[kstrauss]

I hate to belabour the point but...
A round bottomed slot (from breaking at the hole) is wretched for gripping with a screwdriver
and
The shoulder would be firmly planted on the plastic piece so a bending force on the probe tip would tend to pull the "fuse" out of the probe rather than break it.

[bobeson]

I'm loving this discussion on tip breakage and prevention. I've gone through far too many probe tips myself, and have dreamed of many different improvements that could be so easily incorporated into the probe design to mitigate these problems. Now I'm happy to see Keen taking an interest in this, and I hope you proceed into making something useful to further improve upon the basic passive probe design, I'd love to buy a new passive probe to replace my broken-two-too-many-times 'active probe' and then upgrade the new probe with your improvements.

I will note that I experienced a bug in Mach (documented and workaround announced shortly following my experiences) which caused the probe to rapid over-travel in Z at random times, causing multiple breakages that ended up 'coining' the whole spring-loaded probe tip subassembly into the probe body when I didn't manage to e-stop fast enough. One of those coining events was the coup-de-grace that has sidelined my probe until I rebuild a few of the interior components. I think having z over-travel protection included in your design would be very much worthwhile. Even if this particular bug is no longer relevant, others will happen.

I think the best way to control tip breakage is for the control software to be smart about limiting probe movements following a sensor trip. I thought I read that PathPilot had this feature when it was first announced, but it doesn't sound like it does if folks are still having all of these breakage problems.

I dream of making a wireless probe that has built-in smarts, to control the machine travels and spindle rotation (assuming a suitably-equipped spindle) for automatically calibrating itself onto the spindle centerline. I think this could be accomplished by using concentric-eccentric adjustment rings for both the centerline and deflection angle ajustments. Calculating runout direction and magnitude could then be done while rotating, in order to perform hole locating and centering in the highly-accurate manner of co-axial indicators, along with the typical edge-finding function as exists currently. I suspect that a simple laser diode reflected off a probe-mounted mirror and sensed with a video imaging chip would allow all of the displacement measurements to be performed with accuracy. Power for the electronics and adjustment actuators could be generated by the rotating machine spindle and stored in super-caps, so that no batteries are required. Making this in low volumes for an affordable price with all of these features would be more-than-a-bit challenging, no doubt, but it would be an awesome product to own! I doubt I'll ever get around to working on these ideas, but hopefully somebody else will, so that I can actually use one someday.

[MichaelHenry]

Hi Michael. I am having trouble posting today - I hope this does not appear multiple times!

Re Z over travel. The stylus and hub are free to move a long way into the probe body as the spring compresses - and if it deflects sideways it will disengage - see the next video for more on this.

Keen

Thanks - I didn't see the video yesterday but it is here today. Looks good to me. Would this also work on a Haimer probe? I'm on my 6th or 7th probe tip there but only break them once a year now.

[keen]

[QUOTE=Philbobb;1778530]It's designed to break where it threads in. It looks like it should work, but I haven't tested it. Attached photo is the TI M4 030 07 020 styli and BI M4 000 07 008 fuse. $29 for the 20mm long styli with 3mm ruby ball and $4.50 for the 8mm long fuse. I unthreaded it for the picture, it does seat all the way.
Attachment 295568

If you break the stock ceramic tip, save it. Boil the end to release the glue, ruby or base side, and remove broken stub. Square the break up on the good part and reinsert adding CA glue if needed. Recal and your good. Cry if you want cause your probe is now shorter than everyone else's.

Thanks Philbobb - In theory that product may somewhat undermine the market for a resettable probe.

Looking at it, I would say it would take quite a load to break the fuse. The Switch triggers at 110g - 150g and my prototype resettable overload protector would be around 1000 g - I would say that fuse would break at more like 5000 g - and that would vary according to the overload direction. If I am guessing correctly then the probe internals could be damaged at those levels, and also the M4 thread in the Acetal hub could strip out - are you game to test it ?

But still - it is also about perception, and it is a cheaper option.... thanks for the information. keen

[keen]

Thanks - I didn't see the video yesterday but it is here today. Looks good to me. Would this also work on a Haimer probe? I'm on my 6th or 7th probe tip there but only break them once a year now.

Yes I am looking into Haimer option also. Keen

[Philbobb]

I hate to belabour the point but...
A round bottomed slot (from breaking at the hole) is wretched for gripping with a screwdriver
and
The shoulder would be firmly planted on the plastic piece so a bending force on the probe tip would tend to pull the "fuse" out of the probe rather than break it.

Scrounged up a little beer bravery and tested the fuse. I think you're right. I used my hands, not enough courage to use the mill lest I mess up the whole probe. I applied pressure with my thumb on the end until it started hurting, I'm guessing on the order of 20-30lbs. It didn't break but did bend the fuse a little at the stud. It did make impressions on the face where the shoulder seats. I tried a washer to prevent that, but I may be a little worried about the threads getting cocked in the plastic. I don't think it occurred to me that the 3-armed holder was plastic before you said something. I'm sure it works great for a Renishaw or similar commercial high $$$ probe. I think for these an aluminum stud that breaks with less force may be better. The stud threads in and out fairly easy so as long as the thread don't get mangled I think a screwdriver in the rounded slot would work reasonably well. It did give me 8mm more reach so it wasn't a wasted $4.50.


My posts weren't meant to take away from the intent of keen, but to show an option already there. Now that I've tried the fuse I don't think I would rely on it as a fuse. The original tip didn't break because of a crash, but bumble hands and dropping the probe.

[bevinp]

I dream of making a wireless probe that has built-in smarts, to control the machine travels and spindle rotation (assuming a suitably-equipped spindle) for automatically calibrating itself onto the spindle centerline. I think this could be accomplished by using concentric-eccentric adjustment rings for both the centerline and deflection angle ajustments. Calculating runout direction and magnitude could then be done while rotating, in order to perform hole locating and centering in the highly-accurate manner of co-axial indicators, along with the typical edge-finding function as exists currently. I suspect that a simple laser diode reflected off a probe-mounted mirror and sensed with a video imaging chip would allow all of the displacement measurements to be performed with accuracy. Power for the electronics and adjustment actuators could be generated by the rotating machine spindle and stored in super-caps, so that no batteries are required. Making this in low volumes for an affordable price with all of these features would be more-than-a-bit challenging, no doubt, but it would be an awesome product to own! I doubt I'll ever get around to working on these ideas, but hopefully somebody else will, so that I can actually use one someday.

Wow Bobeson,
That is a great dream and a feasible build (disregarding likely sell price). Yes it would be a fun project to attempt.

But having just read about scientific research into the way bees steer their way through leaves etc, a more exotic solution to auto centering a probe could follow the same principle. Apparently, bees fly through gaps by matching the left right vision motion; ie, when flying closer to the right obstacle than the left obstacle, the right vision movement will be faster than the left vision movement. The bee steers to make each apparent speed the same.

So my solution would be to super glue a bee to a lever that will make the necessary centering adjustments while “flying” the bee vertically down the hole. Then… hmmmm… I probably need to think about this a little more.
Bevin

[keen]

Scrounged up a little beer bravery and tested the fuse. I think you're right. I used my hands, not enough courage to use the mill lest I mess up the whole probe. I applied pressure with my thumb on the end until it started hurting, I'm guessing on the order of 20-30lbs. It didn't break but did bend the fuse a little at the stud. It did make impressions on the face where the shoulder seats. I tried a washer to prevent that, but I may be a little worried about the threads getting cocked in the plastic. I don't think it occurred to me that the 3-armed holder was plastic before you said something. I'm sure it works great for a Renishaw or similar commercial high $$$ probe. I think for these an aluminum stud that breaks with less force may be better. The stud threads in and out fairly easy so as long as the thread don't get mangled I think a screwdriver in the rounded slot would work reasonably well. It did give me 8mm more reach so it wasn't a wasted $4.50.


My posts weren't meant to take away from the intent of keen, but to show an option already there. Now that I've tried the fuse I don't think I would rely on it as a fuse. The original tip didn't break because of a crash, but bumble hands and dropping the probe.

Many thanks Philbobb! Very useful info. And thanks for posting the info initially also (no take aways there) Facts and logic are what matter here, and the more the better.

Yes I thought it might be a bit strong in SS - Luckily you did not strip the M4 thread in the Hub !...but it could be sleeved if that did happen.

Cheers Keen

[keen]

Wow Bobeson,
That is a great dream and a feasible build (disregarding likely sell price). Yes it would be a fun project to attempt.

But having just read about scientific research into the way bees steer their way through leaves etc, a more exotic solution to auto centering a probe could follow the same principle. Apparently, bees fly through gaps by matching the left right vision motion; ie, when flying closer to the right obstacle than the left obstacle, the right vision movement will be faster than the left vision movement. The bee steers to make each apparent speed the same.

So my solution would be to super glue a bee to a lever that will make the necessary centering adjustments while “flying” the bee vertically down the hole. Then… hmmmm… I probably need to think about this a little more.
Bevin

Bevin the problem with your solution is that any coolant on the bee or the bore of the hole might cause it to waver off course. I suppose a small coolant proof jump suit could be developed, but it is its wings I am worried about.

[bevinp]

Gee Keen, you're right again, as well as a caring naturalist.

[rowbare]

DAMN! Just snapped yet another Haimer probe tip to oblivion....UGGGH!!! The keyboard slipped off from the workstation and hit the ground resulting into a RAPID right into my fixture. I soooo hate doing that! My total spent on various probe tips for FY2015 has just swelled to $245...uggh!

An optimist would say that you have saved over $3000 worth of Haimers this year...

bob

[bobeson]

bevinp, you are actually on to something with your bee-vision proportional-feedback-control idea.

In fact, this exact control concept has already been widely exploited with the use of 'quadrant photodiodes.' These devices are simple light-sensing photodiodes that have been divided into four quadrants, cut through the center like a pie would be. Each of the resulting sensor segments then provides a reading of the relative percentage of the light source that is falling within that quadrant, which allows you to calculate the center-point offset of the (round-ish) light image being projected onto the sensor. These were used for the primary homing function in early heat-seeking missiles, for example, along with many other industrial auto-positioning applications.

[bevinp]

Bobeson,
Sorry to be slow in replying. Was on a four day holiday break without internet connection, which did not have any effect effect effect....on me me me.

When I was in the Australian Air Force flying bombers, we used to play target for the fighter pilot training in radar homing and Sidewinder lock on. The fighter pilots would call us dumbies (short for dumb targets) and we would call them knucks (short for knuckleheads). They liked to broadcast "Boom boom you're dead" or some other unimaginative comment as they flew past. I don't understand why they thought they were so clever in "making a kill", particularly since we were flying a simple race track pattern, not allowed to take any avoidance action. But then again it was at night, probably to stop them cheating, but we still had to have Nav lights showing.

I didn't know the actual heat-seeking design. So once again, nature got there a few millions years before us. And we think we're so smart!!.

Bevin