PDA

View Full Version : Need a lesson in dynamic braking



Highpower
02-11-2011, 09:20 AM
Just installed a braking resistor for the lathe VFD, and have been doing some experimenting with the "fine tuning". I was trying to get an idea of what the limits are and toying with it to see what it will do. I found out that it will come to a dead stop consistently from about 1300 (spindle) RPM in 0.8 seconds with no issues.

Below that, it will decel for about 0.6 seconds and then the braking kicks out and it free wheels for another couple of seconds until stopping. It is NOT tripping the inverter nor setting any over voltage/current errors. I'm just trying to understand the "why" because I don't have any parameters set for the motor to freewheel. And I am not using DC braking either - previous posts have scared me away from that stuff. ;)

This is just out of curiosity since a 2 second decel seems to be a good level for my lathe. That and the fact that after a couple of 0.8 second stops my new Fenner drive belt is letting me know it's not happy about my experimenting. (Squeek... squeek... squeek...) :o

Still it would be nice to fully understand the reason this happens.
Thanks.

squirrel
02-11-2011, 10:09 AM
Becarefull when you have a lot of mass on the spindle nose, if you have a screw on chuck be carefull, it might unscrew itself. Extremely short brake times are intended for emergency situations and special applications with components designed for it.

Black_Moons
02-11-2011, 11:06 AM
Maybe the motor is actualy breaking free of the rotating magnetic field (Wayyy overtorque for the current)? So its just cogging insted of braking, And then once the 0.6 seconds is over, the VFD stops driving the motor because it thinks the motor has stoped.

I seem to recall dynamic braking is only like 150% normal operating torque. But I could be wrong on that as my memory is a little vauge on the subject.

Also note that while it might work at 0.8 seconds now, Try chucking a 50lb chunk of steel in the lathe and spin it up to 1000rpm and stoping it...

Hopefuly, you don't figure this out while threading towards a shoulder. (of course, why you'd be threading at 1000rpm on a 50lb work peice..)

A slightly more realistic situation might be changing a 3jaw chuck for a slightly bigger 4jaw.

Highpower
02-11-2011, 12:54 PM
Nope, no threaded chuck. (D1-4 spindle)

Don't plan on spinning any heavy work pieces at high speed either.

Threading to a shoulder at SLOW speeds - yes. Thus my interest in trying to obtain a quick stop. I don't have a mechanical brake and I don't want to use a 3-wire control.

I only used spinning the empty chuck up to create some inertia to get an idea of the braking torque available. I don't have any idea of how to test this at slow speeds. New territory for me. :o

squirrel
02-11-2011, 01:05 PM
Nope, no threaded chuck. (D1-4 spindle)

Don't plan on spinning any heavy work pieces at high speed either.

Threading to a shoulder at SLOW speeds - yes. Thus my interest in trying to obtain a quick stop. I don't have a mechanical brake and I don't want to use a 3-wire control.

I only used spinning the empty chuck up to create some inertia to get an idea of the braking torque available. I don't have any idea of how to test this at slow speeds. New territory for me. :o
If I am reading this correctly, you need to stop it quick before crashing while threading. Yes, that is a challange, myself, install a precision limit switch on the bed and when the carriage makes contact it shuts down the drive. You should be able to monitor the status lights on the drive while adjusting the switch by hand. You will pull your hair out trying to calculate this problem, its much easier solved in real life.

RobbieKnobbie
02-11-2011, 01:30 PM
Many drives have multiple decel ramps available... so if you hit the E stop you get a very short stopping time, and if you hit a 'normal' stop you set it to come to a more gradual (less taxing) stop.

Best practice I've found is to set your regular decel to a comfortable, controlled time and leave it.

For the E stop, you don't want the brake to trip out and have it coast down like you've seen happen so far, so set your decel like ABS brakes... set the decel time for the quickest possible stop under the worst possible conditions... highest speed and greatest rotating mass. Cause if there's ever an emergency, it'll invariably happen when you're spinning a big chunk of metal really fast and not when you're threading a 1/4-20 screw!

Black_Moons
02-11-2011, 02:01 PM
Well, Depends. At 'slow speed', but high gear (low hz, speed reduced by VFD) you have exactly the same torque to overcome as at high speed, But your starting at less RPM's so you have less total power to disipate, and the motor rotar will add very little momentium, But you will have limited torque avilable to stop the chuck.

At slow speed, but low gear (normal hz, speed reduced by gearbox) you only have the momentium of the motor rotar (and some of the gears) working against you, the chuck is geared down and hence requires much less torque to slow down. (Remember, your motor has torque limits), But you have much more momentium in the motor rotar to slow down.

So depending on the weight of the motor's rotar and the chuck, You may find a balance for best stoping speed at low RPM's beween VFD speed reduction and geared speed reduction. (of course, VFD speed reduction also reduces max HP avilable)

macona
02-11-2011, 02:59 PM
A lot of VFDs will cut off at the very end. Also as you slow down at a point it will not generate anything to dynamically brake. At this point you apply DC braking which will bring it to a stop. The drive should have a mode where it dynamic brakes then switches to dc braking. This is fine for stopping but dont use it to hold the motor, that will eventually overheat it.

Forrest Addy
02-11-2011, 03:24 PM
VFD's in my experience proportion the accel/decel to the speed ref setting. If set to 30 Hz and the decel parameter is set to 2 seconds at 60 hz the motor will stop in 1 sec and so on.

When you want quick stops at say 80 RPM, gear for 500 and set the speed ref pot for 9.6 Hz and the motor will decel to a stop in 1/3 second - almost instantly by perception. The motor at 9.6 Hz will have about 7% of its kinetic energy of full speed and the inertia load et for 500 even a smaller pecentage of its top RPM.

lakeside53
02-11-2011, 05:46 PM
what Macona said... DC inecjtion can be tricky to get right or just to find an acceptable balance. For quick braking you want it on only for a short time, but for longer braking the DC needs to be on longer to be effective. If the braking times are variable.. (see below) then err on the "too little" side.

Also... from higher speed or with higher masses, make sure you have enough time to stop without tripping out from over-voltage on the DC buss. Some VFD's (Hitachi for example) have a parameter to set that give "best effort" braking - i.e it holds the DC buss at or below a specified voltage and decelerates based on that. Mills require this infrequently, but my lathe uses it a lot. At low speed or low mass, the braking is quick. At higher speeds and/or masses, it can take many seconds more, sometime 10-15 if a big mass at 2500 rpm. Much better to allow the "continue" than tripping out.

Highpower
02-11-2011, 09:57 PM
VFD's in my experience proportion the accel/decel to the speed ref setting. If set to 30 Hz and the decel parameter is set to 2 seconds at 60 hz the motor will stop in 1 sec and so on.

When you want quick stops at say 80 RPM, gear for 500 and set the speed ref pot for 9.6 Hz and the motor will decel to a stop in 1/3 second - almost instantly by perception. The motor at 9.6 Hz will have about 7% of its kinetic energy of full speed and the inertia load et for 500 even a smaller percentage of its top RPM. OK, I did not realize there was a correlation between the decel rate and the frequency that the motor is running at. My limited "testing" was all done at 60 Hz in high gearing. I see now that I should have done it in a way that matches my expected usage, ie. low gear and a lower frequency since I'm aiming for a slower spindle speed while threading. That is the type of education I was hoping for - thank you.


A lot of VFDs will cut off at the very end. Also as you slow down at a point it will not generate anything to dynamically brake. That is what was puzzling me because the VFD didn't trip or throw errors, and I could not find any parameters set that would make it free wheel to a stop after braking. DC braking and things like low speed torque boost are areas I'm not ready to get into just yet. I'm quite certain I would get the timing wrong and cook something. Those waters are too shallow for me to blindly dive into. :o

lakeside: You're right. I got a couple over-voltage trips previously which prompted me to bite the bullet and get a braking resistor. Have not had a single trip since. :)

As of now I'm just pickled tink with the quiet smoothness of the new 3 phase, and stopping times that don't require a calendar anymore. :D

J Tiers
02-11-2011, 10:34 PM
ANY electrical braking is speed-dependent. Even DC braking. For vector drives this is different, as mentioned later.

For standard decel, the braking effect comes from slowing the drive frequency. The drive forces the motor to "generate" instead of drawing power, and that power is transferred to the DC bus by the inductive kick from the windings when the drive IGBTs turn off.

At low speeds, this is not very effective, and often the drive will stop doing that for some percentage of the total decel time. That MAY be settable, if you want, read the settings section of the manual very carefully to find out.

DC braking is more effective at lower speeds, but even it needs some speed to create any braking effect. Many drives shift to DC brake at the end of decel. You may notice this as a 'clunk" at the end of decel.

Vector drives can sense the rotor turning, and can do a pretty good job of holding the rotor steady. Better than DC braking, because they can actually drive against the motion of the rotor, if set up that way.

To get a better stop, you may be able to choose and set "DC brake at end of decel".

But since the electrical braking is related to electrical drive, you usually can't get braking a lot better than accel.

lakeside53
02-11-2011, 10:49 PM
DC injection braking will lock the rotor if you let it (and it does if the brake period exceeds actual - not programmed - the decel time). Why do you say it needs some speed for it to be effective?

I set mine to start at about 33% - it takes over the last part of the cycle when regenerative braking is falling off.

alsinaj
02-11-2011, 10:49 PM
Don't want to hijack the thread, but has anyone seen mechanical braking used, and if so, how was it set up? No limit to how fast you can stop mechanically (so long as the machine is bolted to the floor and can't tip over).

lakeside53
02-11-2011, 10:54 PM
Many times...

My big lathe has an electromechnical brake that activates every time the spindle power is cut. Very nice...

Several lathes I've used have a foot pedal to a brake device (pads, disk etc) that stops the spindle real fast.

J Tiers
02-11-2011, 11:50 PM
DC injection braking will lock the rotor if you let it (and it does if the brake period exceeds actual - not programmed - the decel time). Why do you say it needs some speed for it to be effective?


I said that because it is true.

You may be thinking of a vector drive. They CAN virtually lock the rotor in some modes.

The way DC injection works is to turn the motor into a generator. The DC in the motor windings creates a field which, IF THE ROTOR IS TURNING will induce a current in the rotor bars. The resulting rotor field interacts with the DC field and produces a counter-torque that slows the rotor.

It CANNOT "lock" the rotor, because there is zero braking unless the rotor turns. A true mechanical brake locks the rotor so it cannot turn until a certain torque is input, that overcomes the brake. The mechanical brake allows no "creep". This is not true of DC braking.

With DC braking, you can turn the rotor slowly with little counter torque. As you turn it faster, counter torque goes up, the actual amount depending on how much DC is "injected". There can be a considerable amount of counter-torque, with heavy DC injection, but it is limited, and it cannot prevent "creep".

Brake counter-torque is limited by electrical parameters, and usually cannot exceed the maximum motor "driving" torque.

djc
02-12-2011, 12:39 AM
...has anyone seen mechanical braking used, and if so, how was it set up?.

Some food for thought.

http://www.youtube.com/watch?v=glYDe94hxb0

lakeside53
02-12-2011, 09:40 PM
Jerry - I have a volt/hertz (not vector) Hitachi drive. With DC injection on, the rotor is locked after rotation stops for as long as I program it for. I can't rotate the chuck by hand until that period times out. I try not to do that - real bad for the motor.. but...

Highpower
02-12-2011, 09:55 PM
I set mine to start at about 33% - it takes over the last part of the cycle when regenerative braking is falling off.So how do you know when the regenerative ends, in order to know when to start your DC braking?

J Tiers
02-12-2011, 10:42 PM
Jerry - I have a volt/hertz (not vector) Hitachi drive. With DC injection on, the rotor is locked after rotation stops for as long as I program it for. I can't rotate the chuck by hand until that period times out. I try not to do that - real bad for the motor.. but...

Doesn't matter.

DC can't do anything without some motor rotation to generate rotor current. With sufficient DC, it can *seem* pretty stiff, but torque is limited, and it CAN creep, even with heavy injection.

If DC is equal to motor current, brake torque will be equal to motor torque. If you can't hold the chuck against the motor torque, then you shouldn't expect to turn it against DC, especially with the belt/gear ratio from motor to chuck working against you.

What you are seeing is simply a very similar effect to the generation of torque in the motor due to "slip" when it is running. It's powerful, but not unstoppable. Braking can be fast, but not instant.

For instance, the reason the "sawstop" uses the aluminum "cartridge" mechanical "jam-brake" and not electronic braking is because electronic braking is torque limited and works too slowly.

lakeside53
02-13-2011, 12:32 AM
For all intents, the rotor is locked. OK, let's just call it "stopped" and "held". I can't turn the chuck by "normal" hand, but I haven't tried to force it. Why do I know this? If the chuck stops with the key hole in the wrong position, I have to wait the few seconds (stopping from slow speed or with low mass) for it to release before I can get the key in place.

I find DC injection a pain on a lathe. Rough on the gearbox (the big clunk when it comes on is both from the motor and the transmission) and hard to find a balance between positive stop and inertial load. If it ends late, the chuck is "held"; if early the rotation continues somewhat. Maybe there's some smarter VFDs that can stop DC injection at zero RPM, but those I deal with don't.

lakeside53
02-13-2011, 12:37 AM
So how do you know when the regenerative ends, in order to know when to start your DC braking?

You don't. There's a parameter or two to set that starts the DC injection at some thresh hold. I set mine 30-50%, but you can go higher. The problem is overheating of the motor, and you need to watch the VFD duty cycle. Some VFD's will track this and stop injection if you exceed certain a (programmable) amounts.

Highpower
02-13-2011, 09:50 AM
To get a better stop, you may be able to choose and set "DC brake at end of decel".
If only they made it that simple... :(

http://img.photobucket.com/albums/v82/Highpwr/VFD/DC_braking-2.jpg