VFD stop mode options on a pallet hoist?

[bikerdude]

Hi,

I have to replace a burned out & obsolete VFD on a machine with another make & model. The VFD will be driving a pallet hoist. The hoist drive motor (5HP) couples to a right angle worm gear reducer. The reducer output shaft drives a pair of shafts via roller chains & sprockets. This pair of driven shafts has lifting bars attached that the pallets sit on. On the end of the worm shaft opposite of the drive motor there is an electrical/mechanical brake that is attached to the gearbox housing. This brake is supposed to stop the hoist from moving when there is no power applied to the brake coil. It is energized to release in other words. The brake coil is energized from a relay. The relay is controlled by the machines SLC500 controller. There is a relay output card in the SLC500 chassis for driving the VFD digital inputs. There is an analog in/out card in the chassis for getting the speed signal to the VFD. A shaft encoder is installed on the motor shaft in between the motor & gearbox. I can’t access the parameters in the old drive & there is no documentation of VFD parameters that I can find. I don’t have a brake resistor available in the cabinet although the VFD does have DC brake resistor contacts installed.

I have all the bits & pieces here to make the machine do what I want. I’m just stuck on making a decision about the timing of the brake coil & VFD stop command.

Here is my question: What stop mode should I use in the VFD?

The Choices are:

• Coast
• Ramp
• Ramp to hold
• DC brake

The final position of the pallet when stopped must be accurate & repeatable within 10 mm each time. The run /stop cycle is normally repeated once every 15 seconds. My current plan is:

• Release brake coil: wait 50 ms, then issue run command. ( about 800 rpm)
• Run the hoist up until the pallet height detector is satisfied
• Issue stop command ( ramp ½ second)
• When shaft encoder detects “zero speed” apply brake to hold the shaft.

The current SLC program just releases / energizes the brake coil in series with any run command. This causes a high wear rate to the brake pad and provides a “jerky” movement. I can just leave the program the way it is & use a “Coast to Stop” setting, but I thought maybe I could make some improvements to the motion & reduce the brake wear while I’m mucking around with the drive.

What do y’all think I should do?



BD

[bikerdude]

mountainbikedude,

Who is the Palletizer/Depalletizer made by?

I have read the original blue prints (yeah they are really blue) several times but I can't remember. Seems like the De-Palletizer was originally built by "Crown Simplimatic" but don't hold me to that. I'm at home right now so I'll have to get back to you on that one. Probably one of the many manufactures gobbled up by Goldco_Barry-Wehmiller whatever.
BD

Edit: hey!! there waas a reply here a minute ago, now it's gone.
mountainbikedude must have retracted it. Anyhow.. this picture is real close to the actual machine.

[mountainbikedude]

Who is the Palletizer/Depalletizer made by? I was a field service engineer for Goldco for a few years.

You can go very deep into this subject

Just for your reference since you are not using a DC braking resistor which I do not think you need, your choices are limited to either
Coasting- When a stop command is issued the drive just shuts off power and the drive coast to a stop, since your motor attached to a gearbox it would probably stop quickly
Ramp- When a stop command is issued the drive uses your decel time to ramp the drive down to zero speed

Using your solution below I would set it to ramp

The final position of the pallet when stopped must be accurate & repeatable within 10 mm each time. The run /stop cycle is normally repeated once every 15 seconds. My current plan is:

• Release brake coil: wait 50 ms, then issue run command. ( about 800 rpm)
• Run the hoist up until the pallet height detector is satisfied
• Issue stop command ( ramp ½ second)
• When shaft encoder detects “zero speed” apply brake to hold the shaft.

Without any regard to stop position I would adjust your release brake coil time and decel times until you have a nice smooth stop.

After you have that if the pallet is out of position adjust the pallet height detector sensor or if the stop is done off of the encoder value in the plc adjust the stop value

Caution with adjusting the decel time.
Sometimes at the beginning or the endo a cycling depending on if its a depal or pal the hoist runs faster than normal so...

The bottom picture is definitely a Goldco PET Depalletizer, I designed(controls) and did the startup on alot of these machines.

One thing to be careful with is when the pallet has been fully emptied, The speed at which the pallet hoist lowers the empty pallet is normally faster than a typical raise speed, you dont want your pallet slamming the bottom of the machine

If you need some more help let me know, they are fun machines tune

[bikerdude]

dude,

I thank you for the help.
I'm pretty sure there is a prox about 18" from the bottom of the hoist that is supposed to used to set the motor to a slower speed. This prevents the slamming effect when used in the program properly. The raising of a "Fresh" pallet is faster than a normal raise cycle, but the height detector is a different device & the accuracy isn't as important.
When a fresh pallet is raised it will stop when a PE is made. The machine then goes into "Manual" while the operator cuts the pallet banding. When the operator pushes the "Auto" PB the machine begins the normal cycle.
I'm just curious how accurate the final stop position will be if I use the (ramp/then brake) instead of the (coast while braking). I assume the most aggressive stop method will be more accurate but cause more wear. I should be able to find a compromise between jerking to a stop & easing into the wrong position.
BD

[OkiePC]

Originally Posted by bikerdude

I'm just curious how accurate the final stop position will be if I use the (ramp/then brake) instead of the (coast while braking). I assume the most aggressive stop method will be more accurate but cause more wear. I should be able to find a compromise between jerking to a stop & easing into the wrong position.
BD

Ramping will provide much better control and less wear than coasting into a brake, as well as control varying loads much better. Don't use DC injection braking (Ramp to Hold) because your duty cycle may not support it. If that thing moves full travel in 18 seconds non stop, you will cook the windings of the motor with DC injection braking (if that is what is meant by "ramp to hold")

Also, I would ensure that the brake circuit is in series with the drive fault command and any other device that may need to prevent the brake releasing. Monitor the brake relay too, so that you can ensure the drive RUN command goes off if the brake relay does not pick up within a very short time.

That will help you avoid slippage of the suspended load.

In the past in situations like yours, with high inertia loads on a drive, I have placed a "drive running" status drive relay output hardwired in series with e-stop and PLC enable relays to the brake relay coil.

I didn't have any issues with timing this way, the brake relay will come on right when the drive applies motor power, in fact it reduced slippage over the motor starter method.

The trickly part is going to be acceleration and deceleration under heavy loads if the gear train is easily backdriven.

Since the program is already written to control the old drive via analog, you can probably look at the code for a clue as to hte drive settings.

I tend to program high inertia loads with the lowest possible accel and decel rates, and then vary them in the PLC program.

If the PLC program merely switches speeds, then your observation of the past operation is the best clue to setting the rates.

All this is dependent on the machine and the loads and speeds...

What's the range of pallet weight?

How many feet/inches per seconds?

Paul

Sounds like a fun project!
Good luck!

[bikerdude]

Okie,

Good advice.
Weight ranges from empty pallet = 45lbs.
Full pallet.................... = 300lbs.
There is some backdriving due to wear on the worm gear, I consider it to be normal amount. Very slow drift in other words.
Full speed is aprox. 2-3 feet per second.
I am making the program edits right now. Below is screen shot of the brake rung.


Thanks Again
BD

[DickDV]

As OkiePC has said, the best control, the most predictable stopping time and distance, and the least wear on the mechanical brake is almost always acheived using ramp to stop with snubber braking on the drive and engagement of the mechanical brake for holding purposes as soon as you reach zero speed. Since a PLC is involved, it should be fairly easy to get mechanical brake release delays when starting to avoid a short drop or jerkiness when starting a heavy load up in mid travel.

[bikerdude]

DickDV,
Thanks for the reply. It looks like 3 for 3 voting on the "Ramp" to stop. I hope 1/2 second decell ramp from 1750 rpm to zero is OK. This will only happen once per pallet when a new or "Fresh" pallet goes up to the unbanding height at full speed. Not sure what is meant by "snubber" braking. All the other stops will be from 800 rpm to zero.

I'll set the brake release delay the same on all upward & downward movements. Probably end up in the 20ms range.

For the delayed brake set after a stop command is sent, I'll wait 1/2 second then let the brake come in. Should I check for zero speed just before applying the brake?
The original program just uses the zero speed signal to enable another start command. Probably to prevent the operator from quickly changing directions in manual mode.
Just to be clear on the details, the encoder will connect to the PowerFlex-70 VFD, not a HSC module. I plan to set:
Parameter 80 to "Encoder" for speed feedback selection.
Parameter 412 to "Qaudrature" for encoder type.
Parameter 413 to "1024" ppr for encoder resolution.
Parameter 342 to "speed feedback" to get the speed signal out as analog 4 to 20ma sent to an analog input module. Hopefuly parameter 342 will cause the analog out signal go to 4ma when the encoder pulses stop, and it will be accurate & fast.
BD

[DickDV]

Snubber braking is sometimes called dynamic braking altho that is really a DC motor term. Snubber braking involves a voltage sensing solid state switch which watches the drive's DC bus. If the voltage rises too high, the switch closes diverting some of the excess bus energy to a brake resistor.

I use the term snubber braking as opposed to regenerative braking which involves inverting the excess DC bus energy back into the AC supply for re-use elsewhere.

[DickDV]

One other thing. Be sure the control scheme will engage the mechanical brake when the drive faults. Otherwise, the load will drop to the bottom on drive fault.

[bikerdude]

DickDV,

So... will I need a "snubbing" resistor to disipate the excess DC buss voltage that wil be developed when ramping down to stop from 60hz in 1/2 second? If there is no resistor will there be a DC Buss Overvoltage fault? probably depends on the amount of overhaul from the load. Hmmm...
Thanks
BD

[DickDV]

The brake chopper part (that's the voltage sensing part) may or may not already be included in the drive. If not, you will need to add it plus the brake resistor.

There will be a DC buss Overvoltage Fault is that is not done, especially when stopping a descending load.

[bikerdude]

DickDV & Others,

Thanks for the help!
The drive was ordered with a DC brake option, but I didn't order a brake resistor. I'm learning more things about VFDs everyday with help from the forum members.
BD

[Clay B.]

Looking at your pictures that Palitizer is real similar to some I used to work on in a can plant. To keep from having to constantly compensate for the brake wear and pallet drift what I did was get an encoder splitter and send the encoder data back to the PLC. This allowed me to control my ramping directly from the PLC and gave me better position control in the PLC. Our cycle had to be less than 7 seconds and empty aluminum cans do not take jolts very well.

With what you have said you are ordering I would use dynamic braking. Just rememeber when you are setting it up. Check your motor temp as it runs and reduce your brake voltage to get the motor to stay at a reasonable temp. The downside to reduced DC voltage is that it takes longer to stop. Also instead of waiting till you get to zero speed to engage your mechanical brake try using a limiter somewhere above zero that can allow your mechanical brake to help out your dynamic and vise versa.

The times when you are not there the non PLC guy can adjust the spring on the mechanical brake to compensate for drift and you can stay home sleeping. A compound system is more complex but it gives you options if you work it right.

[bikerdude]

Clay,

From what I can tell from the manual, the IGBT brake chopper transistor should already be installed in the unit I ordered last week. I am currently trying to size/source a resistor. The chopper is directly connected to the DC buss I believe. It is always best to find a compromise between motor heat & brake wear. The motor (Baldor Vector) has a 120vac cooling fan attached to the tail end. There is also a 2 wire temperature cut-out installed in the windings, not too sure if I'll hook it up or not.
Thanks for the help everyone.
BD