[Gerry Kmack]

I've had my 1100 mill for long enough now (arrived last November), that I am making making production parts and am fairly happy with the results.
Now that my first parts a looking close to where I want them to be, I am working to improve the overall accuracy of the parts. Carefully rechecking tool offsets, etc has improved DOC innacuracies, but I am now trying to figure out why X & Y cuts are not precise as they should be.
So, today, I started with a test blank that had been milled flat in all directions. I measured the X & Y thickness (using my Mitutoyo calipers):
X=1.3805", Y=1.3790"
Next I mounted the blank in my vise, and installed my Haimer 3d Taster in the spindle. I used the Haimer to sense the left side of the blank, and set the "X" DRO in PathPilot to "0".
Now I used the Haimer to sense the right side of the part. At this point, the DRO should have read 1.3805", or close to it. Instead, the DRO reads 1.3427". - that's almost 40 thou off!
I have tried resetting everything & rechecking my measurements numerous times, and I still get the same result. I get very similar results in the "Y" direction as well.
I checked the measurements using a different set of calipers, and got nearly identical readings, so it's either the Haimer, or the mill DROs or I'm just doing something wrong (entirely possible).
Any ideas what could be causing the discrepancies?
Thanks
Gerry Kmack
Cave Creek Woodworks
Pagosa Springs CO


Sent from my iPad using Tapatalk

[Steve Seebold]

Accuracy on my 5 year old 1100 has never been a problem. I consistently do work that requires tolerances of +-.001 and I have done work as close as +/-.00075 with no problem.

I even cut 718 Inconel to +/-.0015.

I have used the Haimer since I got the machine. I bought my Haimer in 2000 and have used the heck out of it.

[tmarks11]

I don't have a Haimler, so I have a stupid question: isn't there some way to calibrate the probe? It depends upon knowing the diameter of the ball on the probe tip (0.080"), and like any mechanical device occasionally needs to be zeroed.

EDIT: so the instructions (such as they are) have a procedure to adjust the concentricity of the 3d probe:

http://eland.org.uk/deckel/haimer_3dsensor_page1.jpg


When was the last time you checked the calibration of your mic? You need the 1" standard that (sometimes) comes with the mic. Or use a Jo block.

So take the haimler out of the equation. Program your mill to cut the outside of a square, and then measure the edges of the cut surface with your (calibrated) micrometer. Make sure to mic the diameter of the endmill as well (0.5" endmill isn't always 0.5" diameter).

[Gerry Kmack]

Steve - thanks for the input. I'm hoping my mill wil work out as well as yours has.
tmarks11 - OK fair enough; my calipers didn't come with anything to check accuracy, but I do have two 1-2-3 blocks. I just measured both blocks with both sets of calipers & all measurements were within (.0005").


Sent from my iPad using Tapatalk

[MFchief]

Have you checked the tool table to verify the diameter of the tip for you Haimler? They sell 0.2 inch tips and 4 mm tips, as well as other sizes. The diff between 0.2" as 4 mm is almost exactly 40 thou.

Terry

[Hirudin]

I'd guess the inaccuracy lies with the calibration of the Haimer. I guess the first question that seems pertinent is: did you calibrate your Haimer? If you didn't calibrate it: it's not likely to be accurate.

A test you can perform with the Haimer is to find the same surface with it rotated at different angles. For example, use the Haimer to find the left side of your fixed vise jaw and zero the DRO, then rotate the Haimer 90° and find the exact same surface. The DRO should read close to "0". Do the test again at 180° and 270°.

Take a close look at the stylus. Is there anything wrong with it? Push on it to see if it bends/flexes more than it should.
Not too long ago the ruby ball of my touch probe had cracked in half, but the split line was real close to being in line with the X axis, so it looked OK if you only glanced at it quickly. Maybe the ceramic shaft of the OP's stylus got cracked somehow.

I got coolant in my Haimer and it stopped working correctly, the mechanism got gummed up. Is the Haimer "spongy" in any way?

I'm going to guess your calipers aren't the problem, but an easy way to eliminate that potential problem is to simply use a block of a known size. Like that 1-2-3 block.

Have you checked the tool table to verify the diameter of the tip for you Haimler? They sell 0.2 inch tips and 4 mm tips, as well as other sizes. The diff between 0.2" as 4 mm is almost exactly 40 thou.

Terry

That's a good call! Perhaps Gerry Kmack ended up with the wrong model of tip attached.
But is that something that will be in the tool table? Because of the way Haimers work, I can't think of a reason for the ball diameter to be entered into the software.

[MFchief]

Hirudin, I think you are correct, when I put mine into the tool table I included the tip diameter, but I guess it isn't really used.

Terry

[MFchief]

As I think about it, wouldn't you get the error if you had the imperial tip on a metric Haimier, or vice versa? When the scale say zero you wouldn't really be at the center???

Terry

[adamvs]

I'll just add my $.02, which may or may not be relevant:

I use the cheapo Tormach TTS electronic edge finder ( about $40 ). I found that the rotational position it was in the spindle could make for up to a 0.003 difference in runout on the ball. If oriented just right in the spindle, run-out is almost zip, and if re-oriented at that spot, zero run-out is repeatable. So I made a mark on the edge finder and the spindle to line it up. I suspect the same is true of the Haimer even after the runout is zeroed the first time, ie, it needs to be put back in the spindle at the same place rotationally.

[tmarks11]

I use the cheapo Tormach TTS electronic edge finder ( about $40 ). I found that the rotational position it was in the spindle could make for up to a 0.003 difference in runout on the ball. If oriented just right in the spindle, run-out is almost zip, and if re-oriented at that spot, zero run-out is repeatable.

For an edge finder, you are supposed use it while the spindle is running at lower rpm (<1000). That removes any run-out related errors. If you think about it, run-out affects your end mill just like it does the edge finder, so any slight wobble in your end mill is going to cut slightly deeper then the width of the end mill. Same run-out with the edge finder pushes your centerline of the spindle that much from your workpiece, canceling the effect.

Here is a video from Tormach showing that. The electronic edge finder starts at time 1:56:

https://www.youtube.com/watch?v=f0od-cp_9dg

[MichaelHenry]

I recently checked my Haimer by touching off on either side of a 0.2000 gauge block and taking the difference between the two resulting DRO readings. I ended up with a 0.0003" error. You might try the same sort of thing with your 1-2-3 block.

BTW, calipers aren't really a good instrument when really accurate measurements are needed. In those cases, I try to use a quality brand tenth-reading micrometer that has been calibrated against a standard or a gage block.

[tmarks11]

As I think about it, wouldn't you get the error if you had the imperial tip on a metric Haimier, or vice versa? When the scale say zero you wouldn't really be at the center???

Which is apparently why you are supposed to re-zero the Haimier after replacing the tip.

http://eland.org.uk/deckel/haimer_3dsensor_page1.jpg

I would recommend the OP do the zeroing procedure that Hirudin described (and is in the link above) to ensure that isn't the source of error.

Even really expensive mechanical measuring devices can go out of whack with repeated use.

[adamvs]

For an edge finder, you are supposed use it while the spindle is running at lower rpm (<1000). That removes any run-out related errors. If you think about it, run-out affects your end mill just like it does the edge finder, so any slight wobble in your end mill is going to cut slightly deeper then the width of the end mill. Same run-out with the edge finder pushes your centerline of the spindle that much from your workpiece, canceling the effect.

OK, My 'edge finder' is a contact one with a .400" ball, I don't think it was meant to be rotated. There's no reason to do so ( and mar the part ), cause runout is virtually zero.

[pickled]

Haimer probes are darn accurate after the initial calibration. I marked my spindle and the haimer to make sure that I am always putting it in "clocked" as it was when I performed the initial calibration. Repeated measures studies that I have done here pretty much have the Haimer reliable to 0.0015" of total measurement error (I'm talking 3 standard deviations of data not eyeball error) for X and Y dimensions. This is after I have broken the tip off the sucker probably 7-8 times too.

[CadRhino]

Anyone have a link on recalibrating a haimer?

Thanks

Nathan

I'd guess the inaccuracy lies with the calibration of the Haimer. I guess the first question that seems pertinent is: did you calibrate your Haimer? If you didn't calibrate it: it's not likely to be accurate.

A test you can perform with the Haimer is to find the same surface with it rotated at different angles. For example, use the Haimer to find the left side of your fixed vise jaw and zero the DRO, then rotate the Haimer 90° and find the exact same surface. The DRO should read close to "0". Do the test again at 180° and 270°.

Take a close look at the stylus. Is there anything wrong with it? Push on it to see if it bends/flexes more than it should.
Not too long ago the ruby ball of my touch probe had cracked in half, but the split line was real close to being in line with the X axis, so it looked OK if you only glanced at it quickly. Maybe the ceramic shaft of the OP's stylus got cracked somehow.

I got coolant in my Haimer and it stopped working correctly, the mechanism got gummed up. Is the Haimer "spongy" in any way?

I'm going to guess your calipers aren't the problem, but an easy way to eliminate that potential problem is to simply use a block of a known size. Like that 1-2-3 block.


That's a good call! Perhaps Gerry Kmack ended up with the wrong model of tip attached.
But is that something that will be in the tool table? Because of the way Haimers work, I can't think of a reason for the ball diameter to be entered into the software.

[tmarks11]

Anyone have a link on recalibrating a haimer?

yes. Pretty close to what Hirudin described.

EDIT: so the instructions (such as they are) have a procedure to adjust the concentricity of the 3d probe:
http://eland.org.uk/deckel/haimer_3dsensor_page1.jpg

[CadRhino]

yes. Pretty close to what Hirudin described.

Thanks
I was hoping there was a video out there.

[Doubleoh9]

https://youtu.be/nkYpKWbA2P0

First result if you search "Calibrating Haimer" on YouTube

[CadRhino]

Have you checked the tram in your machine?

I've had my 1100 mill for long enough now (arrived last November), that I am making making production parts and am fairly happy with the results.
Now that my first parts a looking close to where I want them to be, I am working to improve the overall accuracy of the parts. Carefully rechecking tool offsets, etc has improved DOC innacuracies, but I am now trying to figure out why X & Y cuts are not precise as they should be.
So, today, I started with a test blank that had been milled flat in all directions. I measured the X & Y thickness (using my Mitutoyo calipers):
X=1.3805", Y=1.3790"
Next I mounted the blank in my vise, and installed my Haimer 3d Taster in the spindle. I used the Haimer to sense the left side of the blank, and set the "X" DRO in PathPilot to "0".
Now I used the Haimer to sense the right side of the part. At this point, the DRO should have read 1.3805", or close to it. Instead, the DRO reads 1.3427". - that's almost 40 thou off!
I have tried resetting everything & rechecking my measurements numerous times, and I still get the same result. I get very similar results in the "Y" direction as well.
I checked the measurements using a different set of calipers, and got nearly identical readings, so it's either the Haimer, or the mill DROs or I'm just doing something wrong (entirely possible).
Any ideas what could be causing the discrepancies?
Thanks
Gerry Kmack
Cave Creek Woodworks
Pagosa Springs CO


Sent from my iPad using Tapatalk

[Gerry Kmack]

Wow! Great feedback! I got busy in the shop yesterday and am just now reading all of this...
First of all, it did not occur to me that my Haimer might need calibration... I will look into that.
Also, I need to check the ball on my sensor probe, to see if it is correct or possibly damaged.
Someone mentioned measuring the actual diameter of my endmill... I do need to do that, as I have two 1/2" EMs that I use regularly, and have noticed that one cuts wider than the other....
Question: how do you measure the true (cutting) diameter of an endmill?
FWIW, at the start of this, I checked the tool holders for runout (it was very small), and mic'd the shaft of both endmills (within ~.0005 of 1/2"), yet the difference in their cutting width seems much greater than that (~.03)...?
When I get to the shop this morning, I will take a close look at my Haimer, and let you all know what I found.
Thanks again for the great feedback.
Gerry Kmack


Sent from my iPad using Tapatalk