I've read some info on the different types of home switches but am honestly a little overloaded. My goal is for something that has a high level of accuracy and repeatability so that when I pair it with a tooling plate I can reduce/eliminate some of the set-up time when switching between different fixtures.

Mechanical switches seems to be the least expensive and most widely used. On the downside it seems to be the least accurate, please correct me if I'm wrong.

There is a photo kit from cnc4pc that is interesting but I don't see an accuracy listed with it:
http://www.cnc4pc.com/Store/osc/product_info.php?cPath=29&products_id=183

I know I saw a proximity system listed somewhere in the forum a while ago. I can not seem to find it again no matter the amount of searching I have done???

Any advice and recommendations on products would be greatly appreciated.

Thanks again!

Mark [mkenney]

Editted to say HOME switched instead of the Limit switch I intially listed it as inaccurately. Thanks Crevice!!!

I think you mean HOME switches. Limit switches do not worry about accuracy/repeatability--They just need to stop motion when limit of travel is reached. Actually, with a stepper system and mach3 soft limit settings, you could forego limit switches altogether.

Mechanical home switches need to be roller type and are only as accurate and repeatable as their mounts and trigger blocks are solid. Good roller switches solidly mounted and protected from chips/coolant are very accurate/repeatable.

CR.

I think you mean HOME switches. Limit switches do not worry about accuracy/repeatability--They just need to stop motion when limit of travel is reached. Actually, with a stepper system and mach3 soft limit settings, you could forego limit switches altogether.

Mechanical home switches need to be roller type and are only as accurate and repeatable as their mounts and trigger blocks are solid. Good roller switches solidly mounted and protected from chips/coolant are very accurate/repeatable.

CR.

Thanks Crevice, I updated it to say home switches as it should of, I miss-speak on this often, going to have to watch myself better!!

Are these roller type?
http://i360.photobucket.com/albums/oo44/kenneme1/mill/200802/DSCF0069.jpg

Are we talking .001 in repeatability or better then that?

Thanks again for the help!!!

Mark [mkenney]

Yes, and they are very repeatable, but only if they or their trigger ramp don't move.

CR.

Repeatability also depends on the accuracy of your steps/inch settings. If Mach3 says your axis moved three inches, but the axis actually moves 2.95 inches, then repeatability will be non-achievable.

CR.

The usual method of wiring home switches is Normally Open. It will close and complete the circuit when actuated. This requires three inputs and one ground connection on the Break Out Board.

Limit switches are usually wired Normally Closed and in series. If a switch opens (Or a wire breaks) all axis motion is stopped. These require only ONE input and a ground connection to the BOB.

The switch on the left of your pic can be wired either NO or NC. The one on the right will only be one or the other.

CR.

Repeatability also depends on the accuracy of your steps/inch settings. If Mach3 says your axis moved three inches, but the axis actually moves 2.95 inches, then repeatability will be non-achievable.

CR.

Been meanning to check this. I intially set it based on the screw rate and stepper specs. From what I'm gatherring, this doesn't mean it is set correctly.

On the switches, my current ones are definately not mounted as rigid as it sounds like they should be! Also used the standard arm microswitches without the roller as that is what I had at the time and was waiting for the other's to arrive.

And the learnning curve goes on:) Thanks for the continued assitance!

Mark [mkenney]

Your SPI may or may not be correct, depending on the mechanical accuracy of the axis. Make a fixed distance cut or line in each axis, measure it, and compare the actual travel to the Mach3 DRO travel to be sure.

CR.

I was really impressed with the accuracy, low cost, and durability of Vince's proximity sensor setup:

http://www.cnczone.com/forums/showthread.php?p=541846#post541846

Cheers,

BW

I currently have roller micro-switches for home/limits on my knee mill, and they work OK. Repeatability seems to be at least 0.001". However, the little roller arms are delicate, and easily damaged, and, of course, they are not coolant/oil/chip friendly. I am switching to fully enclosed optical limits, which I expect to give much greater accuracy, and, being completely sealed, will be oblivious to all the nasty stuff flying around a mill. This also allows me to have separate home and limit switches in a small space, by putting two optos next to each other, both operated by the same flag.

You normally want to connect home switches as normally closed, and simply connect them all in series. There's no advantage whatsoever to having separately wired switches. Also, using normally closed, you'll know immediately if one fails. If wired normally open, you won't know if one has failed until you try to home, and the axis sails right past the switch and hits the hard stop. Similarly, all my optical switches are normally conducting, and turn off only when the home/limit is hit.

Regards,
Ray L.

You may want to look into Hall Effect switches. You would be amazed at how many places they are used and their accuracy. They have three connectors - +v, -v, output. For my G540 I connected them directly to my 24v psu and the output directly to the G540 - works like a charm! Accuracy is a smaller number than I can measure with the equipment I have - a .001 dial indicator. The hardest part is getting your magnets aligned to where you want them. The other nice thing about them is the cost - $2.30 from DigiKey - p/n US5881LUA-ND. The magnets you can get just about anywhere, I salvaged mine from some cheap, $1.00, wooden boxes from a Michaels craft store - 2 magnets per box!

Gary

You normally want to connect home switches as normally closed, and simply connect them all in series.


I would think you want independent homes, and tied together limits no?

I would think you want independent homes, and tied together limits no?

No. It doesn't matter if you connect 3 separate switches to 3 inputs or all 3 to one input. Mach starts moving in the first axis and sets home when the switch opens, then it goes to the next, and then the third. Even with limit switches it won't matter - Mach stops all movement when a switch is opened, it doesn't matter which switch.

Gary

I would think you want independent homes, and tied together limits no?

Nope. The way Mach works, it makes absolutely no difference. For both home switches and limits, all Mach knows is you're at a home/limit. It neither knows, nor cares, which one you're at. It homes one axis at a time, so it assumes when a home switch it triggered, it's the one for that axis (seems reasonable....). For limits, it doesn't care which one is hit, it simply STOPs. So, functionally, there is no advantage to running all the extra wires to make them separate.

Regards,
Ray L.

Thanks again for all the info. I ordered some hall effect switches to try out. I really like the idea of them being so swarth resistant from the sound of it. My home switches are currently wired normally closed as is rcomended. This bit me a couple times recently after installing the flood coolant and not properly protecting the switches. Bad sounds when the switch stay closed while homing:-)

Thanks again for the help so far!

Mark [mkenney]

I was really impressed with the accuracy, low cost, and durability of Vince's proximity sensor setup:

http://www.cnczone.com/forums/showthread.php?p=541846#post541846

Cheers,

BW

I'm trying to wire this up now. Not getting the hall effect to do anything yet, but know I'm missing something. Looks real great, now to get it working!!

Mark [mkenney]

I posted this question over in this forum:

http://www.cnczone.com/forums/showth...846#post541846 (http://www.cnczone.com/forums/showthread.php?p=541846#post541846)

I'm trying this setup and struggling so far. I have the same sensor's and swear I thought I have things wired correctly, apperently not. I have +5v going to pin 1 (VDD), pin 2 to the input on the break out board and pin 3 goes to the ground. I have a 1k resitor between the +5v and the input on the breakout board. I wave different magnets different ways and get nothing. What am I doing wrong, been one of those days?

I assume you're not really using application circuit 12.1 just what you have listed above correct? I would wire up the circuit in 12.1 exactly, hooking grounds to the ground on the BOB and the Vcc to the 5V on the BOB. Before hooking up the Vout to the BOB, do you have a multimeter or better yet a scope? Personally I test on a breadboard on the bench (electronics, not machining) before it goes near the controller.

this is something that i should really put some time into on my machine, as it sits now i have no limit switches, or home switches or anythg.. i have some of the little roller switches, but running full flood coolant makes me worrie about them...

I have a pile of home door and window alarm magnetic contatcs, i wonder if i could use thoes, cause i can seal them up, and then i dont have to worrie about the coolant getting in them...

one other thing to think about, if your using any kind of magnetic switching, watch your ferrous swarf.. it could build up on the magnets and cause problems...

I assume you're not really using application circuit 12.1 just what you have listed above correct? I would wire up the circuit in 12.1 exactly, hooking grounds to the ground on the BOB and the Vcc to the 5V on the BOB. Before hooking up the Vout to the BOB, do you have a multimeter or better yet a scope? Personally I test on a breadboard on the bench (electronics, not machining) before it goes near the controller.

Yep, just using what I listed. I ordered the 470pF capacitor that I did not have. Was trying to get away with just a pull up resitor as the orginal poster of this was able to do, not my luck I guess:-) I do have some breadboards and a multimeter and orginally bread boarded it and was never able to get any change in the output?? I assume when they say that the output is active low that it turns into a gorund when the hall effect sensor sees a magnetic field, is this correct?

Thanks for the help!

Mark [mkenney]

this is something that i should really put some time into on my machine, as it sits now i have no limit switches, or home switches or anythg.. i have some of the little roller switches, but running full flood coolant makes me worrie about them...

I have a pile of home door and window alarm magnetic contatcs, i wonder if i could use thoes, cause i can seal them up, and then i dont have to worrie about the coolant getting in them...

one other thing to think about, if your using any kind of magnetic switching, watch your ferrous swarf.. it could build up on the magnets and cause problems...

Standard roller micro switches will not last long in coolant, as they're not sealed, and will eventually get coolant inside and stop working reliably. The magnetic alarm switches will have very poor repeatability. They might work as crude limit switches, but I'd be amazed if they were sufficiently accurate and repeatable to be used as home switches. And, they are almost certainly not sealed either. Best bet is either optical switches, or good magnetic proximity switches. The latter is covered in another active thread in this forum right now.

Regards,
Ray L.

The latter is covered in another active thread in this forum right now.

Regards,
Ray L.

Moved some info over to the other thread to try to combine. Torn as to which thread is better for other's to have this in. The other thread is the orginal but started with something totally different. The Subject line of this one will be easier to find for other's when working on limit switches. As stated in the other thread, I yield to the admins and others that know better to advise on what to do. My fault for asking the question in two places, I should know better...

Mark [mkenney]