[ckirchen]

I have a 2011 TM-2 that has a problem and I think it has to do with the servo tuning. When it is cutting in certain directions, it chatters and squeals like crazy.

The most telling test I've done so far is to mill a large circular boss. Imagine a clock, with 12:00-6:00 aligned with the Y axis and 9:00-3:00 aligned with the X axis. The cut is stable and smooth for the entire circumference, except from 4:00 to 5:00 and from 10:00 to 11:00. It does not matter what type of tool I'm using (coated, uncoated, carbide, HSS), what type of tool holder (collet or end mill holder), or what material I'm cutting (brass, aluminum, mild steel, alloy steel); the chattering occurs in those two places.

I've had the way covers off and confirmed that nothing is loose and that there is little to no backlash.

Using a tenths indicator, I've measured from the machine casting to:

• the end of the screws (to check for axial play);
• the threads of the ball screw
• the body of the ball nut assembly
• the table

In all cases, I could not discern any backlash. It's not to say that there isn't any, it's just that with the tenths indicator I have, I couldn't detect any.

All items were checked by jogging at the controller and by pulling and pushing on the table (both by hand and with a 2" x 4").

While I had the motors off, I did notice that I could easily turn the output shaft by hand by about 3° to 5°. The motor tried to get back into position, but it wasn't trying very hard. I understand that's how a servo drive works, it has a control position and it is trying to keep the servo as close as possible to it. I've worked on other mills and the servos are so stiff that you can't even get them to rotate a fraction of a degree.

On the Y axis, the end of the ball screw that is closest to the front is exposed. When I physically push or or pull on the table, the screw rotates by the same 3° to 5° and is slow to return to position when I let go.

My thinking is that the servos are not tuned properly (i.e. not aggressive enough) and at those two positions (4:00 to 5:00 and from 10:00 to 11:00), the lack of motor torque/stiffness combined with the cutting direction, causes the table to push away momentarily and then bounce back, and chatter occurs.

When I compare the parameters from the TM to my other Haas machines (two lathes and a mill), the P/I/D gains on the TM are much lower.

I am wondering if anyone has any info on turning these drives. I have successfully tuned servos drives before, but I don't know the specifics of the Haas controller (such as the maximum and minimum values for each gain is). I don't want to enter random values only to enter a number that is out of the range and corrupt the parameter file and brick the controller.

I've talked to my local HFO and they don't have any tech documents that they can provide. They've confirmed that the current parameters match the factory parameters, but I have a feeling the current parameters are initial defaults that are there only to get the machine up and running at the factory.

The parameters on the TM are as follows:
2 X P GAIN = 4
3 X D GAIN = 500
4 X I GAIN = 32
16 Y P GAIN = 4
17 Y D GAIN = 500
18 Y I GAIN = 32
30 Z P GAIN = 4
31 Z D GAIN = 500
32 Z I GAIN = 64

For reference, the parameters on my 2007 MM are:
2 X P GAIN = 32
3 X D GAIN = 4000
4 X I GAIN = 64
16 Y P GAIN = 32
17 Y D GAIN = 4000
18 Y I GAIN = 64
30 Z P GAIN = 32
31 Z D GAIN = 4000
32 Z I GAIN = 64

And on my 2009 TL-2, they are:
2 X P GAIN = 48
3 X D GAIN = 4000
4 X I GAIN = 64
30 Z P GAIN = 24
31 Z D GAIN = 3000
32 Z I GAIN = 64

Even if anyone else out there has a 2011ish TM-2, I would love to know what the above parameters are on your machine.

Thanks for your time,
Chris

[ckirchen]

I should add that when I grab the red X hand wheel on my TL-2, which is connected directly to the screw and offers a good deal of leverage, the servo is very stiff and I have to try very hard to get any rotation.

On the TM, I can grab the ball screw by hand and easily rotate it at least 5° before my hand slips.

[cygnus x-1]

From what you're describing it does sound like the tuning is a bit sluggish. Awhile back I remember messing with the tuning on my TM-1 a bit, but I can't remember exactly what I did. When I get back to the shop on Monday I'll take a look and report back. My TM-1 is from around 2000 so it won't be exactly the same, but it's another data point.


C|

[cygnus x-1]

Here's what I've got for settings on my 2001 TM-1 (all factory values):

2 X P GAIN = 48
3 X D GAIN = 6000
4 X I GAIN = 32
16 Y P GAIN = 32
17 Y D GAIN = 4000
18 Y I GAIN = 32
30 Z P GAIN = 32
31 Z D GAIN = 4000
32 Z I GAIN = 64

The hand wheels (yeah, it has the silly hand wheels ) are pretty stiff. I can get them to move but they correct very quickly and I can't hold them out of position at all.

Looks like your P/D values are off by a factor of 8, which is quite odd.

C|