VFD not decelerating as expected. 4

[itsmoked]

I'm using a MA7200 to run a lathe motor. The original motor was a 3hp. This one is a 5hp VFD motor. I have everything working fine except the deceleration is anemic.

I've set accel and decel to 1 second. The motor ramps to 1720 in one second but refuses to decel in less than about 4 seconds. I have a mondo load resistor that doesn't rise above room temperature (which is saying a lot since it's ^*#@^%! 45F).

Running it in sensorless vector.

I've tried "Full Range" DC injection braking with only trivial improvement.

When STOP is commanded the current changes from about 5A at full speed with no load, to 7A on its leisurely stroll to zero.

Conversely when I switch instantly from full speed forward to full speed reverse the VFD dumps 30A into the motor and has it running full speed reverse in about a second.

I'm sure it's one of those cases where some quasi-related parameter is limiting the maximum stop current but I can't find it after more than an hour of weeding thru the manual and trying everything even remotely suspicious.

BTW: Never once have I had a single fault so it's not like it physically can't decelerate as fast as I'm asking it to.

Here's the: Manual

Keith Cress
kcress -

http://www.flaminsystems.com

 

[rhatcher]

Did you set Sn-15, stall prevention selection during deceleration, = 0 (disabled)? When Sn-15 = 1 (enabled), the deceleration time is extended to prevent DC bus overvoltage. This sounds like what you are experiencing. Also, if you have a DC ammeter you may want to check to see if the braking resistor is active during decel. Alternatively, I guess you could use a voltmeter since the presence of voltage on the resistor implies current flow and braking. Finally, note that the braking torque is limited to 126%.

 

[jraef]

Keith,
If you want Dynamic Braking, Sn-06 must be set to "0", Deceleration to Stop. By setting it to Full Range DCIB, it will ignore the Dynamic Braking braking transistor and resistors, it just pumps DC into the motor. Not good. DCIB really should only be used at the END of the braking cycle. Dynamic Braking uses the kinetic energy in the load to regenerate through the drive into the resistors. But Dynamic Braking suffers from the Law of Diminishing Returns in that as there is less and less kinetic energy left in the load, there is less and less braking power, so it can never "finish" the job. So you set the DCIB to come on at the very end, like the last 10% of speed.

But when you set Sn-06 to 0, you must ALSO set Sn-15 to 0 as well. This is the Stall Prevention on Decel. The factory default setting is "1", which has it turned on. This will interfere with your Dynamic Braking.


"Will work for salami"

 

[ScottyUK]

How is the chuck mounted? If it is a cam-lock or similar then no problem, but if it is a threaded type then make sure that it doesn't unscrew itself. As far as I know this type of chuck mounting is limited to small training and hobby machines - I've never seen it on an industrial lathe.

 

[itsmoked]

Wow! Such great answers.

I read it as DC injection wasn't as fast as deceleration but since decel wasn't cutting it, I finally gave it a desperation shot. I saw how they seemed to show the DCI only applying at low-end speed but didn't understand why. Thanks for the details Jeff.

I think stall prevention is the ONLY thing I didn't try. YES! I will do that and get back to you all.

Scotty, I too was really concerned about the spindle unscrewing myself but the lathe is a large one and the owner assured me the lathe stopping almost instantly with the old DC(?) motor.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[rhatcher]

If you have not tried deceleration stall prevention, Sn-15, then I'm betting that this is the problem. Also, as you probably know, if you trip on high DC bus voltage (main circuit overvoltage is the term used in the manual), then you need to incrementally increase the decel time until you can reliably stop without tripping. Good luck and let us know what happens!

 

[itsmoked]

No joy.

Stall prevention made zero difference to this issue.

One thing I did notice is that when you dial the speed knob up or down the motor responds to the knob according to the accel/decel setting. I can even hear the dump relay switching the resistor in and out. It just ignores the decel when STOP is commanded.


Keith Cress
kcress -

http://www.flaminsystems.com

 

[John2025]

Any chance the goofy thing uses Bn2 for running changes and Bn4 for stopping changes. I don't see any such thing and doesn't sound right, but do you have you tried setting them both to 1 sec?

 

[jraef]

Keith,
Another word of advice on my experiences with VFD programming.

Sometimes when I run up against a problem and I do a lot of experimenting, I end up creating conflicts by forgetting to reset one of the experimented options, which then gets in the way of the real solution. I suggest doing a full reset to factory defaults and starting over.


"Will work for salami"

 

[itsmoked]

Any chance the goofy thing uses Bn2 for running changes and Bn4 for stopping changes. I don't see any such thing and doesn't sound right, but do you have you tried setting them both to 1 sec?

Oh yeah, tried that several times.

Keith,
Another word of advice on my experiences with VFD programming.

Sometimes when I run up against a problem and I do a lot of experimenting, I end up creating conflicts by forgetting to reset one of the experimented options, which then gets in the way of the real solution. I suggest doing a full reset to factory defaults and starting over.

!!! I was starting to think that myself and almost did it. He told me when he started cutting the spindle slowed about 200 RPM. So I turned on the PID and suddenly the drive no longer cared about the speed pot. Turned out the drive decided to switch from 0~10V input to the 4~20mA input because PID was activated! Thanks for the push. I'll do it.


Maybe someone can explain sn-15 logic to me. I've read that paragraph about 10 times. It seems wrong.
It's on the bottom of page 3-50.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[John2025]

The description of Sn15 isn't very clear, is it?. Are they saying that with Sn15=1, every time the drive gets to +20%, it runs some undefined amount of time at its current speed, allowing the bus voltage to drop back down? That's what the steps in the graph look like to me. But I still don't see why it would be different when go directly to reverse.

 

[jraef]

In general, Stall Prevention is a tool that overrides your speed command in order to keep the motor running by sacrificing speed control in favor of limiting motor power, and by inference, current. So if your motor begins to stall, the increase in current demand from it as a result of the increase in slip is seen by the drive. The drive reacts by changing (usually dropping) the output frequency until the current stabilizes, regardless of what the speed reference is from the keypad or control system.

So if you have that enabled when you want Dynamic Braking, the drive might be incapable of telling the DIRECTION of the current flow and may artificially INCREASE the speed reference to attain the same net effect. But that's not what you are after when braking, so instead, it appears to do little. Just like what Keith was observing.


"Will work for salami"

 

[itsmoked]

Just got off the phone with the Westinghouse VFD factory and after running thru all this with the engineer he said, "You need to turn OFF stall prevention if you want to use the brake resistor to stop the motor more quickly". When I discussed the confusing aspect of the manual with respect to Stall Prevention, his response is that Stall Prevention allows extension of a Decel period whereas turning OFF stall prevention doesn't allow period extension. Which could otherwise lead to DC Link Over Voltage we've all, no doubt, seen.

I thought I'd tried turning off stall prevention but since it defaults to ON I may not have turned it off ever.

Hopefully I can try it in an hour or so.

Stay turned.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[itsmoked]

Finally got it.

Had to turn the Stall Prevention OFF.
Turn the DC Braking OFF.
Turn Decel ON.
Turn Run Time Stall prevention OFF.
Turn Accel Stall prevention OFF.

Now when I hit STOP the lathe spindle screeches to a stop amidst 32A of drive current and the braking resistor warms up.

Now happy I immediately ran into a new shock...

When we start cutting metal the spindle slows about 200RPM. PID to the rescue? Not so fast.

Here this thing is a sensorless vector drive and it can't run PID on its own speed. Can you even fathom that? Turns out you have to feed this sensorless vector drive a feedback signal from a sensor, if you want it to do PID. Remarkable. Astounding. Moronic.

They offer Voltage analog IN, or 4-20mA analog IN, or pulse generator IN as the process value for PID, but not it's own sensorless speed derived number.

I think they allow you to output the speed on an Analog OUTPUT. I wonder if I can feed that back in one of the Analog INPUTs as the process value....

Probably too slow, too delayed.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ScottyUK]

Is the drive still tuned to the motor if you have done a full reset? My lathe runs on a sensorless drive and the transition from no load to a heavy cut in steel is undetectable. My drive is an old A-B 1336 Plus so it's hardly cutting edge any more; yours should arguably perform better given the improvements in drive algorithms.

Is the PID loop not intended for use with an external sensor controlling, say, pump discharge pressure rather than shaft speed?

 

[Skogsgurra]

Yes, Scotty. I have yet to see a drive where the PID is supposed to handle speed. That is usually very well handled by the speed controller in the drive and it should just work if you do the autotuning and parametrization by the book. P(I(D)) is for air pressure, temperature, water level and such things.

Smoked, start over again. From scratch. Attention to details. Book in hand. Do as it says and do not Ass U Me anything (thanks Jeff).

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[itsmoked]

Scotty;
I spent about an hour trying to "RESET" to default settings. As far as I can tell that is not an option. If it can be done I don't see it anywhere in any manual. I finally went through the torture of looking at every single parameter.

There is slip compensation GAIN. I turned that up hoping to 'rigidify' the droop a little but was not able to test it by turning some metal today.

Skogs;
It does seem they want to use the PID for "flow or pressure" type inputs.

I did run the AUTOTUNE immediately after setting "Vector" mode. I was unable to disconnect the motor from the single shaft gear box but did remove the belts feeding everything else, head stock, feeds, etc. So during tuning there was the gear box shaft and the output sheave ~4".

Is ASR something I should be trying? No where can I find what it is even. That and how it's mixed in with PID isn't helping.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Skogsgurra]

The autotune usually does two things: 1. it measures motor parameters like leakage inductance, stator and rotor resistance and such things. 2. it measures inertia of the load and sets P and I in the speed controller (not the PID for external signals).

Let the belts stay in place if you want the speed controller to be optimized for the actual application. If only motor (and reduction gear) are connected, you will usually get a much too low P. Sometimes as low as 1 or 2 while an optimized drive has 10 or 20, sometimes more. That does a lot to the "crispness" of the speed loop. A speed loop with too little P does not react well to a load increase - it "dives" and recovers slowly while a well optimized speed loop barely notices a load change.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[rhatcher]

Keith,
I'm glad you got the decel stall prevention (sn-15) turned off. I had recommended this back on the 12th along with an explanation of why. I guess you missed that post.

Gunnar is correct that tuning the drive to the motor parameters should improve speed regulation from no load to full load. Although he offers a good explanation of why autotune should be performed with the motor coupled to the load, my admittedly limited experience is that the motor should be uncoupled. The drive manual also states this. That being said, Gunnar may be right. He is, after all, a pretty smart guy.

So, if the results of autotuning with the gearbox and sheave attached are less than satisfactory, it would probably be easier to install the belts (load) for the next autotune per Gunnar's recommendation than to disconnect everything. If this doesn't work, you will have to uncouple the motor completely per the manual to perform the autotune. Regardless of which method works, coupled or uncoupled, a proper autotune should improve the speed regulation. The details and instructions for the autotune function can be found on page app-9.

 

[itsmoked]

Interesting Skogs. They always seem to want the load gone but it always made little sense to me. Why not tune the systems instead of the motor?

Disconnecting the motor from the load would take several hours.

The tuning I ran with just the gear box, set all those various motor values. I've written them down so I can put them back if things go really sour while tuning the entire system.

Hey rhatcher! I didn't miss your correct directions! Thanks for them. Got any idea what "ASR" is referring to?

Keith Cress
kcress -

http://www.flaminsystems.com