| | |
my haas 4th axis has gotten slack in it. if you dial it at zero and move it around a few times then home it to zero it will be off a couple thousandths every time. there appears to be some slack in the belt between the motor and the drive pully. Is there an adjustment on this belt or is it just a replacement item? would the belt being worn be enough to cause it to be out ? doesnt it have some way to verify position? sorry for the stupid questions but im new. HAAs vf2, Haas HRT 210 4th axis
If it is like the ones I've had, the motor is moved to add tension to the belt. I would try that first. After that you might need to adjust the worm to the worm gear. It may have lost some adjustment because of wear.Originally Posted by mattfurches
Like most indexers, it gets it's location off of the back of the motor itself, where the encoder is located. A loose pulley, belt, or gear adjustment will not be noticed by the encoder.
Mike
Haas VF-2, HA5C, Hardinge CHNC 1, BobCAD V23
can you give me any insight to how i adjust the screw? I adjusted the slack out of the belt but it still has .002 of runout.
I think you mean it still has .002 of backlash, not run out. Now is the time to shim the gears so they make tighter contact.
http://www.kirkcon.com/
The adjustment should be on the end. That is where the bushing is. The bushing is slightly eccentric, so the turning it will move the worm closer or further away from the worm gear. If your unit is like the one in this picture, it should be under that cap. I have the older unit and it has a dial face on it on that end that is obsolete now.
It is very touchy, so don't know if you want to try it. If you do, take the belt off so that you can turn the gear by hand and feel the pressure on it. The bushing adjustment is very sensitive too. Mark your starting point and make very small adjustments, re-tightening the set screws that hold it in place after each one. It has to be tightened down before you can get the feel of it again. They move when tightened.
Too loose---you have slop. Too tight---you have excessive pressure between the gears and can damage the unit with excess wear and binding of the shaft and motor.
Fun Huh! Also, it is best to rotate the unit a complete cycle to insure that there is no bind in any position and that will take some time.
When you say it has .002 of slop------where? At what distance from center?
Mike
Haas VF-2, HA5C, Hardinge CHNC 1, BobCAD V23
If by ".002 error" you mean 2/1000 of a degree [of rotation], you're getting concerned over an error that you cannot even measure.
On the surface of a cylinder 2 inches in diameter, a .002-degree error will result in an error on the circumference of about .000035 inches. That's 1/3 of 1/10000 of an inch. I don't personally have any way to detect or measure such an error, and I'll bet you don't either.
If you're working on a 6-inch cylinder the surface error will still only be a whopping 1/10000". In my shop that is more than acceptable. I know it's annoying on an intellectual level - our twin-head Haas indexer does the same thing - but we have noticed that the "error" is not accumulative. In fact, over a run of several parts the position fluctuates between a tiny discrepancy (your ".002") and "perfect".
One last note: on our Hass - circa 1992 - the adjustment is a set screw accessed from underneath the worm screw that applies upward pressure between the worm gear and ring gear. So far we have resisted the urge to mess with ir, and if your error is really only .002 degrees I would suggest you do the same.
TW
I am using a straight piece of material about 10 inches long . It is mounted vertically in the rotary and dialed perfectly straight up and down. When i move the rotary back and forth say halfway around then back and i return the rotary to the same position i used to dial in the piece it dials around .002 off. Im dialing around 9 inches from center of the rotary. In other words, at 9 inches from center the rotary is missing its 0 by .002.
That is not a lot of error unless you need to hold positional tolerance within 0.001". Does the error show the opposite way if you rotate the axis the other direction and back? If so, you might can dial out some of the error with the backlash parameter for that axis.
http://www.kirkcon.com/
Have you checked the specs for the HRT210 on the Haas website? As far as I can remember accuracy is +/- 15 arc seconds and repeatability is +/- 10 arc seconds. Do the calculation and I think you will find that 0.002" at a 9" radius is pretty close to spec.
An open mind is a virtue...so long as all the common sense has not leaked out.
Just used my cad system and calculated the angle difference and it is 0.0128 deg. That's pretty darn close to the specs.
Repeat your test and come to your setting from the same direction and see if that does it. How good an indicator are you using etc.
Mike
Haas VF-2, HA5C, Hardinge CHNC 1, BobCAD V23
Okay, now I'm thinking when you write "missing its 0 by .002" you may mean it's off from it's starting position by 2/1000 of one inch... yes?
For future reference, by the way, there are conventions that have been adopted to help others know what units are being discussed. For example, your error may be in inches, millimeters, degrees, etc. So if you were to write '.002" ' then we would know it was .002 inches. If you wrote it as '.002mm', we'd read it as .002 millimeters; if you wrote it as '.002*', we'd read it as .002 degrees. Without a unit designation it's kind of difficult to know what you're actually talking about.
That said, it now sounds like you mean your backlash error is .002". Given your part description, that calculates out to a rotational error of ~ .0127*, which is a bit more significant. Also, if it is backlash error it will be the same regardless of how far you rotate the piece before reversing and returning to "zero". In other words, try rotating it only 10* and back to 0*; you should generate the same .002" error IF it's a backlash issue.
One thing we've noticed on our twin Haas: it will often show a very tiny return-to-zero error (~.002* - .004*) if we stop a program in mid-run, but if the same program runs to completion it will return to exactly 0* without any "error" at all. We've come to believe this is controller related, rather than indexer error. So perhaps you can resolve the issue by incorporating a full 360* return-to-zero step (followed by re-designating the end point as "0").
After all, turning a full 360* to return to your start position cannot possibly take any longer than running forward 180* and then back 180*. Rather than fooling around with adjusting your indexer, perhaps you can resolve you issue within the program itself.
TW
It is a common fourth axis problem. Firstly You should check the tuning of servo motor which is responsible for rotating the fourth axis. If it is ok, then it is the probable issue of backlash in gear.you will be more sure if you check the fourth axis by rotating it 1 degree positive and negative. There are various gear trains included spur gear, bevel gear etc. by dismantling the fourth axis you should check all the gear backlash and if necessary a spacer collar may be introduced to minimize the backlash.
Posting Permissions
| | |
LinkBack URL
About LinkBacks
