Haimer 3D - Calibration Problems

[matt-b2]

Hi,

Does anyone here have a 'Haimer 3D-Tester New Generation' dial? I find I have to realign the concentricity of my dial every single time I use it - Not a single day passes by without the damn thing needing readjustment. Am I the only one who has this problem? Also, does anyone else find it takes a lot of force from the grub screws to get the probe concentric?

The whole point of buying the dial was to save time aligning things on my machine - I was not expecting that I'd have to frig about for 20 minutes every time I want to use the stupid thing. After using the dial yesterday, I accidentally broke the probe tip and almost threw the dial across the room! Wish I hadn't spent the money on it - a five quid wiggler seems to do a better job.

Matt.

[matt-b2]

Well, I guess the complete silence is testament to the value others see in the Haimer - In the junk drawer it goes.

[Neale]

Matt - there might be a lot of people out there, like me, who've never heard of these things before. I've just had a look at the website, and I can see that they might be useful to someone doing second-operation jobs who need to accurately reset workpieces. I suspect that a lot of CNC users don't work this way; one setup per workpiece where profile and internal machining is done at one setting is more common? I'm thinking out loud here, but that's the way I usually work. Vertical mill (mine is still manual - CNC conversion is on the list...) is a different matter. However, I find a conventional wiggler very quick and easy to use. Haimer seem to suggest that calculating half ball diameter is a complex, difficult, job. Maybe US education is a little different from European education, but I don't find calculating half of 0.2500 to be so very difficult. I don't even need to use my fingers to count or anything. And with a DRO, pick up the edge, switch to imperial, punch in 0.125 (+ or - as appropriate), and if necessary switch back to metric. And using a wiggler with a DRO to find centres is trivial.

Haimer seem to be talking of commercial systems where you can permanently put the edge-finder into a toolholder, and I would guess that this takes out recalibration every time you pick it up. Me, I'm swapping collets all the time and I would expect to recalibrate because I'm not sure that ER collets are that good. The wiggler is beautifully self-calibrating. But I don't have customers paying for my time so the time/money calculations are somewhat different.

[Web Goblin]

I did look at one of these a while ago and ended up buying the Tschorn one. Excellent little toy it is too.

[matt-b2]

Thanks for the replies - Most the things I do on my machine require second operations. I bought the Haimer in the hope it would help realign workpieces more accurately than a wiggler - Just seems to cause more trouble than it's worth though. I don't have any problems with my ER collets - No concentricity issues with my end mills.

[uli12us]

Are you sure, you have no spindle or toolholder problem. Maybe a little chip that is clamped between spindle and toolholder.

[Neale]

I know about the "autoleveller" software that will probe the surface of a piece of work and modify gcode so that depths are accurately referenced to the real surface of the work and not where the surface should be.

Is there an equivalent that anyone's written (Mach3 add-on?) that would do the equivalent for second-operation work? Clamp the work down without needing too much care, probe a few reference points, and then modify the toolpath to allow for misalignment (mainly rotation, I guess). After all, with CNC machines, we don't really need to align work with a fixed table axis in the same way as with manual machines (he says, having been clocking bits of aluminium plate this evening for coordinate drilling hole patterns).

[uli12us]

If you have a good control SW, than you can enter a turn of your part in each axe.
But thats not as good as you clamp your part parallel to X-Z.
You'll see the small steps, when the control correct the position, against a move only in 1 axe.
And a autoleveller works best with a switched 3D-Tester, because you can move, and the control automatically stops and take the measure if the switch is actuated.
With a manually operated tool you must take over each measure manually.

[matt-b2]

Are you sure, you have no spindle or toolholder problem. Maybe a little chip that is clamped between spindle and toolholder.

Hi uli12us, no - There is no noticeable run-out on the Haimer's shaft when it is in the collet holder, but the run out on the probe can be visibly out (1-2mm) only a day after calibration. I alway keep the collets and spindle clean - Never have a problem with end mill run out.

[komatias]

Hi uli12us, no - There is no noticeable run-out on the Haimer's shaft when it is in the collet holder, but the run out on the probe can be visibly out (1-2mm) only a day after calibration. I alway keep the collets and spindle clean - Never have a problem with end mill run out.

Matt,

can you run us through your process a bit more?

What spindle taper do you have on the machine? R8 MT-3, ISO?

Are you using a single collet chuck that you torque-untorque each time and replace endmills or drills with the haimer?

If you can spare it, se aside one collet chuck and use it only for the haimer. That way it is clean and will be as repeatable as possible (not trying to teach you anything you don't already know)

Was helping a lad out that had bought a 20mm LED/Beeping type edge finder by Vertex. He had it in a ER40 collet chuck and was also seeing massive runouts due to the length of the tool, basically the taper of the collet and that of the chuck were out and could be tweaked by hand. Once the tool was tweaked, it was fine...until he rammed it into the side of a vice