[Project_Hopeless]

I'm looking for recommendations on relatively inexpensive Home and limit switches for a DIY build. If possible I would like to use 3 limit switches also as Home switches. Then 3 less expensive/precise switches for the opposite end of stroke. What are you using that works well?

[Jim Dawson]

I have used these successfully for precision positioning (+/- 0.0005 inch). Requires a little fiddling with the ramp on the operating cam to get that repeatability, needs to be about a 90° edge with almost no radius so you get an instant snap over..

https://www.ebay.com/itm/10pcs-V-156...frcectupt=true

Can't get much cheaper than that.

[rcheli]

I used these. Pretty much the same as Jim used.
https://www.sparkfun.com/products/13119

Mount them so you can either adjust the ramp or the switch.

[Project_Hopeless]

I have used these successfully for precision positioning (+/- 0.0005 inch). Requires a little fiddling with the ramp on the operating cam to get that repeatability, needs to be about a 90° edge with almost no radius so you get an instant snap over..

You don't have issues shearing the roller off or the arm bending over time?

[Jim Dawson]

You don't have issues shearing the roller off or the arm bending over time?

You have to set the arm in exactly the correct position, so the cam hits a couple thousandths forward of the roller C/L. You also have a very small radius on the cam. Like I said, it's a little fiddly, but it works. Or you can buy a $60 precision limit switch.

[davida1234]

I actually used ones similar to those mechanical switches on my old machine but without rollers to eliminate any play/deterioration in the roller.
Working fine for over seven years but you should try to mount them with the arm facing down so that dust/debris doesn't pile up on the joint.
Had to clean junk off the arm once in a while.

[routalot]

I actually used ones similar to those mechanical switches on my old machine but without rollers to eliminate any play/deterioration in the roller.
Working fine for over seven years but you should try to mount them with the arm facing down so that dust/debris doesn't pile up on the joint.
Had to clean junk off the arm once in a while.

Now that is quality advice!I am not too far from doing this sort of thing myself and I hadn't even considered this aspect.Thank you.

[davida1234]

Very welcome....

[109jb]

I personally like the inductive proximity switches for limits and homing switches. They are very accurate, about 0.0002" from my testing, and have been very reliable on my machines. They are non-contact and just require a metallic target, which can be part of the machine or an added target.

https://youtu.be/IGO47l62T2o

[Project_Hopeless]

I personally like the inductive proximity switches for limits and homing switches. They are very accurate, about 0.0002" from my testing, and have been very reliable on my machines. They are non-contact and just require a metallic target, which can be part of the machine or an added target.

I'm surprised you get that accuracy from a prox! The ones I'm familar with on conveyors are no where near that accurate.

I worry with a prox setup if something is slightly out of wack and the axis drifts it could smash into the prox. How far is the metal target to the prox when it trips?

[davida1234]

I worry with a prox setup if something is slightly out of wack and the axis drifts it could smash into the prox.

They do seem to be about as accurate as the mechanical switches.

You don't have to arrange them such that the metal hits the prox.
Usually, the metal needs to come close like 4-8mm to prox.
It could just pass underneath.

Mechanicals are definitely easier to set up and don't need
a power source. They can also be arranged in series which
works better. It's more complicated to do that with e-switches
if at all possible.

[Project_Hopeless]

Do you get the same repeatability if the target passes underneath vs. approaching head on?

[davida1234]

I don't know, never risked it that way.

But I can imagine that you need to be closer with passing by vs. approaching.

[CitizenOfDreams]

Still, I personally don't like relying on a switch for long-term repeatability. Mechanical switches wear over time; proximity sensors may change sensitivity with voltage, temperature etc. The traditional approach - switch+encoder - seems more robust.

[wizard]

I think the important thing to understand here is that there is no perfect solution. For Example Micro Switch type switches of the type linked to above can be very good up until you get crap or swarf in the lever mechanism. Prox switches are great unless you are machining metallics that the switches can trigger off. In a lot of the machinery I work on axis limits are either mechanical switches with enclosed mechanisms or prox switches. Homing on these is a different issue because home is refined off the zero pulse (servos).

While a precise rehome can be beneficial it might not be worth the trouble. It is pretty common in the CNC milling world to probe the part (sometimes the fixture) before machining. That may or may not be useful for a router machine and the work you do. What I’m m getting at here is that if you go cheap with limit switches homing accuracy might not be a big deal. Say you are doing sheet goods, if the home position is off by a couple of thousands does it really matter if you are cutting parts out of a 4x8 panel.

Another thing worth considering is plunger style micro switches. That is switches that delete the lever arm for a plunger. The primary advantage here is the lack of a lever arm to bend or get cruded up. These require a bit more thinking when it comes to mounting though you can get plunger type switches with rollers.

Also if you want to get real cheap with the electronics/switches you can use a single switch with a long cam. That long cam though is usually more work that it is worth. However you can design in the “cam” at design time. You also loose the ability to know which direction faulted if your controller supports it. You also loose the ability to move the CAM or switch if you are in chicken mode with a new fixture.

[Project_Hopeless]

What I’m m getting at here is that if you go cheap with limit switches homing accuracy might not be a big deal. Say you are doing sheet goods, if the home position is off by a couple of thousands does it really matter if you are cutting parts out of a 4x8 panel.

That is my situation exactly. If I get 1/100 inch ,0.25mm or better repeatablity that will be more than adequate for my needs.

[109jb]

Was on vacation for a few days, but will try to address some of the questions that came up about my prox switch use.

1. The video I linked was really just a hack together for testing. I later made things much neater, but it showed the general arrangement I used on that machine which is approaching the target head on.

2. I have also used the prox switches with a target that passes the sensor from the side in a slide-by arrangement. I find this arrangement to be just as accurate and there isn't a possibility of crashing into the prox sensor because the target can't come into contact when set up this way. I did similar testing to what is shown in the linked video, and the repeatability seems to be as good as the head-on arrangement. Thinking about it, you could probably set up some kind of wiper arrangement to sweep off the target to tackle the swarf issues.

3. In regard to potentially crashing the target into the sensor, I still have the machine I used in the video with the head-on target and have never had it happen. This is a small machine and it decelerates quickly enough that it hasn't been a problem, but using the slide-by arrangement solves this.

4. Swarf can be an issue with false triggering, but the same is true for any switch. On the machine in the video, I would up putting the X-axis prox sensor and target under the table where swarf has not been an issue. For the Y-axis, I built a guard that keeps the swarf off, and the Z axis sensor is up high enough that it can't easily get swarf on it..

5. Accuracy of the prox switches is great, but as pointed out, how good you need is dependent on what you are doing. My machine that I linked the video to is a milling machine and not a router. For my use, I need high accuracy for cases where I want to shut the machine down, and then resume the job another time. I want the homing of the machine to be very close so that once homed it is ready to load a program and go., The prox switches provide high enough accuracy for my needs. Is it the best? No. As pointed out earlier a setup that uses a switch to get close and an encoder index to fine tune is the best, but not many people need that kind of accuracy and in the case of that machine in the video is open loop so the encoder index isn't there anyway. The +-0.0002" I have been able to get with just the prox switches is fine for me. Whether you need that much accuracy is up to you, but personally I see no reason to not put on accurate switches, be they prox, mechanical, optical, or whatever because the cost difference between a switch that isn't accurate and one that is isn't much. One of the reasons I do like the prox switches is because they are very accurate and also very inexpensive.

6. As far as prox sensor accuracy being affected by temp, voltage, etc. OK, I'll concede this as a possibility, but we are talking about pretty small changes in the sensitivity and even this is dependent on the particular user situation. My prox sensors are connected to a regulated power supply so there will be very little voltage change to worry about. As for the temperature, in the past I researched this and at least for my use, the temperature effects were going to be negligible. The change in sensing distance with temperature goes from about +10% at -13 deg F to -10% at 140 deg F. This is about 0.065% of the sensing distance per degree temperature change. My shop is always within about 20 degrees whether summer or winter, so that equates to a maximum change due to temperature of about 0.001". Again, it depends on your use, but that is fine for me.

[wizard]

Personally I’d go with a roller plunger type switch. Something like: https://www.automationdirect.com/adc...yABEgJ28PD_BwE. Note that the price is rather high but you can get something molar switches from China pretty cheap. Such switches are fairly resistant to damage and crude and thus are good in exposed areas. Save the micro switch lever arm units for protected areas. A decent switch should repeat reliably to within a few thousands.

While not cheap as you wanted such switches will likely be problem free for years.

That is my situation exactly. If I get 1/100 inch ,0.25mm or better repeatablity that will be more than adequate for my needs.