Hall Effect

[Agent_zed]

Can anyone confirm that the arrangement in the second image would also work?
View attachment 178259View attachment 178260




I'm not yet in a position to carry out empirical tests

These types of sensor are common in old car engines where you are converting from points to an electronic ignition or fuel injection.
If you look here https://www.motortrend.com/how-to/1604-how-to-convert-a-crank-trigger-ignition/ you'll see they mount the sensor exactly as per your second image. If an engine can turn at 6k+ rpm I would guess it would be fine for your motor as I assume you won't be spinning at that speed?

Hall sensors were also a standard fit in some cases, my kitcar (ford pinto engine) has original bosch electronic ignition on the distributor and it uses a hall sensor that is pretty small. This isn't my picture http://media2.turbosport.co.uk/2009/12/2010061615278550903dizzy_reluctor.jpg but it shows the hall sensor on the right (little black plastic box bit) and the 4 points that pass it to trigger the sensor. I don't know if all hall sensors are the same in terms of output etc but if you could use one of these it would be much more compact.

 

[Spectric]

Does a quadrature encoder give you position? Speed and direction, but not position, surely. Enlighten me.

When comissioning the encoder in a system you set the "home position" which is given by the home / Index / Z output and then the A / B outputs give you position relative to this home position, I suppose I should have used the term relative position.

 

[J-G]

Very interesting to see that the motor trade do exactly as I'd imagined, but reading the write-up does raise small questions about being able to 'detect' the Magnetic Flux - as has been mentioned.

Before I'd viewed the latest posting, I'd developed a new strategy based upon previous advice. I now intend to add a 'ring' to the top of the pully - I happen to have some 60mm dia. acrylic tube. I'll inset and glue that into a groove and the magnet will be pressed into the edge - a tight fit in a dovetail slot - in case the polarity does prove to be an issue I can test and reverse if necessary before final fix.

Two more images - - - a 2D elevation and a 3D view - which should give a better understanding.

Although the 'edge on' arrangement may well be perfectly acceptable for the speed range envisaged (0 - 2k) I suspect that this new arrangement will be more efficient or possibly less prone to error. Not that the actual speed is at all 'critical' - it's more of an "I'd like to know" situation.

 

[Spectric]

Why are you using a hall effect sensor ? there are many other types which work on various principles that you could use, optical sensors are easily interfaced into a system and many come with the emiter / detector in a simple package where a simple disk is used with some slots.

 

[Yorkieguy]

Very interesting to see that the motor trade do exactly as I'd imagined, but reading the write-up does raise small questions about being able to 'detect' the Magnetic Flux - as has been mentioned.

Before I'd viewed the latest posting, I'd developed a new strategy based upon previous advice. I now intend to add a 'ring' to the top of the pully - I happen to have some 60mm dia. acrylic tube. I'll inset and glue that into a groove and the magnet will be pressed into the edge - a tight fit in a dovetail slot - in case the polarity does prove to be an issue I can test and reverse if necessary before final fix.

Two more images - - - a 2D elevation and a 3D view - which should give a better understanding.
View attachment 178285View attachment 178286
Although the 'edge on' arrangement may well be perfectly acceptable for the speed range envisaged (0 - 2k) I suspect that this new arrangement will be more efficient or possibly less prone to error. Not that the actual speed is at all 'critical' - it's more of an "I'd like to know" situation.

That will work fine.

In 2017 I made an RPM tacho for my lathe using one of those cheap Chinese Hall effect sensors and a cheapo 4 digit LED display. I didn't use a round magnet on the shaft - I used two small oblong neodymium magnets one behind the other each about 5mm x 10mm so they adhered better to the curvature of the shaft. (In my case, outboard of the lathe headstock on my woodturning lathe). It's worked fine in the ensuing seven years.

I made a bracket for the sensor head which is held onto the pully housing with magnets and can be adjust for distance from the rotating shaft. If you look at the sensor head placement, you'll see the little oblong magnet just above it.

I built it into a little black ABS project box, glued magnets in the back of the box at each corner and it adheres nicely to the pulley housing.

Typically under £15.00 post free from UK suppliers, (Under a tenner from China if you aren't in a hurry):

https://www.ebay.co.uk/itm/17586005...lp:4429486&itmmeta=01HT2MZCQ5XRMRS5FWP8PKTJZB

Hope that helps.

 

[throbscottle]

I was going to suggest that if there's space, attach the magnet to the spindle. But looks like you've worked it out now.
Suggest fitting a counterweight though. Or several. At 1000's or even 100's of rpm that's going to be pulling on your acrylic ring until the glue fails. Also an extra eccentric load (albeit very small) on the bearing you could do without. Why not just cut a wooden disk to fit up against the pulley, it can then be held central by the shaft?
Anyway I'll shut up now!

 

[J-G]

That's re-assuring @Yorkieguy - thanks.

I'm only using a round magnet because that is what came with the 'kit'. I do have a (small) box full of magnets that I recovered from old hard disc drives (along with motors & discs it must be said!) but as yet I've not had the courage to cut one up in case machining them destroyed the magnetic properties - essentially, I simply don't know how neodymium magnets are produced - - - they could well be 'sintered' just as TC tips are for all I know!

Should the concensous of opinion be favorable, I could easily create a square magnet to fit the dovetail slot in the acrylic tube.

 

[throbscottle]

Why are you using a hall effect sensor ? there are many other types which work on various principles that you could use, optical sensors are easily interfaced into a system and many come with the emiter / detector in a simple package where a simple disk is used with some slots.

My experiments with optical sensing showed that it can be very sensitive to vibration so you get false counts. But then, I was making a very vibratey thing. Also susceptible to dirt and whatnot in the encoder wheel. I think the hall sensor would provide inherent immunity to these issues before any "technical measures"are taken.

 

[ajs]

I wonder why the Hall effect device is so big? Is it maybe an industrial device?
The actual semiconductor sensors are quite small..

Probably a built in amplifier, the Hall effect in practice gives a very low level signal that needs to be boosted to a practical level as soon as possible, otherwise it risks being swamped by noise and interference.

 

[Agent_zed]

That's re-assuring @Yorkieguy - thanks.

I'm only using a round magnet because that is what came with the 'kit'. I do have a (small) box full of magnets that I recovered from old hard disc drives (along with motors & discs it must be said!) but as yet I've not had the courage to cut one up in case machining them destroyed the magnetic properties - essentially, I simply don't know how neodymium magnets are produced - - - they could well be 'sintered' just as TC tips are for all I know!

Should the concensous of opinion be favorable, I could easily create a square magnet to fit the dovetail slot in the acrylic tube.

I thought the way they did it on the motoring website i linked was quite sensible https://www.motortrend.com/uploads/f/103902724.jpg?w=768&width=768&q=75&format=webp very easy to drill a hole to push the magnet in from the side and then it can't fly out. Although not sure what the black square is?

 

[John Brown]

Probably a built in amplifier, the Hall effect in practice gives a very low level signal that needs to be boosted to a practical level as soon as possible, otherwise it risks being swamped by noise and interference.

I appreciate that, but amplifiers are also very small these days.

 

[J-G]

A suplimentary question - - - I'm now looking at the wiring loom and have just thought that the signal from the sensor might be affected if the power line (24v) is laid adjacent (and parallel) to the sensor cable. ie. in a short length of 15mm truncking.

Am I being 'silly' and over-thinking potential problems, or is it somethiong I should be aware of and avoid?

 

[Sideways]

24V (DC, yes ?) shouldn't disturb a pulsed signal from the hall sensor and the sensor pulses at probably under 12V won't disturb the 24V feed to the motor either.
It is wise to separate control / sensor wiring from power wiring in many contexts but I doubt this merits it.

 

[J-G]

Thanks @Sideways - having just one 15mm trunking will be tidyer so I'll work on that principle.

 

[Yorkieguy]

That's re-assuring @Yorkieguy - thanks.

I'm only using a round magnet because that is what came with the 'kit'. I do have a (small) box full of magnets that I recovered from old hard disc drives (along with motors & discs it must be said!) but as yet I've not had the courage to cut one up in case machining them destroyed the magnetic properties - essentially, I simply don't know how neodymium magnets are produced - - - they could well be 'sintered' just as TC tips are for all I know!

Should the consensus of opinion be favourable, I could easily create a square magnet to fit the dovetail slot in the acrylic tube.

As I mentioned earlier, I just used two 5mm x 10mm x 2mm thick neodymium magnets one behind the other as they're closer to the curvature of the shaft (which in my case was about 4cms diam). Not sure how many on here are familiar with how powerful neodymium magnets are, but if you allow two of those small magnets to stick together you'll struggle to separate them. They come in all sorts of shapes and sizes - square, oblong round.

The ones I used were like these:

https://www.ebay.co.uk/itm/20326678...BS9Rn5RuTbxrAIdWyxkCyJMBzmHqKQCwaAktgEALw_wcB

10 x 5 x 2mm. Five for £3.49 post free from a UK seller.

They're so light in weight that I wouldn't be concerned about dynamic balancing if you consider the imbalance when say truing up a 30cm bowl blank on a woodturning lathe. (I'm not putting rockets into space).

I've also got one of those non-contact laser tacho jobbies - under a tenner. The readings of both compare exactly with each other. I got that before I made the Hall effect one. I just point it at a white dot on black tape stuck to the headstock spindle on my little metalworking lathe:

https://www.ebay.co.uk/itm/38560881...CBzyXrbiiTr6APumndmCxyaRbgWz8XdAaAnR-EALw_wcB

 

[cemannion]

A suplimentary question - - - I'm now looking at the wiring loom and have just thought that the signal from the sensor might be affected if the power line (24v) is laid adjacent (and parallel) to the sensor cable. ie. in a short length of 15mm truncking.

Am I being 'silly' and over-thinking potential problems, or is it somethiong I should be aware of and avoid?

Shielding the cable, or using shielded cabling is worth the small extra cost as it avoids interference, I too liked the magnet drilled into the face of the car pulley, I assumed that the black diamond was a visual clue and it means that you can use your sensor on the edge of the pulley. I have used just the hall module, which is much smaller, but not robust so you then need to enclose the whole area. I came to the conclusion that the one you have is the best
Chris

 

[throbscottle]

Agreed. The motor will generate interference because it's switching a lot of current, very fast, through some significant inductors - so you get significant back-EMF. So it's worth shielding the power and sensor wires from each other, even if you only (carefully!) wrap some foil around each set (connect the shields to supply common at the supply). Or keep them well separated. A ferrite bead on the motor wires will also be a big help, it will suck up a lot of noise.
Which reminds me. There is another approach - you can count the back-EMF pulses and derive the speed from that with a bit of electonikery (I am hopefully making an analogue tacho based on exactly this)

 

[Spectric]

There is another approach - you can count the back-EMF pulses and derive the speed from that with a bit of electonikery (I am hopefully making an analogue tacho based on exactly this)

With large industrial motors there are sensors that are strapped to the motor which provide a reference speed to the motor control cabinet .

 

[J-G]

It's quite heartening to see postings which point out these potential 'issues' - it all add to the learning curve!

I'm sure that you all realize that my knowledge of 'electronics' is qute rudimentary - one of the reasons (probably the prime) that I've bought in pe-built components rather than design & build from the ground up - so I'm having to consider all comments in the light of the limited information I provide as to EXACTLY what is going on.

It so happens that the 24v DC power cable that I am laying alongside the sensor cable is NOT the cable that is driving the motor. That cable will take a completely different route.

Here is a circuit diagram - not as you might expect from someone with real knowledge of electronics but a 'hybrid' which I find understandable

The 24V DC Cable feeding the variable speed controller - and hence the Motor - will be some 300mm distant from the Hall effect sensor.

Only the 24V line going to the display will be alongside the sensor cable.

 

[throbscottle]

Can't fault your drawing - perfectly clear, which is all that is needed