Endstop switches: mechanical or opto ?

Hello All,

New to this forum but have been lurking for some months.

I am in the process of upgrading my Mini Kossel Delta to 32 bit Duet

Is there any advantage (accuracy & repeatability etc) in changing the XYZ top micro-switches for slotted optos

Now would be a convenient time as I am removing the old electronics and tidying up the wiring.


Many Thanks

JuJuDelta.
Warwickshire, U.K.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
My setup:
T3DP3D Mini Kossel
Arduino MEGA 2560 Ramps 1.4
Slic3r
Raspberry Pi Octoprint WiFi server
Upgrading to Duet & DC42 IR Z sensor

On the large Kossel described in my blog, I still use the microswitches that came with my Mini Kossel. They give results reproducible to better than 0.02mm, so I don't see any need to replace them. I am developing some small Hall effect homing sensors to try instead, but that is out of interest rather than necessity.

Additionally, with the Duet firmware, auto calibration is so fast that I auto calibrate every time I turn the machine on. If you avoid homing after auto calibration, then any inconsistency in the homing switch trigger positions is taken care of by the auto calibration.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I´ve mounted opto-endswitches in my delta, because that way the carriers can pass by the endswitch without breaking them.
That is important, if you want to print full width at full height. Then the carrier needs to move higher than the endswitch.

People who mount the endswitch at the highest possible position and then brag about their printheight are betraying themselves.
-Olaf

Hi,

Both do the same thing, but optoswitch are more complicated to wire and to set up.
Anyways, you don't want to put your switches on the way;
or you may destroy them if something goes wrong.
Whatever this is opto or regular.
Use the type you have on the shelf.

++JM

Why not go with a cheap set of hall switches and some neodinum magnets. They are working well for me. Optical switches are also cheap. But cant beat less than 20c per chip on hall effect switches. One thing I noticed with my mechanical switches was fatigue of the sprint arm.

Here is how I built mine and how I tried to measure their accuracy.

[www.youtube.com]

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

Almost all CNC machines use mechanical endstops for the axes. Some of these machines cost upwards of $100k and much more. If mechanical works good enough for those its good enough for my $500 printer. If opto was more reliable or more accurate i gaurantee they would be using those instead.

Hall effect for the win.

Quote
GITRDUN
Almost all CNC machines use mechanical endstops for the axes. Some of these machines cost upwards of $100k and much more. If mechanical works good enough for those its good enough for my $500 printer. If opto was more reliable or more accurate i gaurantee they would be using those instead.

One more thing that those CNC machines have is a service contract. Does your 500$ printer have one? Hall effect is as reliable as mechanical with no fatigue points. One more thing, the switches used by those CNC machines are of much higher quality that the crap we buy for these 500$ machines. My .02c.....

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

The Homing Mechanisms on our 2 large scale CNC are magnetic sensors. Using optical endstops on a machine where dust is likely to be an issue is just crazy anyway. Mechanical is fine imo.

I've built my Prusa i3 Rework with mechanical one (talking Z axis here, I think there is absolutelly no need to consider optoswitches for X/Y). It was doing fine - repeatability, accuracy, just quite a hassle to adjust every time. Although it's more of an issue of proper mounting/adjusting mechanism.

Of course as a first-time-builder of this amazing thing, I started to upgrade it. And optostop for Z axis proved to be a "downgrade" for me. Accuracy/repeatability got knocked out quite bad - I think it's due to the fact, that the optoswitch, in reality, is not a sharp on/off switch. You can even see the indicator LED fade in/out as you move an object into slot. And somewhere in this zone occurs "the switch". To make the long story short - lost all the "must-haves" (accuraccy/reapeatability), gained no benefits.

Actually, there are none of them offered by optos, anyway - they're NOT more precise, and NOT more durable (even small amounts of dust will be a problem over the long run). So as "cool" as it seemed to have them installed - I vote a big fat NO for them.

Next step from mechanical switches (to remove mechanical wear/fatigue element) should be hall sensor, as already mentioned. I'll test those out myself and let know the results here :]

Quote
GITRDUN
Almost all CNC machines use mechanical endstops for the axes. Some of these machines cost upwards of $100k and much more. If mechanical works good enough for those its good enough for my $500 printer. If opto was more reliable or more accurate i gaurantee they would be using those instead.

Just so you know, cnc equipment use the stop as a reference, they only use the stop so they can get within 1 revolution, and then go to an encoder count on the servo. The servo motor holds the accuracy, not the end stop

It's common to use a Z probe for Z homing instead of an endstop switch.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Old thread but still a pertinent question. So a couple of things not said above.

mechanical switches are classic example of the application of positive feedback. Once they get to a certain point they snap in the opposite state because a spring reaches a 'tipping point' and applies even more forward force. This gives a relatively fast and reliable transition.


opto sensors gives a continuous analogue output, so if you just wire them into a digital IO pin somewhere it will be very prone to noise and rather arbitrary triggering levels. They are basically an opto-diode set up opposite an opto-transistor. The solution is to add some +ve f/b. The simplest way is probably a Schmitt trigger circuit next to the opto.

Hall effect devices are also analogue though a purpose built position detector would presumably have some +ve f/b cct around it to provide a digital output.

With mechanical switches, quality is very important in determining whether they trip reliably and consistently. It's similar to click-stop torque-wrenches: pro-quality ones are very repeatable; however, cheap ones give less repeatable torque setting than a basic torsion bar model. Often I can feel by hand that I am not getting the same torque on successive bolts with cheap devices and won't waste my time using them.

Switches also suffer from switch bounce, which is totally unpredictable. Again quality is key to reduces the effect.

I would suggest anyone posting what does and does not work for them include a reference to the device being referred to. There is no one-size-fits-all answer to all switches or all optos etc.

Edited 2 time(s). Last edit at 04/08/2018 08:28AM by wrapperz.