Oscillating C axis when clamped.

[MazatrolMatrix]

C axis is vibrating on our Integrex when the brakes are on. You can even see it shaking and so it is pretty severe and it affects surface finish. When you have it unclamped it goes away though.

It is possible to make it go away temporarily by adjusting a parameter, but it returns soon enough.

Anybody have any experience with this?

 

[ViktorS]

Yes, but..

Do you mean M211 brake or M210 clamp? Just to clarify.

If M211 drag brake, the solution is actually hillariously simple. Grease the brake disk mounted behind the spindle.

The problem is that the brake gets too grabby with time and dust, introduces massive stick-slip. Some sticky grease solves this and is actually a recommended service point.

I have a habit of only using M211 when really needed to prolong the intervals. IE. when running low force chamfers etc. I leave the C axis with M212 instead.

 

[noname777]

yes. It makes sense, when your C -axis in servo mode, any sticky or dirty brakes will disturb C axis P.I.D control and turn the whole thing into mechanical oscillator.

 

[MazatrolMatrix]

I would believe I'm talking about M210, the breaks being clamped.

Typically when we bore a critical hole, we go to manual mode, and then into C-axis menu and switch softkey 'locked' to 'unlocked', then go over to auto mode and continue the cut from there. This causes the vibrations to completely disappear. This then only relies on the servos ability to keep it in place, right?

The Mazak tech guys couldn't resolve the issue and instead provided us with a parameter that we can change the value of to get it to disappear. Again though it comes back fast.

I'll look into greasing and cleaning the brakes. I read you comments again and it seems I speak of the complete reverse scenario though..

 

[noname777]

think of it as C-axis mode where Spindle goes into servo control mode and has its onw P.I.D parameters and filters and whatever. That is how C axis is operated. When you apply brakes for dragging function, C-axis milling as an example, you bring external mechanical disturbance for servo controller to deal with. Of cause it needs to be properly adjusted as a compromise in between brakes are applied or not.

 

[noname777]

u can also try to tune pressure reducing valve when it comes to brakes. U should have it shown in your hydraulic diagram

 

[ViktorS]

This causes the vibrations to completely disappear. This then only relies on the servos ability to keep it in place, right?

Yes, that is correct. Problem is, both for M211 and M210. The "Release all brakes" code is M212, which is probably given when you press unlock.

Try to verify wether it's M210 or M211 that is the problem, they require opposite solutions. Try putting it into C axis mode and then in MDI run M211. Jog the C axis around a bit, how does it behave? Then try M210, you can't jog it in this mode but you can see how it behaves.

When M210 clamp command is given, a timer of a couple of seconds start to count down. When it is done the C axis power is drastically reduced. This is probably to stop the C axis from fighting a solidly clamped spindle. This is why I think M211 is the problem, the C axis should not have enough power to jerk in M210.

Also, M211 is the one that always act like this in any C axis lathe, so the odds are it's this one this time too.

Long story short:
If it's jerking in M210, the brake is not grabbing enough and need MORE friction.

If it's jerky in M211, the brake is too grabby and needs LESS friction.

Curious, what kind of integrex? And what parameter have you been instructed to change?

 

[noname777]

I just want to mention that this JERKY behaviour can be observed on SERVO DIAGNOSTIC screen once C axis becomes servo axis as a DROOP. All u need to do is see the DROOP. If it is stable, CPU is controlling closed loop. Once it is not stable, CPU is working hard to maintain its PID characteristic. MITSUBISHI guys have absolutely everything available to you. God bless Mitsubishi

 

[MazatrolMatrix]

Yes, that is correct. Problem is, both for M211 and M210. The "Release all brakes" code is M212, which is probably given when you press unlock.

Try to verify wether it's M210 or M211 that is the problem, they require opposite solutions. Try putting it into C axis mode and then in MDI run M211. Jog the C axis around a bit, how does it behave? Then try M210, you can't jog it in this mode but you can see how it behaves.

When M210 clamp command is given, a timer of a couple of seconds start to count down. When it is done the C axis power is drastically reduced. This is probably to stop the C axis from fighting a solidly clamped spindle. This is why I think M211 is the problem, the C axis should not have enough power to jerk in M210.

Also, M211 is the one that always act like this in any C axis lathe, so the odds are it's this one this time too.

Long story short:
If it's jerking in M210, the brake is not grabbing enough and need MORE friction.

If it's jerky in M211, the brake is too grabby and needs LESS friction.

Curious, what kind of integrex? And what parameter have you been instructed to change?

Mitsubishi

I follow. Not on the DROOP thing though, I'm completely lost as for what you are talking about there.

I did a quick check today.. and I found it difficult to connect the issue to one specific 'mode'.

The operator was finished with his workpiece and had taken it out, the machine was in M210 mode when I got there. It hade not done any C(Y) axis interpolation as the part just had some holes. It was vibrating but with very low frequency (this can be low or very high depending on.. something).

So I did M212, again all vibration was gone. I put it then in M210, no vibration. However a small pulsation could be felt when putting my hand on the chuck.

Then I did M212 and then M211. Initially, just the same pulses could be felt, however after joging it around a bit, the vibrations came back and could clearly be felt when touching the chuck. But it seemed to turn smoothly, I could not feel any jerkiness when touching the spindle while rotating. The vibrations seemed to primarily be felt when idle. As to what initiated them is another thing of course.

I know for a fact though that it is vibrating while in M210 sometimes though, perhaps M211 can start it and then M210 can escalate it.

/////

The machine is a E410II machine.

I looked up the parameter, and a description of what the root of the problem is in a word document on the computer. The operator who has written it probably did not understand the technician completely as I find the description awkward. But here it is:

"The reason for the vibrations is that the C-axis overcompensates the goal value which causes the rotation of the axis to be bad"

"In order to correct it do as follows:
Diagnosis
Version
Right softkey
Type: 1131 and then Input
Spindle monitor
Parameters
Scroll to Parameter: SP71 and go to column SPDL 2. The setting range here is 1-999, although values over 500 is not recommended"

Thomas: - "I put in value 70 here and it seemed to work"
Thomas: "now I had to change it to 300 which seems to work "

 

[ViktorS]

Alright.

When the axis is moving there is no "stick-slip" effect, only "slip". So hunting is usually not seen except with really small increments in movement.

I stand by my first solution. Grease the disc and the problem will dissappear.
On a 200Y it takes about 5 min to remove the back cover and get access to the brake disc, not sure on E series. But try it, it is an easy and inexpensive fix. Before that is ruled out I see no point in trying anything else as it gets way more complicated after this.

PS.
The DROOP value is position error in encoder counts (I think). One can see how much error there is at any given time for any given axis.

 

[MazatrolMatrix]

Alright.

When the axis is moving there is no "stick-slip" effect, only "slip". So hunting is usually not seen except with really small increments in movement.

I stand by my first solution. Grease the disc and the problem will dissappear.
On a 200Y it takes about 5 min to remove the back cover and get access to the brake disc, not sure on E series. But try it, it is an easy and inexpensive fix. Before that is ruled out I see no point in trying anything else as it gets way more complicated after this.

PS.
The DROOP value is position error in encoder counts (I think). One can see how much error there is at any given time for any given axis.

It’s certainly worth a try of course, I have never done it before but I will look into it and let you know how it goes. In some ways I hope it isn’t that simple since I’d lose all hope in the tech guys then.. If I haven’t already.
Recently they tried to bill the company 10K$ for ’trying’ to fix a problem with another Integrex, which they ultimately were unable to do. We refuse to pay.

Thanks, I get it.

 

[noname777]

I dont really understand meaning of SP071:

SP071 - DIQM - Variable current limit during deceleration, lower limit value
Set this parameter to adjust the deceleration time by changing the current limit value during deceleration depending on the motor speed.
As shown below, set the lower limit rate of the current limit in SP071 (DIQM), and use with SP072(DIQN).
When DIQM is set to 100%, the standard current limit value in deceleration (TMLR) is applied.


But, the attached page is what they used to do to tune C axis in MDS-C1 generation.
I think MDS-D series has different set of dedicated parameters.