Hall effect proximity sensors

[Voicecoil]

I'm thinking of using these for homing sensors as they have a longer sensing range than the standard proximity sensors which I'm using for limits.Just wondering if anyone had experience of them and what the accuracy was like please.

[the great waldo]

I'm thinking of using these for homing sensors as they have a longer sensing range than the standard proximity sensors which I'm using for limits.Just wondering if anyone had experience of them and what the accuracy was like please.

Duh Link???

[Voicecoil]

I was just asking for general advice rather than for a specific device - but likely something like the NJK-5001C

[the great waldo]

I was just asking for general advice rather than for a specific device - but likely something like the NJK-5001C

Hi Voicecoil
I'm not sure if you checked your original post but you wrote "I'm thinking of using these for homing sensors" however "these" didn`t refer to anything, so a sensible reply would have been difficult. Maybe this was just a typo ?
Cheers
Andrew

[Voicecoil]

Hi Voicecoil
I'm not sure if you checked your original post but you wrote "I'm thinking of using these for homing sensors" however "these" didn`t refer to anything, so a sensible reply would have been difficult. Maybe this was just a typo ?
Cheers
Andrew

"These" referred to the title of the thread.......

[the great waldo]

Ahh

Ok I got it.

Cheers
Andrew

[cropwell]

https://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=13&cad=rja&uact=8&ved=2ahUKEwjciqvx3PbiAhUGbcAKHTZ3AC4QFjAMegQIABAC&url=https%3A%2F%2Fsensing.honeywell.com%2Fhallbook .pdf&usg=AOvVaw0unuV1LzDWNnAf7sglZ7fP

Page 64

But I personally would stick to Inductive sensors.

[magicniner]

2 inductive sensors in series on each axis, one sensing the linear position and one sensing a rotating tab or chopper on the ball screw or stepper would give you accurate, repeatable homing.

[Voicecoil]

2 inductive sensors in series on each axis, one sensing the linear position and one sensing a rotating tab or chopper on the ball screw or stepper would give you accurate, repeatable homing.
Having a rotating tab on the ballscrew is a good idea - essentially magnifies the resolution - I'll see if there's room to incorporate it.

[cropwell]

2 inductive sensors in series on each axis, one sensing the linear position and one sensing a rotating tab or chopper on the ball screw or stepper would give you accurate, repeatable homing.

I presume these would have to be the Normally Open types. I think it is over the top for woodwork, but probably OK for watchmaking.

[johnsattuk]

2 inductive sensors in series on each axis, one sensing the linear position and one sensing a rotating tab or chopper on the ball screw or stepper would give you accurate, repeatable homing.

Generally I don't use N/O types for limit switches, they dont cover broken connections. Could use two N/C in parallel

[magicniner]

I presume these would have to be the Normally Open types. I think it is over the top for woodwork, but probably OK for watchmaking.

Series for NO, so the circuit only closes when both close.
Parallel for NC, so the circuit only opens when both open.
:D

I'm planning this with my Chester Mill CNC conversion, I've become accustomed to being able to shut down my X7 and know there will be no positional errors when I turn it on the next day, select the WCS and slot a part into a fixture, I can't imagine going back to having to indicate every time I fire a machine up, I don't think a homing system which requires you to re-indicate a fixture which hasn't been moved is acceptable, in fact I think that's the definition of Useless.

[Voicecoil]

Generally I don't use N/O types for limit switches, they dont cover broken connections. Could use two N/C in parallel
Equally well, N/C types don't cover short circuits, had a few of those in my time.... :topsy_turvy:

[cropwell]

Series for NO, so the circuit only closes when both close.
Parallel for NC, so the circuit only opens when both open.
:D

I'm planning this with my Chester Mill CNC conversion, I've become accustomed to being able to shut down my X7 and know there will be no positional errors when I turn it on the next day, select the WCS and slot a part into a fixture, I can't imagine going back to having to indicate every time I fire a machine up, I don't think a homing system which requires you to re-indicate a fixture which hasn't been moved is acceptable, in fact I think that's the definition of Useless.

Why not have 1 NC and 1 NO, then it wouldn't matter whether you wire in series or parallel :joker:

It took some time for me to understand the second part of the reply as I don't get what part the Wildlife Conservation Society play in CNC. If I am cutting a few parts and have to turn off the machine then I zeroise the Z and 'Ref All Home' to send the X and Y back to the zero point. That way, when I switch on the next day, I know here the machine is.

BTW, I love these things, Engineers Lego......

https://www.arceurotrade.co.uk/Catalogue/Workholding/Vee-Blocks-Angle-Plates/Stevensons-Metric-Blocks

Cheers,

Rob

[johnsattuk]

Why not have 1 NC and 1 NO, then it wouldn't matter whether you wire in series or parallel :joker:

BTW, I love these things, Engineers Lego......

https://www.arceurotrade.co.uk/Catalogue/Workholding/Vee-Blocks-Angle-Plates/Stevensons-Metric-Blocks

Cheers,

Rob:thumsup::thumsup::thumsup:

[magicniner]

If I am cutting a few parts and have to turn off the machine then I zeroise the Z and 'Ref All Home' to send the X and Y back to the zero point. That way, when I switch on the next day, I know here the machine is.


My point is that with a well designed and implemented system that shouldn't be, and in fact isn't necessary :D

Work Coordinate System BTW

[cropwell]

Work Coordinate System BTW

Hi Nick,

WCS - I looked it up on Google. I actually rejected the idea that it could be the wildlife thing, toyed with Workflow Control System, Warehousing Computer Systems and World Coordinate Systems (Getting close there - that was defined as something to do with CNC). I did not think that with this latter CNC TLA you needed to know the precise latitude and longitude to define the machines X,Y 0,0, but apparently there it is. Even more incomprehensible is the UCS (universal Coordinate System, which even takes into account the Space/Time Continuum.

Maybe it is time that we had some TLA control, such as an Acronym Registration System (ARS). Maybe I could become an Acronym Registration System Executive (ARSE), although a lot of people probably think I already am.

Rob-T

[magicniner]

An Acronym is a set of letters pronounced as a word, if the individual letters are voiced as in VCR or TLA it is an Initialism thus, somewhat amusingly, TLA is not a TLA :D

[cropwell]

To get back on topic....

I do not use work offsets very often, which I presume depends on the accuracy of determining machine zero by homing all axes (axises ?). Most of what I do is one-offs and for that I zeroise locally when the piece is bolted down. If I have to do two or more, then I use a simple jig and, as I said before, leave the machine in a known position if I have to switch off and return to it next day.

It would be interesting to determine the increased accuracy with a dual sensor set. I would guess that it would not be significant for most home built machines.

What do the big boys, like HAAS, do?

[cropwell]

an acronym is a set of letters pronounced as a word, if the individual letters are voiced as in vcr or tla it is an initialism thus, somewhat amusingly, tla is not a tla :d

this forum is changing case - wtf !!!

[magicniner]

Hey, if you don't need it or want it don't do it! :D

With decent screws and nuts I think a repeatability of 5 microns isn't out of reach, or even difficult, that's where I'm aiming with my next build, machining accuracy is likely to be closer to 15 but if you aim low you're bound to achieve shite results! :D

[cropwell]

Hey, if you don't need it or want it don't do it! :D

With decent screws and nuts I think a repeatability of 5 microns isn't out of reach, or even difficult, that's where I'm aiming with my next build, machining accuracy is likely to be closer to 15 but if you aim low you're bound to achieve shite results! :D

I work to an accuracy measured in Angstroms +/- a foot.

25970

This is good enough for me, it is only woodwork :beer:

[routercnc]

I’ve just skim read this thread (funny in places !) so my answer to the first question which I think was essentially can I use Hall effect sensors instead of proximity sensors for homing -

They require a magnet to get close to the target to trigger and so like all magnets in the workshop will grow metal hairs and become unreliable. I have one to play with on my new E-kart foot throttle. Great if the magnet and sensor can be sealed in a box but I think one part must ride on the gantry for CNC machines if they are the sole sensor so they can’t be protected.

However for the auxiliary use as discussed they can be enclosed on the end of the stepper shaft at the rear of the motor or off the ballscrew and be sealed off. (Like the twist throttle of an electric bike).

[cropwell]

They require a magnet to get close to the target to trigger and so like all magnets in the workshop will grow metal hairs and become unreliable.

By Jingo, the man has a point !

[magicniner]

By Jingo, the man has a point !

So use a micro-switch for the linear sensor and an optical for the rotary! :D

[Voicecoil]

So use a micro-switch for the linear sensor and an optical for the rotary! :D

My thought as well, since slotted opto switches are common, just attach a slotted disc thingy to the ballscrew.

[cropwell]

Just out of interest, I thought I would see where the stepper positions itself with repeated homings using an inductive proximity sensor and a 5mm pitch leadscrew.

https://www.dropbox.com/s/xayeqtbo8l5qrqq/SUNP0009.AVI?dl=0

The video shows the stepper coming back to the same angular position and I tried it out 6 times with the same result, before I set up the camcorder.

I will leave you to draw your own conclusions about the necessity for a chopper

Cheers

Rob-T

[magicniner]

I will leave you to draw your own conclusions about the necessity for a chopper


I was only able to conclude the need for manual focus :D

[johnsattuk]

I was only able to conclude the need for manual focus :D:highly_amused:

[cropwell]

I was only able to conclude the need for manual focus :D

I am sorry that my camcorder was too close (AND TOO CHEAP) to sharply define the image. However, I do believe that the definition was enough to allow you to observe that the stop point of the stepper was within a few degrees. In fact, I thought that it probably stopped at the same point and a stepper motor, when energised by hold current, would settle at the same point on the step. My crude test would probably indicate +/- 1.8 deg and if you regard it as valuable, you may like to repeat the experiment with more accurate indication of the angle that the stepper motor settles on.

I am interested in knowing more about hold points and microstepping, but even if the homing point is one full step either way, then on a 5mm pitch screw you have a positional error of +/-0.025mm, which is good enough for woodwork.

[Voicecoil]

The thought strikes me that if we're talking about switching machines off and coming back to them, the best accuracy you can expect repeatably is one full step (1.8deg)?? as the motor will tend to "click" to the detent position either side once the driver is switched off? or am I missing something?

[johnsattuk]

[QUOTE=Voicecoil;109314]The thought strikes me that if we're talking about switching machines off and coming back to them, the best accuracy you can expect repeatably is one full step (1.8deg)?? QUOTE]

Assuming 2 phase steppers :smile: