A Low Profile CNC Tool Height Setter.

[Tweakie]

My existing ‘touch plate’ has served me well over the years but as this relies on electrical contact between the ‘tool’ and the ‘plate’ it was just not possible to use it with the diamond and ceramic tipped tools I now have occasion to use.

Something a bit more sophisticated was required and an active, ‘probe type’, low profile, tool height setter seemed to be on the cards.

This is my first prototype and as luck would have it, works just as intended. It is still work in progress, as testing is on-going, but so far so good.

An internal contact ring rests on 3 contacts under spring pressure and the overall height of the completed probe entered into my probing script as ‘plate thickness’. Any downward movement of the plate breaks one or more contacts for the G31 detection. Although the contacts are ‘normally closed’, the transistors invert this logic so the probe output is normally a weak High and only taken to a strong Low (GND) when contact is broken.

The amount of movement necessary to operate the probe is incredibly small but it is largely dependant on the flexibility in the top cover and operational movement is currently smaller than two of my Z axis micro-steps. It does not seem to make any difference if the plate movement is parallel (centre contact) or tilted (off-centre contact) there is never any contact bounce – it either makes contact or it doesn’t.

This type of ‘contact-break’ detection is used in some of the Baumer Electric limit switches – OK they have far more accurate manufacturing facilities than I have but they claim 1 micron repeatability.

My Z axis only moves in steps of  0.001875mm (although the accuracy of micro steps is a long standing point of discussion).
Jogging downwards with the axis set at 0.0010mm step size and zeroing at the exact point of electrical contact a further 2 steps were required to trigger the probe with the Z axis DRO now reading  -0.0037mm.

Of no practical purpose this 0.5mm text was engraved to a depth of 0.05mm just to see if it could be done.

Tweakie.

[PekkaNF]

Hmm. Looks good. Any chance of a photo of contacts/mechanics? How do they relate to those three screws wired to two transistors?

PekkaNF

[Swarfing]

Tweakie

Looks great and would be very interested in the drawings and BOB if your willing to share please? 

[Tweakie]

Thanks guys.

Once I have finalised the design and have completed a few more tests I will post the full constructional details, drawings etc. for those who wish to make this device (or something similar).

Here is a pic of the completed contact ring assembly and another of the machined parts.

Tweakie.

[philf]

Hi Tweakie,

The Height Setter looks very nice. 

One question - Is it failsafe? (If a wire breaks does it already think it has sensed the tool?)

How much overtravel have you built in?

Mine (which also doesn't need continuity through the tool) isn't very low profile (- about 30mm tall) uses an opto-isolator to pass the signal to the breakout board which doesn't invert the signal. If I don't plug the setter in or don't turn it on then the Z-height setting routine doesn't run.

 

Phil.

[Noitoen]

You can use a piezzo disc from a buzzer as a sensor mounted on some spring suspension just to act as a little over travel damper, with a little metal disc glued to the piezzo. Anything that touches the sensor will generate a voltage spike to be amplified and generate the home reset pulse even if the tool pushes the damper springs.

[Tweakie]

Hi Phil,

Good points you have raised and thank you for your comments.

Yes, I have incorporated what I consider to be ‘fail safe’ characteristics but these are not totally 100% foolproof and I am sure could be improved upon in the light of experience and others suggestions.

Currently the probe is connected to the machine with a 3 core cable and miniature jack plug, there is no on/off switch and it is always active.
If either the Positive or Ground wires were cut (or the probe not connected at all) this is detected (at the breakout board end) and the probing script reports ‘probe already grounded’ and no Z axis movement will be made when operating the ‘auto tool zero’.
If the Sense wire is cut the script runs as normal but the contact plate then over travels onto one or more of 3 ‘over travel’ mechanical switches. This is sensed (at the breakout board end) and in turn, triggers an ‘immediate stop’ by activating a machine limit.

The over-travel switch feature, operating a machine limit was, I must admit, not part of my original design but has subsequently been added (as a result of a suggestion made on the Artsoft forum) to protect the tool from damage in the unlikely ? event that the G31 command (or associated probing script) goes wild.

My available ‘over travel’ distance is currently 2.0mm but this will be reduced to approximately 0.5mm (before the over travel switch is contacted) when the updated switch PCB is fitted.

As always, this is an ongoing project and I am learning as I go but so far so good.

Tweakie.

[philf]

Hi Phil,

Good points you have raised and thank you for your comments.

Yes, I have incorporated what I consider to be ‘fail safe’ characteristics but these are not totally 100% foolproof and I am sure could be improved upon in the light of experience and others suggestions.

Currently the probe is connected to the machine with a 3 core cable and miniature jack plug, there is no on/off switch and it is always active.
If either the Positive or Ground wires were cut (or the probe not connected at all) this is detected (at the breakout board end) and the probing script reports ‘probe already grounded’ and no Z axis movement will be made when operating the ‘auto tool zero’.
If the Sense wire is cut the script runs as normal but the contact plate then over travels onto one or more of 3 ‘over travel’ mechanical switches. This is sensed (at the breakout board end) and in turn, triggers an ‘immediate stop’ by activating a machine limit.

The over-travel switch feature, operating a machine limit was, I must admit, not part of my original design but has subsequently been added (as a result of a suggestion made on the Artsoft forum) to protect the tool from damage in the unlikely ? event that the G31 command (or associated probing script) goes wild.

My available ‘over travel’ distance is currently 2.0mm but this will be reduced to approximately 0.5mm (before the over travel switch is contacted) when the updated switch PCB is fitted.

As always, this is an ongoing project and I am learning as I go but so far so good.

Tweakie.

Thanks Tweakie,

Mine has an overtravel of 2mm but hopefully I never need to worry about it. In operation I can't see any movement at all.

When I first set mine up I didn't realise the significance of some of the numbers in the macro and entered a number less than than the height (then 30mm) of my height setter. The 2mm of overtravel soon disappeared followed by the "ping" of another 2mm carbide cutter as it tried to bury itself in the top of the plate.

After cleaning up the plate the setting gauge is now only 29.74mm!

 

Phil.

[AdeV]

Ooh, eck, I've got all this to come, haven't I?

Luckily, my controller is so old, it doesn't have a tool height setting program; I do have a rather accurate handwheel thingy though, so I can manually set tool heights without having to worry about overtravel.

How repeatable are microswitches? I'm guessing they must be reasonably repeatable as my machine uses them as both limit switches, and reference marks, when setting itself up. It had occurred to me if they were good to, say, 0.01mm, they'd make an ideal tool height setter without having to worry about conductivity.

[HS93]

 Ade  would this do, Knowing you you probably have half a dozen already  http://www.machine-dro.co.uk/z-axis-dial-tool-setting-gauge-aluminium-body.html


Peter

[Tweakie]

You can use a piezzo disc from a buzzer as a sensor mounted on some spring suspension just to act as a little over travel damper, with a little metal disc glued to the piezzo. Anything that touches the sensor will generate a voltage spike to be amplified and generate the home reset pulse even if the tool pushes the damper springs.

Hi Noitoen,

Thanks for the suggestion but I am not sure of the performance of piezzo with relation to slow probing rates - it is my understanding that the voltage generated is tightly related to speed of impact but I will certainly try one and observe the results.

Tweakie.

[Tweakie]

Thanks Tweakie,

Mine has an overtravel of 2mm but hopefully I never need to worry about it. In operation I can't see any movement at all.

When I first set mine up I didn't realise the significance of some of the numbers in the macro and entered a number less than than the height (then 30mm) of my height setter. The 2mm of overtravel soon disappeared followed by the "ping" of another 2mm carbide cutter as it tried to bury itself in the top of the plate.

After cleaning up the plate the setting gauge is now only 29.74mm!

 

Phil.

Hi Phil,

Sorry to hear about the 2mm cutter.

It may be of interest but for my setup (probing at 200mm/min) the over travel (assuming a decelerated G31 stop) was calculated at 0.0543500044871 (d=v^2/2a) then adding the operating distance of 0.0037 indicates that a total over travel clearance of 0.0058mm should be allowed for so I am reasonably confident that 0.5mm should be more than enough. Only time will tell.

Tweakie.

[Tweakie]

How repeatable are microswitches? I'm guessing they must be reasonably repeatable as my machine uses them as both limit switches, and reference marks, when setting itself up. It had occurred to me if they were good to, say, 0.01mm, they'd make an ideal tool height setter without having to worry about conductivity.

Hi Ade,

I suppose it all depends on the quality of the microswitch but it is certainly worth a try.

Tweakie.

[AdeV]

Ade  would this do, Knowing you you probably have half a dozen already  http://www.machine-dro.co.uk/z-axis-dial-tool-setting-gauge-aluminium-body.html

Hi Peter,

Nice find, that's definitely something I would use.  But at nearly 70 quid including contribution to the Chancellor's beer fund, I think I might try adapting a dial guage... Nice bit of  for me anyway. And if I get bored, at least I know there's a commercial option! I've got a couple of old Mercer 10ths dial guages which I briefly tried to use on the lathe, before discovering that trying to chase down 0.0001" in an 80-year-old 10" 4-jaw chuck mounted to an 80-year-old lathe was a bit like trying to nail jelly (aka Jello) to a ceiling; they've been sitting quietly in the cupboard waiting for a project, this will be ideal

[Tweakie]

This is my 'over-travel' limit switch, well there are actually 3 of them but they are wired in parallel and operate on the 'first home' principle.
The total thickness, including the PCB is 2mm and it fits below the contact plate assembly in the base of the probe and operates a machine limit in the event that the G31 probing routine fails. That's the theory, only time will tell.

Tweakie.

[Lew_Merrick_PE]

How repeatable are microswitches? I'm guessing they must be reasonably repeatable as my machine uses them as both limit switches, and reference marks, when setting itself up. It had occurred to me if they were good to, say, 0.01mm, they'd make an ideal tool height setter without having to worry about conductivity.

I suppose it all depends on the quality of the microswitch but it is certainly worth a try.

Several years ago I was doing some development work for a major software/hardware company.  One of the areas involved qualifying SPST-NO and SPST-NC pushbutton switches for make point and break point repeatability.  C&K Switches cheapest (about $4.50/gross at the time) pushbutton switches (with .080 inches/2 mm overtravel) had a repeatability of .002 inches/0.05 mm.  Now, mind you, the actual make point or break point could vary by .025 inches/0.64 mm from switch to switch.  Their (C&K's) mid-range priced SPST switches were slightly better (I would have to dig out my lab book from that period to come up with numbers) and their high-priced SPST switches were more rugged, but no more repeatable.

Now, mind you, a small capacitor in parallel with the switch is needed to get this type of repeatability on almost any switch -- and many manufacturers build that into their switch unit.  I was sadly disappointed by Cherry Switch's showing in these test -- and I grew up with Cherry being the gold standard of switches.

[Tweakie]

The wiring between the over travel switch and the main PCB had made the probe difficult to assemble / disassemble so the case has been re-designed to include a small pocket to accommodate a slightly longer loop of wire. This seems to have overcome the problem - so far so good.

The new case has been CNC machined, lightly sand blasted then hard anodized.

Tweakie.

[Brass_Machine]

Nice looking case Tweakie!

[Tweakie]

Nice looking case Tweakie!

Thank you sir.

There was bit of an anxious moment when I had accidentally selected mm/sec. instead of mm/min. for the feed rate on the tool change - fortunately, I just caught it in time.

Tweakie.

[Brass_Machine]

I bet!

Are you going to be selling these? Or releasing it as kit or the files?

Eric

[Tweakie]

Hi Eric,

It is my intention to make this whole project 'open source', with all the drawings and CNC files made freely available on my website so that others who may be interested can make one (or something similar) for themselves.

I still have some work to do on the layout for the control board and more testing under varying conditions before it is finally ready but I am getting there slowly.

Tweakie.

[cnc-joe]

Hi Tweakie,

- Hey, I'd be very interested in making a probe like this.
- Have you had any luck in finishing the design/drawings up?

- As always- thanks for sharing your work with the rest of us.

CNC-Joe

[Tweakie]

Hi Guys,

Although it was my original intention to make this an ‘open source’ project for anyone to construct, commercial interest was shown. They took their time deciding but following a reasonable offer I have now sold the design.

Tweakie.

[Brass_Machine]

Hi Guys,

Although it was my original intention to make this an ‘open source’ project for anyone to construct, commercial interest was shown. They took their time deciding but following a reasonable offer I have now sold the design.

Tweakie.

Congrats on that Tweakie! Always awesome to make a little cash.  Too bad tho, I was looking forward to making one!