Mitsubishi A800 160kW VFD Higher Than Expected Current

Long time reader first time poster, I will try to keep it short but there is a lot of variables here. I have a Mitsu A846 running a 1000 ton press. I have it set to real sensorless vector control. It is a 160kW 4 poll motor 460V and 240A. We are replacing some Powerflex 700s drives and the one that is still installed is running a peak current of 190A. Yet for some reason the mitsu is peaking at 260A with ruffly the same load. I would expect them to be about the same I have the FLA set at 260 and its kissing that when the press cycles. I did a static tune (offlline) since we are using the motor control we are. I will run a rotating tune when I get a chance. Just hoping someone on here has seen this before.

Oh we have two of them installed both running higher than expected current so no mechanical issues causing more load.

Are you reading the current from the drive display or a hand held meter?

Hand held meter. It is the same using FR configurator software.

Assuming you have the accurate motor nameplate data entered in both drives, I would look first at your speed loop tuning. On stamping presses, its probably best not to be too tight on the tuning, letting the speed run up a bit on the downstroke and run down a bit on the upstroke.

If you try to tune the speed loop too tight, you are needlessly asking the motor to hold constant speed over heavy motoring overload and heavy braking overload. On a press, there is no particular need for this tight speed regulation and just makes the motor work harder than is necessary (higher peak amps).

Let us know what you find, please. Crankshaft presses are deceptively challenging applications and after having done dozens in the past 15 years, its always interesting to see what happens with other brands of drives. I've always used ABB so my experience is a bit narrow regarding other brands.

Thanks your pretty close!!

Thanks for the reply. You are partially right here, well your right. However there are some variables that do not allow me to let the speed fluctuate.
Its a tandem press line with robots doing the transferring. I have a fault set if the speed fluctuates more than about half a SPM or 2 Hz either way. If you play with the params that control "response time" and open up the fluctuation fault window the robots will run into the side of the press lol trust me on that one!

I could do this and slow the machine down but in the manufacturing world as you know that's a no go.

The manufacturer recommended I play with some of the Gain settings like torque control gain. I had a min to play with it today and was able to head a little in the right direction but to much she starts fluctuating speed. Tomorrow I will hopefully have it all day. I will post my results when I get a chance.


Thanks again.

What did they have on it before?

I agree with DickDV, as I have done plenty of presses of different types with a few different manufacturers' drives.

Regulating the Stroke RPM vs. the Motor RPM is a game that can be frustrating. The last thing I want to do is waste the energy that was put into the flywheel getting it back up to the top of the stroke, so let it go a little over-speed on the way down and droop a little on the way up. The key is the Stroke RPM being consistent.

You say the robots crash into the side of the press if the speed is not tight enough? Aren't they using position-based control for the 'iron hand'? Using strictly timing is scary.

One of the tricks I've come to use is droop-mode, or negative slip. Some drives have it, some don't. But if it has it, you are still in control of the motor within the slip settings, and not just beating the heck out of it trying to regulate the speed every mS Say for example you have -1% slip in the settings, you'll pick up 18rpm on the down-stroke, but lose it on the up-stroke... Should come out even.

Food for thought!

One issue might be that the PF700 has ways to do this for you automatically on eccentric loads like presses, maybe the Mitzi doesn't, or how they do it is different and not as simple (from a setup aspect) as it was on the PowerFlex.

jraef said:

One issue might be that the PF700 has ways to do this for you automatically on eccentric loads like presses...

Click to expand...

I wasn't aware of that. What setting do they have, a 'droop mode'? I'd like to see what they do.

Gene Bond said:

I wasn't aware of that. What setting do they have, a 'droop mode'? I'd like to see what they do.

Click to expand...

In the PF7 class drives, it's simply referred to as the "Bus Regulation" feature that can be turned in or off, and if turned on there is an adjustment for the gain on it. So with that enabled as the cyclical load approaches regeneration, it backs off the frequency so that regen doesn't take place, then automatically ramps back into it as the load needs motoring again. If you download what's called the "Reference manual" for the PF700 and search for Bus Regulation you can read details on it. With some Asian drives I've worked with there is often a similar feature, but you need to know the cycle rate and configure the drive to recognize that to respond appropriately. It's not as easy to do, especially if your load varies. It can stil be done, it just takes a little more investigation and observation.

Gottcha. I believe I've seen it called 'Stall Prevention on Decel', or something like that.

I think we are all skinning the cat here.

So you guys are on the cause and effect. I have made a few changes I set up easy gain tuning for load inertia control you can look up parameter 819 set to 2 in the Mitsubishi FR-A800 detailed instruction manual. It reads like it sets a control gain based on the load inertia and response time. This brought current down quite a bit without having the speed fluctuate like I thought it would. I am definitely learning on this one. Used to setting 10 params max and it runs great haha. I will be having some support on this in the coming weeks I will try to let you guys know what else they do.


Replacing Powerflex 700S. External braking to a resistor, No encoder just speed ref A over device net if that tells you anything about what we took out. Both great drives same bird different feathers I suppose. The PF750 series are nice too.

Having Q series PLCs I get more features out of the A800 with less and it for sure fits the next CPU upgrade.



Thanks

Update, I got the current down but for some reason two die sets where still high on the downstroke. Found the counter balance settings where high due to improper new die set up.