greetings, We have an osa-lator 90 degree indexing table in a robotic mig and spot welding cell that has been taking teeth out of the gears on the...
Replies
3
Views
3,899
Hello, I'm in the process of converting a slicer from an Ultra3000 drive to a Kinetix 5100 using the KNX software version 4.00. The current...
Replies
0
Views
498
Hello! The equipment I'm working on uses a lot of indexed tags I guess you'd call them. I thought it was indirect addresses maybe but all the...
Replies
9
Views
994
Hello, Please see the attached rungs of some sort of indexed sampling routine that is currently in a SLC500. I am tasked with migrating to a...
Replies
2
Views
1,262
Hi everyone. Been banging my head against a wall trying to figure out tracking on a servo driven conveyor. I already have a tracking array setup...
Replies
12
Views
3,354
Back
Top Bottom