Reversing motor thats not completely stopped with older VFD?? Is this ok?

Hi guys wondering if its ok to start my motor in reverse if its still slowing down? I have a feeling this is probably fine on most newer VFDs but just want to be sure it wont damage my older Mitsubishi Freqrol Z300. See manual here: http://www.vfds.com/manuals/mitsubishi-z300-manual.pdf

Im working on a rigid tapping program for my retrofitted Fadal with Kflop control. I see some people wait for a signal from drive that it has stopped. Just wondering if this is really necessary? I was planning on adding a slight delay and then reversing, but just want to be sure i wont damage anything if i happened to switch it in reverse while its still rotating. Am i overthinking this? Should i just switch forward relay off and reverse relay on right away? Thanks in advance.

The "damage" issue has to do with the VFD protections. Reversing is a high current condition.

It may likely be possible, but does not buy you much time, and does definitely have the potential to cause an overcurrent trip, unless you can program a restart, or a "limit current but do not trip" behavior.

You can do it. It would be better to know what kind of VFD you have, and it's abilities. Do you have a breaking unit? Do you have a breaking resistor? What is the max...=> speed, waight of turned peice(inertia) you were trying to reverse?
Reversing a 200 pound piece at 1000 Rpm in two seconds is quite a bit different then reversing a 5 ounce piece at 60 rpm in two seconds.

I'm guessing you don't have a reversing clutch?

It's a Mitsubishi Z300, on a CNC (Fadal of some flavor), so............. He gave a link to the VFD manual.

you can use dc injection to stop the motor, then restart as normal.

page 51 or page 56 of the pdf provides the information needed and the caution provided that the dc injection braking will overheat the motor. the problem is a static dc voltage applied is not the optimal dc volts needed to stop the motor the quickest.

Usually you use regenerative braking, then turn the dc brake on.

The only way to speed up the reversing process is to find the cross over point between dc injection and ac injection at reverse direction, probably around 4hz or so. so when it has slowed down to say, 100 rpm for a 4 pole 60hz motor, you stop the dc injection and start the reverse ac frequency injection to accelerate the motor in the opposite direction. starting the motor while its still spinning in the wrong direction will cause the current to increase immediately upon turn on, you will need to drop the "boost volts" to prevent overcurrent within 50ms of turn on. this however depends on the reverse rpm at turn on, and perhaps the rotor's inertia.. so you might keep the boost volts at 20% or whatever it is, then figure out the maximum rpm you can change directions at without overcurrenging the drive.

Well sort-of.

It depends on the mass of drivetrain AND work, as Matt Matt mentions. It's a matter of energy, and power.

A heavy workpiece stores more energy at a given rpm than a light one, or the drivetrain alone. In order to reverse its rotation, you have to remove the energy it has, stopping it, and then put a similar energy back in, rotating in the opposite direction.

The drive has, as its only connection to the rotating mass, motor current. That is limited my the design of the VFD, which can provide only a certain current. Acceleration needs current above that needed to just spin the mass at a constant speed. Slowing is negative acceleration, so the VFD normally cannot decelerate the mass significantly faster than it can accelerate the mass.

DC braking sidesteps that problem, allowing as much or more braking than the maximum normal motor torque, but provides other ones. A fairly small DC current in the windings can produce a very large rotor current, which heats up the rotor. Since it is stopping, it then has no fan to cool it unless an external fan is provided, and a few such stops can very severely overheat it. Also, The rotor current is affected by rpm. DC braking is extremely violent if applied at higher speed, as well as heating the rotor more.

So, the VFD may be easily able to reverse the motor when unloaded, but may produce regular overcurrent faults, or time-current faults, if asked to stop the motor in a given time when the machine has a heavy workpiece in it (mostly important for lathes).

It's most common to do a regular decel until a low speed, then DC braking, DC braking is usually still effective at low rpm, despite what you are often told. Try it sometime, put DC through a winding of a small motor, and spin the rotor. It will have quite a surprising amount of drag, even with a low DC current.

DC braking does not apply to PMAC motors.

Figure that you cannot decel any faster than you can accel without an OC fault, and see if you can find a mode that will limit current without causing an OC fault.

dc injection braking is probably the easiest way to cut the time interval to reverse the spindle as it applies to this thread. but you need some way to determine how fast the spindle is moving and when to turn it on, even if its ok to turn it on while its still spinning other direction, why wait a static amount of time for all inertial loads?


i'm not sure what forward relay and reverse relay mean as it applies to OP's question. if the cnc control uses relays to provide voltage isolation between the vfd and the control, that's great, in fact the vfd may be able to be programed to handle the reverse command properly regardless of the rpm. i'll have to read the manual again later.

Thanks for the help guys. Its a Fadal VMC 4020 with cat40 spindle so no real heavy inertia load. Its running in a 2:1 reduction so i would think biggest inertia by far would be rotor itself. Its not a regenerative drive, it has a large external braking unit and 2 huge resistors. The relays are just used to turn on the CW/CCW signals to the drive. Motor does have an external fan. Machine will only be tapping holes occasionally so i wouldn't worry too much about heat issue in the motor. And were only talking about reversing from a relatively low speed. Max im hoping to use on this machine would only be about 1500 rpm (750 at spindle), and I may end up having to run slower still. Also the acceleration is set a little lower on this machine to reduce start up current.

I did some testing with it tonight and ran into issues with keeping Z axis synced while adding a delay to wait to stop. So i rolled the dice and went from forward to reverse right away and it seems to be fine. Nice and smooth and no obvious current surge. The machine light dims a little when ramping to high rpm and this was not noticeable when reversing from a low speed.

Think im safe? Or should I try and find a way to verify its stopped?

By regenerative braking i mean the vfd is pulling power out of the motor and sending it to the vfd. i don't care if it gets back to the mains or not, that's what the brake resistors are for. dc injection braking wastes all of the energy stored in the mechanical inertia of the motor and dumps it into the rotor of the motor. this is a thermal problem for high speed operations, or very high inertia situations. if you are below around 25% of the motor's nominal rpm, there simply isn't enough energy stored to be a problem, provided the interia of the system is limited to just the motor.


anyhow, it seems the problem you are going to run into is just what you discovered, the control had a problem keeping the z axis synced.

i can speculate on why that happened given the information you've already provided but i don't think it would help you.

if you can hook up a linear encoder to the Z axis and feed the output from the spindle encoder into a computer, you could easily plot how much the error was and after accounting for backlash and stiffness in the entire machine structure, you could determine if it would break a tap or not. --or you can try and figure out what settings don't result in the synchronization error and determine this yourself. but depending on how large the tap is, the additional load will change the situation significantly.

You may be best to try to do it direct as you mention.

I was not considering the synch issue of tapping, if the Z has to know speed so as to stay "with" the tap. That's actually a hard problem, that I would have thought would be solved differently, by simply sensing the rpm directly, and following it with the known tap pitch. (or by not following but letting the tap drive Z once it "bites in"). I do not know too much about some details of CNC. If the machine sensed spindle rpm, it should have been able to follow nearly anything.

But what you describe for your issue is the simplest case, where the only adder to the spindle mass is the tool, which I suppose could be a fairly chunky face mill, but what that really adds to the inertia I do not know. At least you do not have any heavy workpiece to deal with.

And with tapping, there is a minimum of mass, so your approach seems good. Put the biggest toolholder and tap on it, and if it works OK, stop worrying and use those settings. Another variable in the deal is that there is no consideration of cutting forces, which also draw off energy. If the tap is in material and turning there is friction

mmurray70 said:

Im working on a rigid tapping program ...

Click to expand...

Tension - compression holders will give you a little extra safety, even with rigid tapping.