M2 Auto-Leveling Kit (Beta)

[Jules]

Thanks for the info Josh. And thank you for all the hard work !

Second that!

[PcS]

I have bolted the spider directly to the spider bearing plate and converted the z probe to 24v. Seems more accurate with the added voltage. The quality coming off the machine is as good or better than the best bed leveling I have ever done. Never going back .I also tried it with a really outta level bed and it handled it with no problem. Other than it was so far out the while print was slowed slightly so the z stage could keep up. I have a video of it I will post later.

[PcS]

https://goo.gl/photos/cC6mYANBFPhdmphg9

Image of spider.

https://goo.gl/photos/rr97mVeLAE4dVvXC8

video of badly leveled bed.

[jimc]

well that definitely shows it working lol. very nice. once the whole setup is out of beta i will have to get 4 of these.

[PcS]

well that definitely shows it working lol. very nice. once the whole setup is out of beta i will have to get 4 of these.

As far as I can see it is ready for prime time. I will say though I think the probe works more accurately on 24v .

[Matt_Sharkey]

I have a feeling this is answered already but... Is the leveling procedure firmware based or is it a feature of S3D?

I have my MIC6 on the way, and this would be the next logical step I think, but it may mean it's time for me to fork out for the S3D license as well...

[jimc]

Its firmware not s3d. You should kick out for the s3d lic though. You really cant go wrong there.

[PcS]

What he said !!! ^^^^^^^^^

[innkeeper]

I've been following this thread for a bit... frankly i have some concerns about the approach.

to preface...i've been looking at the inductive solution for a bit as a retrofit on other printers..and now the m2 also.

The issues i see are:

• 5v is not reliable over time for this probe. what many do for other printers is run at a higher voltage, then use a voltage divider at the output to get the right 5v level out for the given supply voltage.
• for anyone running heated bed, the sensor is only rated to 50c, and there have been quite a few failures reported for those running heated beds over time.
• When it does fail, failures are not just on and off, but they either become less sensitive or erratic. (research this on other forums you will find these threads).
• Actual usage of the inductive probe, if you change beds, your z offset will have to change .. so in my case, i plan to swap off zebra plate, mic-6 and glass. so each time id have to modify the z offset, that seems to be a pain in the butt, and i know i will forget to do it.

My personal opinion is detecting the actual surface is much better over time than an inductive detection which is detecting the presence of metal, not its surface.

so id think the mechanical probe solution is better as you don't have any of those issues.

for my needs i am working on an implementation of a 3rd way, which is purely optical which would detect the surface of whatever the build plate is by reflection off the surface. the electronics shouldn't have any issues with thermal breakdown. - kind of the best of both words, non touch and detects the actual surface... still waiting on some parts...

please let me know if you think i am off totally base....

[Bratag]

I've been following this thread for a bit... frankly i have some concerns about the approach.

to preface...i've been looking at the inductive solution for a bit as a retrofit on other printers..and now the m2 also.

The issues i see are:

• 5v is not reliable over time for this probe. what many do for other printers is run at a higher voltage, then use a voltage divider at the output to get the right 5v level out for the given supply voltage.
• for anyone running heated bed, the sensor is only rated to 50c, and there have been quite a few failures reported for those running heated beds over time.
• When it does fail, failures are not just on and off, but they either become less sensitive or erratic. (research this on other forums you will find these threads).
• Actual usage of the inductive probe, if you change beds, your z offset will have to change .. so in my case, i plan to swap off zebra plate, mic-6 and glass. so each time id have to modify the z offset, that seems to be a pain in the butt, and i know i will forget to do it.

My personal opinion is detecting the actual surface is much better over time than an inductive detection which is detecting the presence of metal, not its surface.

so id think the mechanical probe solution is better as you don't have any of those issues.

for my needs i am working on an implementation of a 3rd way, which is purely optical which would detect the surface of whatever the build plate is by reflection off the surface. the electronics shouldn't have any issues with thermal breakdown. - kind of the best of both words, non touch and detects the actual surface... still waiting on some parts...

please let me know if you think i am off totally base....

Can't speak to all the issues you stated but as far as the temperature goes I can say that I don't think that will be an issue. My probe is at least 2-3 mm above the print bed and its really only that close while doing the initial G29 sensing grid, not enough time for the heated bed to increase its temp a great deal - especially over an air barrier. Its mounted far enough away from the hotend that it shouldn't be getting too hot from that.

That being said I think another type of sensor, like the one you mentioned is worth looking into. Had not considered an optical solution, would it work for a matte grey aluminum plate, not the most reflective surface.