Standard AC Motor and Drive vs Servo Motor and Servo Drive ?

[Rob S.]

I am working on a few concepts concerning a pick & place . I am picking up a 8 lb part about 1", and moving it about 10".
With my 8 lb part and about 15 lbs of structure , I would like to cycle this part and assembly every 2 seconds. Meaning the pick and place would pick up the part , raise it about 1" , move it forward about 10 " release the part , and then come back ( reverse) empty to pickup another part within 2 seconds , and then another cycle in 2 seconds.

My question is, would a typical 3 phase AC motor and drive , ( if sized larger than I need )work for an application like this , or should I go with a servo motor and drive ?

Thanks in advance for your help.

 

[RussB]

That depends on the placement mechanism and how precise the placement needs to be. If you use linear motion only a servo will give you better speed to placement accuracy. If you use a rotary approach a good VFD will likely work fine.
But to be sure one needs a lot more information.
Can the load tolerate any jerk, open glass of liquid that cant be spilled or a chunk of hard plastic?
etc.

 

[Rob S.]

It is going to be a brick. I am concerned about the forward movement and the g-forces going forward. I would like to use a short accel. and decel time , but not too much to loose the 2 second cycle I am looking to get. The brick will be in a pneumatic gripper.

 

[OkiePC]

2 seconds per complete cycle "sounds" too fast for a VFD method to be accurate, and it would need to be well oversized to handle the duty cycle. The duty cycle is the real killer in my mind. I have seen VFD with position feedback that were plenty accurate...'til you try to go too frequently and hit an overload, over temp or over volts situation.

 

[Rob S.]

Have many of you used SureServo Drives and Motors from
AutomationDirect ? If so , how are they to setup and program ?
Are they durable enough for industrial automation ?

Thanks,

 

[Mark Snodgrass]

As a PLC tech, I hate to say this. If adjustable position placement is not needed, I would go with an all pneumatic setup to get the speed I need for a reasonable price. Plus programing and complexity would be minimized.

 

[RussB]

I know what you mean.
Back in another time I did electrical design and installation for block handling equipment. Most of these types of "transfers" were done with turn tables from above with either mechanical or pneumatic gates attached. The blocks or bricks would slide on UHML plates, or roller tables. Fast and cheap, but they did take some maintaining.

 

[leitmotif]

As a PLC tech, I hate to say this. If adjustable position placement is not needed, I would go with an all pneumatic setup to get the speed I need for a reasonable price. Plus programing and complexity would be minimized.

IF all travels are a straight line I would also favor air cylinders. Acceleration and deleration could be done with "throttle" valves and a taper.

While I do not think a stop reverse every two seconds is good for the motor (even with VFD) it should be OK provided there is adequate cooling for the motor. That is the easy part - the hard part is going to be the gearbox lasting very long - quick starts n stops with reversals wear gears.

Dan Bentler

 

[DickDV]

Just going with my gut and 24 years of experience, this looks like a servo app to me just based on the dynamics alone. If positioning, then certainly a servo.

Precision AC drives are getting very good at torque management these days and there are some remarkably low inertia induction motors out there too but a two second cycle time just seems too demanding to me.

 

[Rob S.]

Thanks for all your responses. When it comes to industrial automation , can you actually have too much air ? Really ,it
seems like going mainly pneumatic is the answer to get the speed I am looking for. But ,it is so taxing on our overall air supply for the plant.

 

[MichaelG]

I disagree with pneumatics - I have been involved in a number of jobs where the air was removed - The air cushioning maintenance of the cylinders fails over time.
One of the jobs was a brick pusher
Continuous extruded clay that were pushed through a cutter to create a number of bricks I think about 16 - cannot remember the cycle time but it was short as the pusher had to be back so the clay would not touch it.

Use a VSD in closed loop (encoder mounted on the back of the motor) - Size it correctly (which you are going to have to do with the cylinders anyway) - this includes external cooling of the VSD (A true closed loop vector motor has an external fan so that it can provide full load current at zero speed)

The benefits are independent electronic control of Accel and Decel rates, also return speed - yes I know that the air can do it with enough fiddly bits but I find VSD's easier to set up, maintain and are more easily adjustable.


Sizing for 2 second cycle time - Without knowing your inertia, I guess that you are border line VSD / Servo - It all comes down to the accel and decel rates required. Servo's can accelerate faster than vector VSD's.

OverVoltage etc - the VSD has not been set up correctly, I have VSD's that decelerate faster then they accelerate - the braking resistor is correctly sized.


Model the system ie get your mechanics sorted - price them (get the reps to do some work), then choose.

 

[alan_505]

If you want to run a normal drive you could use one that allows you to use a servo motor (permanent magnet motor), I believe that the PF750's can do this.

 

[Rob S.]

Thanks, I am leaning toward using Servo. What do you think of SureServo Drives and Motors from AutomationDirect ?

 

[OkiePC]

First, what is the mass and velocity of the total load "as felt" by the motor output shaft (including gearing/drive pulley ratios, etc).

I have not used the A/D servo drives, but I would not hesitate to give them a try with the new PLC Bernie recommended (edit) somewhere in another thread I guess....(/edit) I have been very pleased with every thing else we buy from them.

Check HERE New DoMore PLC for the DL205.

Actually, you could just use the indexing option and not have a PLC at all, unless the index distances needs to be changed.

 

[Peter Nachtwey]

First one needs to make a rough estimate of the motion profiles.
A good estimate for the velocity and acceleration is
v=1.5*delta x/delta t. If x is 10 inches and delta t is 0.75 second the peak speed will need to be about 20 inches per second. The average acceleration needs to be about 4.5*delta x/(delta t)^2 = 80 inches per second squared. This isn't too difficult to do with a VFD or servo motor. However, I question whether one can pick up a brick in 0.25 seconds. If it takes longer to lift the brick then there is less time to move back and forth 10 inches. If delta t is 0.5 seconds the speed increases to 30 ips and the average acceleration increases to 180 ips^2. This is still doable, but if s-curves are used the peak acceleration is 1.5 times the average acceleration if a fifth order motion profile is used. The peak acceleration will be much higher than 1.5 times the average acceleration unless high jerk values are used.