Rewinder Machine Upgrade Question

[strantor]

I think I left out a very important factor .... This machine has never ran for this customer as it was bought used but doesn't work. I think (from the few mins I looked at it) that one of the DC drives are dead, The computer/PLC System is pretty much screwed and only the hyd pump works.. well turns on.

The material is small and the customer is hoping the upgraded controls will allow the smaller material to run.

It's good they're hoping, and not confident. They are, not confident, right? because to get a perfect lay with material that small is going to take mechanical precision. The original specs indicate to me that the mechanics that the machine was designed around do not support the precision you need. All the most advanced electronics won't make up for mechanical sloppiness. You can't have any end play in the spindle or play in the left/right axis. This borders on CNC IMO. You can't make a CNC mill out of a drill press.

 

[Critt]

Yes I agree Strantor.

I'm thinking the Hydraulic traverse will need to be upgraded.

I also agree with your thoughts DickDV.
Do you think I should upgrade both the Bridle (Feed nip rollers) and the rewind drive? They would like to use this machine regularly...

Thanks

 

[andypot]

I would base my decision on whether the machine is in continuous service and whether the DC motors are in good condition. If not in continuous service and the DC motors are in good condition, I would probably leave the DC motors in place and put in a modern regenerative reversing digital DC drive. Bardac comes to mind first. Otherwise, I'd use a high performance sensorless vector AC drive with TENV AC motors good to double base speed (probably 120hz).

At 5hp, auxiliary blowers are nothing but a maintenance nuisance.

With sensorless vector, center-driven software, and a good high performance drive like the ABB ACS800 or ACS850, you'd have a reliable efficient system with no motor encoders and no auxiliary blowers.

Servos are overkill for this application.

I'll second that, but as you stated you intended to use CompactLogic I would go for PowerFlex 755 drives because they integrate so well with the PLC. Should you replace both drives, well you need to cost it up. Do the DC motors need overhauling? If you are going to have to strip them out and overhaul, compare that with the cost of new motors and drives and don't forget to add in your labour time for engineering a solution, pick equipment you are familar with and the cost will be lower.

What are you going to do about safety? If I was installing that machine in the UK I would have to do a complete Risk Assessment and then design a suitable a safety system and guarding, I assume you need to do something similar in Canada. It's likely to come out at least PLd so how are you going to isolate the motors when someone opens a guard or trips the light curtain, with a modern drive you just need a Safe Torque Off card, if you stick with the old drives you will need a pair of safety contactors to isolate each drive.

 

[Critt]

Good Point Andy.. the Safe Off is the easier way to go.

Guess I will start to look into sizing these motors from 5 HP DC to AC VFD's...

As for the application, I would be using Torque Mode and calculating roll diameter as the winder rolls, however, this winder also oscillates which throws a wrench into things....

 

[strantor]

As for the application, I would be ... calculating roll diameter as the winder rolls,

Good thinking. Calculating (or measuring) roll diameter and feeding that into the proportional parameter of you PID is a good way to keep things moving smoothly from beginning to end of run.

however, this winder also oscillates which throws a wrench into things....

When you say "oscillate", you're referring to the in/out "traverse" right? or is there some other kind of oscillation going on? The traverse shouldn't throw a wrench into anything.

 

[DickDV]

I don't think you want simple torque control. You need center-driven winder software which combines speed and torque control to give you nice stable consistent tension. I'm not familiar with the AB PF755 but you need a drive that will do zero speed torque control, internal calculation of roll diameter and mass, and torque management error of less than 1% of motor nameplate torque at all speeds, all without an encoder. If it can do that, use it. Otherwise, I'd go with the ABB product. You might have to communicate with devicenet, profibus, Modbus Plus over Ethernet, or some other open protocol.

 

[Critt]

Thanks for all the help so far...
Strantor - the traverse going in and out is what I was speaking about.

DickDV - I agree that I need a drive with those features. I am familiar with both the AB755 and ACS800 so I will go with the best price.
I think I will also communicate with the drive with either Devicenet or Ethernet since I have done these before.

Any thoughts about how to size these 5 HP DC motors to AC type with gear boxes?

 

[DickDV]

5hp in the AC motor is enough. Choose your gearbox ratio so the motor is turning at 80-90hz at the fastest required spindle speed if you are using a general purpose inverter-duty motor. It would be better to find an inverter duty motor that is rated constant hp up to 120hz. If you choose that kind of motor, set your power train ratio so the motor is at 120hz at maximum spindle speed.

I'd choose a Marathon Black Max or Baldor/ABB RPM-AC motor rated to 120hz.

 

[kekrahulik]

My take on it:

1. Servo is definite overkill and might be more difficult to do. If you end up using torque control, servos can do it, but they are better set to control position and velocity - you can't really control torque and position and velocity at the same time. You'll need to play some games to make it work.

2. I (experience in AB) would go with an AC vector drive (PF70, PF755, or PF525 might do) with an encoder. If you go sensorless vector (no encoder), you lose true torque control - you get current control. They aren't exactly the same thing. You may also have some issues with torque control at zero speed. The encoder and Flux Vector Control are worth it.

3. If you have a dancer for tension control and feedback, I would not bother with torque control. I would use the dancer for feedback in a speed controlled winder (a servo could easily do that, but a servo is overkill).

 

[Peter Nachtwey]

The traverse is where all the precision is needed. I'm not a big hydraulics guy, so I don't know what I'm talking about, but I don't think a hydraulic solution is going to cut it.

This has and can easily be done with hydraulics. If the feed back is a Temposonic rod or Balluff rod it should be good enough. Coordinating axes to the accuracies mentioned is easy. Coordinating the DC motors and hydraulic actuators is easy too if you get the right controller. The winder probably needs to use a speed limit/torque limit low select. This is easy too.

you can't really control torque and position and velocity at the same time.

True but you can limit both with a low select.

The nice thing about DC motors is that you can sense the armature current easily. With Ac motors you only can sense the average current as derived by a model.

 

[jdbrandt]

I would base my decision on whether the machine is in continuous service and whether the DC motors are in good condition. If not in continuous service and the DC motors are in good condition, I would probably leave the DC motors in place and put in a modern regenerative reversing digital DC drive. Bardac comes to mind first. Otherwise, I'd use a high performance sensorless vector AC drive with TENV AC motors good to double base speed (probably 120hz).

I would add a 'like' for BARDAC. I see few other good DC suppliers in this application/market, with up-to-date comms, function blocks, etc.

 

[strantor]

I would add a 'like' for BARDAC. I see few other good DC suppliers in this application/market, with up-to-date comms, function blocks, etc.

+1
Ad addendum Bardac has ready made winder solutions. Check out their smarty software, free download.

 

[Critt]

I would like to thank everyone for their input, and feedback to my post.

I am trying to keep the cost down and I agree with Peter that hydraulics are able to drive the traverse winder with the right control.
Would a Compact logic be a suitable solution for a controller?
Would you upgrade the existing DC drives to current ones and leave the motors?

Comparing DC Drives, how does the Bardac compare to the AB Power Flex DC version, or Baldor DC Drives?

 

[califflash]

I would imagine a 5370 L1 compactlogix controller would handle your application just find. Of course that all depends on the logic and instructions your are using.

I would imagine most of your functions could be run directly from the drives. The 525 has a tech note for controlling rewind tension in the manual. The PF755 could definitely handle this without the need of a Plc. You would need a Plc or some kind of PID control to handle the hydraulics.

 

[Peter Nachtwey]

I would imagine a 5370 L1 compactlogix controller would handle your application just find. Of course that all depends on the logic and instructions your are using.

I would recommend a ControLogix because it is faster and the motion modules will fit in the back plane. Our RMC150E will work too/better because all the processing can be done shortly after the beginning of a new update which can be done every 500microseconds. The M02AE would have to transfer information to the ControLogix on the next RPI then ControLogix will process the data during the next scan and then the results are transferred to the slave during the next RPI. That introduces a lot of delay.

I would imagine most of your functions could be run directly from the drives.

The PLC would still need to coordinate everything. Old DC drives usually aren't smart.

The 525 has a tech note for controlling rewind tension in the manual. The PF755 could definitely handle this without the need of a Plc.

Yes, I would avoid using the PLC for the tight control.

You would need a Plc or some kind of PID control to handle the hydraulics.

Separating the control between different controllers will not improve performance nor simplify the coding.
The right motion controller can do it all unless there is a lot of non-motion related I/O.