[quadtech]

Wow, really beautiful. Those are the best yet.

[bobsch]

Are the threaded nipples steel or brass? If steel, does that affect the switches?

[boblon]

Are the threaded nipples steel or brass? If steel, does that affect the switches?

They are steel. I can't really say for sure if they affect them or not as I haven't really used them yet.

I can say that when I bring a magnet close the LED lights up and when I pull the magent away the LED goes out. That's about all the testing I can do for now.


@quadtech-Thanks

BobL.

[RomanLini]

Wow Boblon, they look like something from StarTrek! Home sensors from the 24th century! Going where no CNC machine has gone before! Ok, ok...

They look very nice and professional, and seem to light up VERY visibly with those high intensity yellow LEDs.

I also like th eway you incorporated the mini-DIN connectors into the design! Good work!

I'm not sure about using steel threads, you will run a chance of getting some residual magnetic buildup, but that depends on the material type of the steel. Some steel (most stainless types) are quite resistant to becoming magnetised.

If they do become magnetised the likely effect will be small, the sensors should still function but the detect point may shift a tiny amount due to the different magnetic field. They will have much less chance of becoming magnetised if the magnet is further away, so I suggest mounting the magnet as far away as is reliable. It will be interesting to see!

[boblon]

Gee, thanks Roman.

I didn't think of the steel nipple picking up some magnetism. So let's say it does. I wonder what pole the sensor end would be?

Brass nipples would be cool (cold in winter maybe, heehee )

BobL.

[RomanLini]

It it gets magnetised it will be the opposite of the pole magnetising it. So with the south pole passing by all the time the thread would become a weak N. That would not stop the sensor working, but would require the S magnet to be a tiny bit closer to activate the sensor (as the magnet S needs to overcome the local N).

Now if you had the sensor in the middle of the axis, and the magnet mounted at 90 degrees in a "slide by" configuration, then the average magnetising effect would be zero I think as there would be both N and S magnet poles going past the sensor.

Anyway I don't see it as being a big problem. Even if it does get a small buildup over time there are many ways to de-magnetise it. I have two different demagnetisers here that are commonly used in electronics work.

[mickelsen]

Hi Roman and everybody,
I'm new to this thread so please pardon me is I ask a dumb question. I am thinking about using these magnetic sensors as slide-by homing sensors on my machine. One thing I need to know is what does the orientation need to be between the little disk magnets and the Hall effect sensors for them to operate best? Should the disk magnets be edge-on to the flat side of the sensors or edge-on to the edge of the sensors? Should the magnets be flat-side-on to the flat side of the sensor, and if so, which side? I have some sensors on order, but they haven't come yet so I can't try anything. I'm not the best electronic experimenter anyway, so I might burn something out. I might not know if I've ruined something or that I just don't have them oriented right. Please help.
Also, my table top router is not terribly large, so I would like to make the switches as compact as possible. Any suggestions that people might have as to how to package them as compactly as possible would be greatly appreciated. Actually, I guess that everybody would like that kind of suggestion.
Thanks,
Mark

[RomanLini]

Hi Mickelson (Mark). Please check out the first couple of pages of the thread, there is a photo of a sensor, a resistor and a LED. That simple setup can be made temporarily just for a quick test and lets you test the magnet orientation.

The sensor is activated by a S (south) pole of a magnet. So if you have the disc magnet flat face pointed at the sensor it needs to be the S face.

Again, the test setup helps find out which end is which on the magnets as those magnets are not marked!

Alternatively, you can mount the disc magnet sideways in slide-by configuration and one side must be S, so that will always work (provided it is NOT right at the end of machine travel as slide-by needs to go past the sensor by up to 1/2" maybe more).

The first few pages also have photos of compact mounts and CAD files people have provided etc. That should be what you need.

[mickelsen]

Thanks for the help, Roman, but there is one thing I don't understand (Ha! only one, that's a laugh). You say that in slide-by mode the magnet may have to go past the sensor by as much as 1/2" in order to work correctly. But in the nice drawing and write-up in your post #83, it says that it may not need any more that 1-2 mm of movement to work. 1/2" seems like an awful lot. You would have a very hard time getting any sort of good repeatability and accuracy with that, it seems to me. Are people working with that kind of requirement?
Mark

[RomanLini]

No that 1/2" doesn't affect the accuracy! The 1/2" may be needed with stronger magnets as once the hall sensor is turned ON by a S pole, it stays on until the S pole gets sufficiently weak. (Until the magnet moves away far enough). The stronger the magnet the more the distance may need to be.

The 1/2" (if needed) is not the distance between magnet and sensor (when they are at their closest). That distance should be about 1mm to 3mm, and depends on your magnet strength.

It might not need 1/2" at all. If you are using the slide-by setup with the S magnet face pointed at the sensor, it will trigger the sensor as the magnet approaches, triggering ON probably from 1/4" to 1/2" away. This is all normal and fine for homing.

The homing procedure should move the magnet away to about an inch, then slowly creep back until the hall switch turns on. That is the sensor home, and the software will apply a calibration offset to that value (so it's not dritical WHERE it turns on, just that it remains in the same place). I hope that all make sense!

[mickelsen]

Whew! I've got to get some magnets and sensors (they are already on order) and mess around with them to see what's really going on.

Does Mach3 have the type of homing routine that you're talking about already built in? Does anybody know?

[RomanLini]

I don't have much knowlege of Mach3 but I believe that is how it works. Nobody seems to have had any troubles setting up Mach3 with the standard homing procedure.

Maybe someone else can answer that one?

[justCNCit]

Stopped reading at page 5 (very interesting)

Q1: With a Y axis gantry on a large CNC does Mach allow you to position the two Y axis homing switches independently through software? Otherwise they would be difficult to adjust

Q2: Has anyone come up with the kit form of these hall-effect switches or a finished product, if so how much and who do you contact. Thanks

[RomanLini]

Hi justCNCit, and welcome to the thread.

re Q1; I don't think people generally use 2 homing switches, they have 1 homing switch (to set the axis "home" position), or 2 limit switches (to limit travel at each end of the axis). They may even have 1 home AND 2 limits. And as far as I know Mach3 can be set up for any position of these switches.

re Q2; Nobody is selling these finished switches or kits that I know of, although they are allowed to make them and/or sell them as the design was released "open source".

From memory there is a commercial finished hall switch someone mentioned in this thread that you can buy, I think it was shown a few pages back before this page.

Actually I flipped back a couple of pages and found it. It is the Hamlin sensor, it was shown in post #434 and the few posts directly before that.

[justCNCit]

Thanks I wasn't expecting this fast of a response.

[ger21]

Q1: With a Y axis gantry on a large CNC does Mach allow you to position the two Y axis homing switches independently through software? Otherwise they would be difficult to adjust

No, you need to be able to adjust the physical location of the switch. On the new router I'm (slowly) working on, I'll be designing some switches that incorporate a micro adjust mechanism (screw )

For Roman. Mach3 allows you to slave motors to drive both sides of a gantry, and each side can have their own home switch, so that the two sides home independently (and simultaneously) to square the gantry during homing.

[justCNCit]

That's cool, it shouldn't be hard to make a micro adjusting switch, should keep us updated because that's exactly what I intend on doing

[RomanLini]

...
For Roman. Mach3 allows you to slave motors to drive both sides of a gantry, and each side can have their own home switch, so that the two sides home independently (and simultaneously) to square the gantry during homing.

Thank you Ger21 and my apologies to JustCNCit, I mistakenly assumed you were talking about "two Y axis homing switches" that were at opposite ends of the travel (like limit switches). There was a recent question about multiple home switches on an axis. Sorry if my answer was confusing but I didn't understand what you were asking!

Some of the home sensor mounts that people have made in this thread have slotted mounting holes, that would allow two sensors to be synchronised. But a fine thread "pushing" screw would be a great idea (as Ger21 said).

[justCNCit]

There is always an urge to speed up development of this hobby machine, it looked like you used transparent silicone to bury the LED and hall-effect transistor into a plastic block. That should work fine right?

[RomanLini]

Silicone can move if the wires are pulled and that might move the tiny hall sensor inside and affect accuracy. Also some silicones can be corrosive if applied directly to the metal connections, you need a "non-acetic" silicone for contact with metals.

I used a clear casting epoxy, but probably the most common and fastest material would be a typical hardware store "5 minute araldite" or "5 minute epoxy" where you mix the two parts and it's all finished and secure after 5 minutes.

A couple of people had mentioned different fillers too earlier in this thread, but mainly it's personal preference as to what you have on hand or what you know the best.