It's nice to see that there are other people out there trying other ideas. I like the BL Touch, it was a mother getting it dialed in, and now that I have it working just right, my EZABL sensor just arrived in the mail. :-) So, I'll give that a try and see what happens.
I have been using the Touchmi for the past 3 weeks without any issues. You are supposed to use a small printed 2mm disk to adjust the probe when you install it.
Hi Michael, great review/tutorial as always. Before I had even seen the video, I checked out the website and manual. It seems a very clunky way of simplifying installation, which doesn't even do that well. Sure, maybe slowing the probing speed may yield better repeatability, but that wasn't necessary for the BL Touch, in fact you sped it up, with no loss of accuracy. As you may remember, I also followed your guides on installing the BL Touch on the Borg (CR-10S4) and it is still serving me perfectly, not one failure. Great to see the WhamBam FBS, I finally managed to buy mine too (also for the Borg) and it is everything I wanted, and then some. The weight saving has meant that I can increase both the jerk and acceleration too. I've also added a 0.8mm nozzle to the Borgs Microswiss hotend, so getting used to the finer tweaks with that. Happy days :-)
@@TeachingTech Thanks mate, much of it is thanks to your tutorials. It seems to be working well for me. Just need to think of something to print with it now. I spend far more time and money on improving/upgrading my printers than actually printing stuff. I thoroughly enjoy it though and that's what counts for me.
Dang, I was really hoping the repeat ability was better. I love the innovation and the idea to remove the extra setup a BLtouch has. Thanks for another great video!
When you do put a BLTouch on your Sidewinder, please post a video about it! Your videos on the Ender 3 have been most helpful and educational, and some facets of the Sidewinder could use the same treatment. Thanks!
Optical ABL solutions have been around for many years now. It is known that light effects the accuracy so the optical sensor can detect the interuption better a very small LED would need to be fitted to help with the issue. As you said probing speed also matters which is a very easy fix in the code.
You could try taking multiple probes in each spot and averaging the result. That should make the bed leveling much more reliable. In Marlin this is done by uncommenting #define MULTIPLE_PROBING 2 in configuration.h and setting the value to 3 or higher. I would love to see a follow up to know if that helped! :)
I think it is a very nice solution proposal because it works with any surface and it could be implemented in old control boards that don't have the extra pin for controling an electromagnet. It just needs 5V and GND. I think problems would be solved soon....
The reason why it doesn't work consistently is because they didn't try it out well enough. An IR sensor is sensitive to the sunlight and some artificial light sources. The solution would be to add an opaque shield around the IR sensor, but besides that, IR sensors are generally not always accurate enough for auto bed leveling measurements. Besides that point, that metal pin is too sharp and will start carving small dips in your print surface after a while. I suggest making it flat like with the 3Dtouch.
I saw a similar solution in a fablab more than 2 years ago. Difference was that initial release of touch probe needed to be done manually, but in the end the pin was electrically raised from the bed. And it worked quite good, but they were using on glass bed. I guess that this design is open source since a while, and not really difficult to make also.
My printer used to use something similar, And worked pretty good. As I had troubles deploying it on a delta, (it was supposed to bump a tower, but before calibration, it doesn't know where the tower is) I do it manually (and dont use it for ever print.) Your's probably inaccarute as it missing a spring to push the pin down when measuring. Using a round shiny metal pin is also hard to detect for the optical sensor. You should test the probe by using M48.
You should check out the pyr0 piez0 auto bed leveling sensor. It's an auto bed leveling sensor developed by one guy. It's not finished yet, but I have heard it's near completion. It has no moving parts and is supposedly more accurate than a bl touch and it works on all surfaces.
Since you're already in Marlin, you can add the G1 Z0.5 step to stow the pin in Marlin's G29_SUCCESS_COMMANDS line in Configuration_adv.h and skip the slicer config.
@@thisiskeithb Hi friend, I installed it on my Ender 3 printer and Touch-Mi with Firmware 1.9.1 and it worked perfectly. But now change the plate for a SKR Mini E3 v2 0 and I want to install Marlin 2.0.1 which is the latest version and I don't know how to configure the TOUCH-MI. Can you tell me if you have the Marlin 2.0.1 configured so that EL Touch-Mi works. Thank you
M48 repeatability test results would be interesting including a standard deviation. Your results remind me of my experiences with the cheap TL touch knock offs... Buy cheap, buy twice.
I'd rather spend the extra money for a genuine bltouch. Abl is such a wonderful feature for me that it's worth it to get something well known and trusted. Interesting design though!
Hi Michael, Great video and thank you so so much for helping. Quick question. I have a CR-10s, with a hotend that leaks (I updated the nozzle to a 5mm (to use wooden filament) and think I may have overtightened the unit) and am considering updating. In your opinion and if you had to buy another creality printer, what would you buy? Note: Whatever I get, auto-bed-levelling is a must-have. Eagerly await your reply. Dujon
I learned my lesson to not cheap out when the BLtouch clone I got was just as bad as this Touch Mi in the repeatability department. I got an original BL Touch V3 a week ago, and it's just perfect. Expensive as hell, though...
Hi, it did not become clear by the video, but did you comment out the G29 and just use the stored settings? My ezabl even show same symptoms as this if I don`t run the G29 in front of every print. I find it to be due to Z-axis not staying in zync on my printer. Did print some blocks for it, after that - the problem where solved on my printer.
Hi, help me deal with BLtouch. I installed everything, set everything up, but there was a problem with (HOME). When you start HOME, then on CORE xy, the carriage goes to the left front corner according to XY - that is, everything is correct, and then the Z axis starts to rise. But the problem is that I do not reach the table with the block, that is, I need my head printer before leaving the Z axis in HOME, first went to the center of the table. Marlin 2.0 SKR1.3
I would paint the top of the pin black so there is less reflection from the top into the photodiode. It should help but YMMV. Blocking the ambient light into the top will probably help too as it will change the amount of beam from the LED that needs to be blocked before the sensor operates. I bought a 3D touch and it was useless compared to the BLtouch I now have. It may have been a fake I have no way to tell.
I have had these same issues with unreliability with a £12 3dtouch from geeetech, tried the probe on an A8, cr10s and a20m. Always similar results, just not reliable enough unless use a 0.3 layer height. With this touch mi, could the magnet no be placed on far x axis, or I know it would take a long time but if we fixed it to the top bar at the top of z axis, there would be less false triggering.
Just a thought here... might the photo sensors reading be affected by room (or your video) lighting setup? Could it be getting bleed in from an external source, as this would vary as it moves to probe other areas or if there was movement between the light source and the sensor.
Howdy Michael! Greetings from Brazil. One of my printers is a Sethi3D FARM (made by a local manufacturer here in Brazil). It doesn't have specifically a Touch Mi, but its Z probe works with the same principle (the actual board PCB looks like an exact copy of it), so I probably think it's configured in Marlin as one (the manufacturer does not release their Marlin source code allegdely to "standartize tech support". Now here's the thing. The printer is out of warranty anyway so I plan on Klipperizing it like I did to my Anycubic Mega X (your tutorials, along with JJ Shankles ones, helped A LOT btw). I might model a mount for a standard BL Touch, but if I could (at least provisionally) make the stock probe work on Klipper, it would be awesome. Any hints on configuring a Touch Mi-style probe on Klipper?
You could do a probe repeatability test with M48 n#, that gives you the mean and standard deviation of n# of measurements of the probe. Just watching the video you can see the slop in the pin relative to the pin housing. I’m guessing that what’s causing the inconsistent probe measurements. Nice idea but poor execution by the looks of it, shame.
This! Please do a repeatability test, and also experiment with probing speed. Also, I wonder if crashing the probe into the printer may have damaged it - it got stuck on the pin, and if it bent the pin or deformed the hole that constrains it it would affect reliability a lot!
Hello friend, on my Ender 3 printer and based on this tutorial I installed the Touch-Mi with the Firmware listed in the description and it worked perfectly. But now change the plate for this: SKR Mini E3 v2 0 and install the Marlin 2.0.1 which is the latest version and I don't know what to do to configure the TOUCH-MI. Can you help me configure the Marlin so that EL Touch-Mi works. Thank you
Looks like an interesting concept, but why no cover over the optical path? Seems to me that ambient light could be bolloxing the accuracy.. I'll stick with my ZedThing ABL.. :-)
Indeed, your right! Even that sensor it's using for sure the infrared light it's still sensitive to ambient light. I'll try to build a closed one just for curiosity and in the price for build it it's right, then the concept might be worthy, what do you think?
@@neagoecatalin879 I would definitely have a go at building a fully closed version.. It's not exactly expensive to build, apart from a few hours design and print time..!!
I use this type of optic sensor for all of my endstops and there is no need for adding enclosure. That black piece with a slit is enclosure. Poor acuracy is result of to much slack between probing rod and frame.
I bought a 3D Touch a while back when ordering some other printer parts from China since it was so cheap I thought I'd give it a try, but it became more hassle than just manually leveling the bed. I used the "Bed Level Visualization" plugin on Octoprint to get a color-coded map of the probed surface that shows high and low spots and tells you on a graph the actual difference in mm. Every time I ran it the probed result looked completely different. Adjusting the sensor with the screw helped a bit but a few days later it started flashing an error that couldn't be cleared and never worked again. It's supposed to go back to a solid blue LED after you clear the alarm and then you are supposed to turn the screw until the LED turns off, then back it off until the LED comes on again to fix the issue, but mine would turn the LED off completely after clearing the alarm and adjusting the screw wouldn't turn the LED back on. It was really frustrating since there is very little information online but eventually I found a forum post detailing the same issue and the answer was "buy a new one"... Rather than buy another piece of junk I ordered a genuine BLTouch instead. I received it a few days ago and the difference is night and day. The solenoid on the genuine sensor is much stronger, the coil has at least twice as much copper wire, and running the color map produced nearly identical results every time (deviation was only +/- 0.01mm). Holding the two side by side the genuine sensor is heavier and overall feels like it's construction is much higher quality. I wouldn't recommend going for the 3DTouch, it's not accurate enough to get good results. The genuine BLTouch is not cheap but from my experience so far it is well worth the money.
TBH... I fully support new companies, local producers (well, I´m in Spain but still nearer as China), but for the hassle that this device and unconsistent results I have to pay double as my bl-touch clone, I prefer the clone... the magnet solution is bonkers and an optical sensor for that...
I've been using a Precision Piezo Orion on my Hypercube for at least half a year. It has had some false positive triggers a few times, it can be a bit tricky to the get the sensitivity just right. Other than that though, I love the Orion. It's at least as consistent as BLTouch in my experience, and the concept of having your actual nozzle act as the probe is a much better solution imho. I just wish it was easier to fit on everything (like a Chimera for instance) so I could use it on all my printers..
BL Touch rocks! This one's accuracy it's also being affected by the ambient light even it's infrared but even if it's closed to the outside, the accuracy it's lower than on the BL. Maybe worth changing the optical sensor with a hall sensor? Think that this concept could be DIY if the price it's lower than for BL Touch (need to calculate but...unlikely).
@@slip1959 I have to admit I dont like the word clone, its used to much to describe a fake product, or an illegal copy of a product . To be fair these things are cheap enough to support the original creator really so let us try to do that or well end up being dictated to by the corporations and the little guys and gals out there just wont bother to make these brilliant add ons anymore.
How unfortunate this has such major drawbacks. I was able to follow this fairly well (the code stuff is still scary to me) and the price was right. Such a shame.
ReactorTV You can use any that fit the configuration (I forget the name). Just do a search on them. Your best bet is 2208 or 2130, although those are the more expensive options. I have the 2130 drivers installed on my Gen L board, and haven’t had any problems as of yet.
@@TeachingTech Maybe they will show the best results with 32bit motherboard and 1/128 microstep? I do not think that a 8bit motherboard with higher than 1/32 microstep is good idea.
you could probably 3d print an insert for the photo detector to make the "view box"? narrower to make it more accurate, i think i might try and make my own, I have everything needed except the magnet and magnetic ball bearing. Both should be easily bought from local stores. Dont have any for of ABL yet so why not try, it should only cost me a few $ to make.
It's more a matter of a repeatability than accuracy. One want to have sensor triggered on the same height, rest is a matter of z offset srtting. In this solution metal rod have short guidence and it's wobble. That can be a source of inconsistence.
@@LoosiuFlying I'm working on a version now that should have very little if no wobble, based off parts i have laying around, ill just have to get some magnets ($8 for 10), a magnetic ball bearing ($10 for 216) and a steel M3 bold/threaded rod. It should just be a matter of getting the clearance perfect. loose enough that the bolt will drop and move without catching, tight enough that it guides the bolt correctly.
My Opinion instead of an optical sensor you can also just let the pin rest on a washer and if the pin is pulled up, it brakes electric contact with the washer, should be very repeatable and accurate
@@badstep495 you would have to either have 2 pads that it bridges, which would cause inaccuracy, or you would need to attach a wire to the pin, restricting its freedom of moment or have some sort of sliding contact, that would again restrict its movements. The optical sensors I have here have a very narrow beam area and will be inside a 3d printed black housing to absorb any refraction. i will also be painting the pin black to help as well. I am looking for a deflection gauge so i can measure repeat ability.
@@m3chanist Yeah, those were all the rage back in 2014. My friend and I made one as well. Though it used a servo that swung a magnet on top of the metal probe to retract it.
We're adding formal support for this in Marlin shortly. Watch out! It will ensure a minimum height for probe deployment. You'll still have to tune movement limits and/or the magnet position so it doesn't drop the probe near the end of the gantry. While the side opposite your endstop is an option, some people like to put their camera trigger there. github.com/MarlinFirmware/Marlin/pull/14101
Seems a very inelegant solution since it has to home each time. Yeah it uses less wires but if you're adding a touch sensor you may as well pay the few extra quid for the Creality kit.
10 minutes of installation video just for you to essentially say "this shits garbage". Creative way to get your viewer retention I guess, as if you opened with that most people wouldn't have watched the installation. RIP 14 minutes.
There is a fair chance that I have an issue I've caused so the instructional part might still be worthwhile. I never clickbait or do annoying things for clicks or retention.
From a repeatability test I saw, the best option is a simple microswitch. Key points: Needs to have a repeatable positive stop position. Remove the metal tab carefully. It's cheap and provides very accurate results. The down side is in solving the automatic deployment issue. You could integrate a servo or something like that, but there in lies the complexity of making it full auto. Semi auto would be a simple magnetic mount that would have a couple stop tabs that keep it positively and repeatedly located.
Love english comment ... Don't know what you doing wrong but i've got a touch mi since 1 year and i get the same value with a bl touch (original). The manual is on the product page ... drive.google.com/file/d/1zAGvagyjzEX3QtdQGDpWTvXLcYBuwUqO/view
Your video needs to be more scientific in its testing no standard deviations or minimum amount of movement for trigger very poor job overall glad I’m not subscribed I didn’t learn anything
It's nice to see that there are other people out there trying other ideas. I like the BL Touch, it was a mother getting it dialed in, and now that I have it working just right, my EZABL sensor just arrived in the mail. :-) So, I'll give that a try and see what happens.
I have been using the Touchmi for the past 3 weeks without any issues.
You are supposed to use a small printed 2mm disk to adjust the probe when you install it.
I used a 2.5mm hex key in the exact same way. The range in the manual is 2 - 2.5mm.
Thank you for the information. I was looking for a bed leveling system. Now this one is out
Hi Michael, great review/tutorial as always. Before I had even seen the video, I checked out the website and manual. It seems a very clunky way of simplifying installation, which doesn't even do that well. Sure, maybe slowing the probing speed may yield better repeatability, but that wasn't necessary for the BL Touch, in fact you sped it up, with no loss of accuracy.
As you may remember, I also followed your guides on installing the BL Touch on the Borg (CR-10S4) and it is still serving me perfectly, not one failure.
Great to see the WhamBam FBS, I finally managed to buy mine too (also for the Borg) and it is everything I wanted, and then some. The weight saving has meant that I can increase both the jerk and acceleration too. I've also added a 0.8mm nozzle to the Borgs Microswiss hotend, so getting used to the finer tweaks with that. Happy days :-)
Your setup is very advanced now!
@@TeachingTech Thanks mate, much of it is thanks to your tutorials. It seems to be working well for me. Just need to think of something to print with it now. I spend far more time and money on improving/upgrading my printers than actually printing stuff. I thoroughly enjoy it though and that's what counts for me.
Dang, I was really hoping the repeat ability was better. I love the innovation and the idea to remove the extra setup a BLtouch has. Thanks for another great video!
When you do put a BLTouch on your Sidewinder, please post a video about it! Your videos on the Ender 3 have been most helpful and educational, and some facets of the Sidewinder could use the same treatment. Thanks!
One heck of a marketing strategy.
Optical ABL solutions have been around for many years now. It is known that light effects the accuracy so the optical sensor can detect the interuption better a very small LED would need to be fitted to help with the issue. As you said probing speed also matters which is a very easy fix in the code.
What an unfortunate or purposely named device
Good idea but at that point this Is a overly complicated bl touch that's both less accurate and more work than its worth
@@kunalshah7539 Leave it to the french...
You could try taking multiple probes in each spot and averaging the result. That should make the bed leveling much more reliable. In Marlin this is done by uncommenting #define MULTIPLE_PROBING 2 in configuration.h and setting the value to 3 or higher. I would love to see a follow up to know if that helped! :)
I imagine it would but I'm impatient and don't like the idea of slowing down the ABL process.
@@TeachingTech Yes that makes sense. It would take a considerable amount of time longer.
I think it is a very nice solution proposal because it works with any surface and it could be implemented in old control boards that don't have the extra pin for controling an electromagnet. It just needs 5V and GND. I think problems would be solved soon....
The reason why it doesn't work consistently is because they didn't try it out well enough. An IR sensor is sensitive to the sunlight and some artificial light sources. The solution would be to add an opaque shield around the IR sensor, but besides that, IR sensors are generally not always accurate enough for auto bed leveling measurements. Besides that point, that metal pin is too sharp and will start carving small dips in your print surface after a while. I suggest making it flat like with the 3Dtouch.
I saw a similar solution in a fablab more than 2 years ago. Difference was that initial release of touch probe needed to be done manually, but in the end the pin was electrically raised from the bed. And it worked quite good, but they were using on glass bed. I guess that this design is open source since a while, and not really difficult to make also.
My printer used to use something similar, And worked pretty good. As I had troubles deploying it on a delta, (it was supposed to bump a tower, but before calibration, it doesn't know where the tower is) I do it manually (and dont use it for ever print.) Your's probably inaccarute as it missing a spring to push the pin down when measuring. Using a round shiny metal pin is also hard to detect for the optical sensor.
You should test the probe by using M48.
You should check out the pyr0 piez0 auto bed leveling sensor. It's an auto bed leveling sensor developed by one guy. It's not finished yet, but I have heard it's near completion. It has no moving parts and is supposedly more accurate than a bl touch and it works on all surfaces.
Thanks for the tip.
No Touchmi for me. BLTouch's have been working fine for all installations for me, and they had been all simple, straightforward installs.
@j mtx The same goes for me too. Still prefer the BL Touch
Since you're already in Marlin, you can add the G1 Z0.5 step to stow the pin in Marlin's G29_SUCCESS_COMMANDS line in Configuration_adv.h and skip the slicer config.
Hi, can you help me, I can't find the Marlin G29_SUCCESS_COMMANDS, 1.9.1. Thank you
@@juancarlos3989 You’ll need to use Marlin 2.0.x.
@@thisiskeithb Hi friend, I installed it on my Ender 3 printer and Touch-Mi with Firmware 1.9.1 and it worked perfectly. But now change the plate for a
SKR Mini E3 v2 0 and I want to install Marlin 2.0.1 which is the latest version and I don't know how to configure the TOUCH-MI.
Can you tell me if you have the Marlin 2.0.1 configured so that EL Touch-Mi works. Thank you
You should try bfptouch, it's cost is very low and not hard to impelent and i think it derserves more love from makers.
M48 repeatability test results would be interesting including a standard deviation. Your results remind me of my experiences with the cheap TL touch knock offs...
Buy cheap, buy twice.
Hi, I'm using the Touch Mi's. Honestly, I'm not that kind of problem.
I'd rather spend the extra money for a genuine bltouch. Abl is such a wonderful feature for me that it's worth it to get something well known and trusted. Interesting design though!
Hi Michael, Great video and thank you so so much for helping. Quick question. I have a CR-10s, with a hotend that leaks (I updated the nozzle to a 5mm (to use wooden filament) and think I may have overtightened the unit) and am considering updating. In your opinion and if you had to buy another creality printer, what would you buy? Note: Whatever I get, auto-bed-levelling is a must-have. Eagerly await your reply. Dujon
I learned my lesson to not cheap out when the BLtouch clone I got was just as bad as this Touch Mi in the repeatability department. I got an original BL Touch V3 a week ago, and it's just perfect. Expensive as hell, though...
touch mi is baguette quality 👌🏻
Why do these manufacturers keep cheaping out using using non-polarized connectors?
Hi, it did not become clear by the video, but did you comment out the G29 and just use the stored settings?
My ezabl even show same symptoms as this if I don`t run the G29 in front of every print.
I find it to be due to Z-axis not staying in zync on my printer.
Did print some blocks for it, after that - the problem where solved on my printer.
Hi, help me deal with BLtouch. I installed everything, set everything up, but there was a problem with (HOME). When you start HOME, then on CORE xy, the carriage goes to the left front corner according to XY - that is, everything is correct, and then the Z axis starts to rise. But the problem is that I do not reach the table with the block, that is, I need my head printer before leaving the Z axis in HOME, first went to the center of the table.
Marlin 2.0 SKR1.3
I would paint the top of the pin black so there is less reflection from the top into the photodiode. It should help but YMMV. Blocking the ambient light into the top will probably help too as it will change the amount of beam from the LED that needs to be blocked before the sensor operates.
I bought a 3D touch and it was useless compared to the BLtouch I now have.
It may have been a fake I have no way to tell.
this is really a smart design,!!!!
wish you would also cover configuration for Duet Reprapfirmware boards
I have had these same issues with unreliability with a £12 3dtouch from geeetech, tried the probe on an A8, cr10s and a20m. Always similar results, just not reliable enough unless use a 0.3 layer height.
With this touch mi, could the magnet no be placed on far x axis, or I know it would take a long time but if we fixed it to the top bar at the top of z axis, there would be less false triggering.
If you can make a video on 3D Touch it will be well appreciated
3:49 better if you try to place the sensor at 2mm (max) above your nozzle when nozzle sticks to the tray.
Just a thought here... might the photo sensors reading be affected by room (or your video) lighting setup? Could it be getting bleed in from an external source, as this would vary as it moves to probe other areas or if there was movement between the light source and the sensor.
That is an interesting idea. I do point the lights at the printer to film.
Howdy Michael! Greetings from Brazil.
One of my printers is a Sethi3D FARM (made by a local manufacturer here in Brazil). It doesn't have specifically a Touch Mi, but its Z probe works with the same principle (the actual board PCB looks like an exact copy of it), so I probably think it's configured in Marlin as one (the manufacturer does not release their Marlin source code allegdely to "standartize tech support".
Now here's the thing. The printer is out of warranty anyway so I plan on Klipperizing it like I did to my Anycubic Mega X (your tutorials, along with JJ Shankles ones, helped A LOT btw). I might model a mount for a standard BL Touch, but if I could (at least provisionally) make the stock probe work on Klipper, it would be awesome.
Any hints on configuring a Touch Mi-style probe on Klipper?
Very interesting engineering, seems a little primitive compared to other systems
The simplicity is part of the beauty for me.
Pallet 2 video coming soon???
Sure hope so!
Been testing with it a lot, don't worry.
You could do a probe repeatability test with M48 n#, that gives you the mean and standard deviation of n# of measurements of the probe. Just watching the video you can see the slop in the pin relative to the pin housing. I’m guessing that what’s causing the inconsistent probe measurements. Nice idea but poor execution by the looks of it, shame.
This! Please do a repeatability test, and also experiment with probing speed. Also, I wonder if crashing the probe into the printer may have damaged it - it got stuck on the pin, and if it bent the pin or deformed the hole that constrains it it would affect reliability a lot!
Fair point. The pin is pretty rigid, but it is possible.
I agree. Lots of slop in the assembly as it wiggles. I'm also curious to see what the standard deviation is.
@@Topy44 Tiene el mismo movimiento que el Bltouch..
Hello friend, on my Ender 3 printer and based on this tutorial I installed the Touch-Mi with the Firmware listed in the description and it worked perfectly. But now change the plate for this:
SKR Mini E3 v2 0 and install the Marlin 2.0.1 which is the latest version and I don't know what to do to configure the TOUCH-MI.
Can you help me configure the Marlin so that EL Touch-Mi works. Thank you
Looks like an interesting concept, but why no cover over the optical path? Seems to me that ambient light could be bolloxing the accuracy..
I'll stick with my ZedThing ABL.. :-)
Indeed, your right! Even that sensor it's using for sure the infrared light it's still sensitive to ambient light. I'll try to build a closed one just for curiosity and in the price for build it it's right, then the concept might be worthy, what do you think?
@@neagoecatalin879 I would definitely have a go at building a fully closed version.. It's not exactly expensive to build, apart from a few hours design and print time..!!
I use this type of optic sensor for all of my endstops and there is no need for adding enclosure. That black piece with a slit is enclosure.
Poor acuracy is result of to much slack between probing rod and frame.
Try it on a bed that does not have a magnetic bed.
When you can buy a BLTouch clone for $13, why buy this for 19 Euros? What about adding UBL to the BLTouch? When are you going to cover this?
is the pin magnetized? are those orings on the bottom and top of pin?
Can you review the 3D Touch in comparison to the bltouch?
it's junk, don't bother. I have it and I have to adjust the z offset almost every print just cause of how inconsistent it's probe accuracy is.
@@Rbanh I have it too ;)
I prefer to support genuine but the choice is for each individual
I bought a 3D Touch a while back when ordering some other printer parts from China since it was so cheap I thought I'd give it a try, but it became more hassle than just manually leveling the bed. I used the "Bed Level Visualization" plugin on Octoprint to get a color-coded map of the probed surface that shows high and low spots and tells you on a graph the actual difference in mm. Every time I ran it the probed result looked completely different. Adjusting the sensor with the screw helped a bit but a few days later it started flashing an error that couldn't be cleared and never worked again. It's supposed to go back to a solid blue LED after you clear the alarm and then you are supposed to turn the screw until the LED turns off, then back it off until the LED comes on again to fix the issue, but mine would turn the LED off completely after clearing the alarm and adjusting the screw wouldn't turn the LED back on. It was really frustrating since there is very little information online but eventually I found a forum post detailing the same issue and the answer was "buy a new one"... Rather than buy another piece of junk I ordered a genuine BLTouch instead.
I received it a few days ago and the difference is night and day. The solenoid on the genuine sensor is much stronger, the coil has at least twice as much copper wire, and running the color map produced nearly identical results every time (deviation was only +/- 0.01mm). Holding the two side by side the genuine sensor is heavier and overall feels like it's construction is much higher quality.
I wouldn't recommend going for the 3DTouch, it's not accurate enough to get good results. The genuine BLTouch is not cheap but from my experience so far it is well worth the money.
That Z offset screen is different than the one on my stock ender 3. Is that from the MKS board?
No just more recent updates to Marlin
@@TeachingTech Awesome! Thanks for all your helpful videos!
Interesting ABL, but somewhat dust sensitive
Thanks for Sharing :-)
TBH... I fully support new companies, local producers (well, I´m in Spain but still nearer as China), but for the hassle that this device and unconsistent results I have to pay double as my bl-touch clone, I prefer the clone... the magnet solution is bonkers and an optical sensor for that...
BL Touch clones is the answer :)
Would be interesting if you could test the system from Precision Piezo (DJDemonD and others on the reprap forum)
I've been using a Precision Piezo Orion on my Hypercube for at least half a year. It has had some false positive triggers a few times, it can be a bit tricky to the get the sensitivity just right.
Other than that though, I love the Orion. It's at least as consistent as BLTouch in my experience, and the concept of having your actual nozzle act as the probe is a much better solution imho. I just wish it was easier to fit on everything (like a Chimera for instance) so I could use it on all my printers..
BL Touch rocks! This one's accuracy it's also being affected by the ambient light even it's infrared but even if it's closed to the outside, the accuracy it's lower than on the BL. Maybe worth changing the optical sensor with a hall sensor? Think that this concept could be DIY if the price it's lower than for BL Touch (need to calculate but...unlikely).
Seems accuracy is a big problem with Toucmi.
PDF link is broken. Good info btw!
huh why i didn't see this before i started remixing the cr-6 mod i seen a few day ago >_>
i have everything to make this like now !!!
can you do a review on the flashforge adventurer 3?
for 12$ you can get a bl clone, so this device has to be dirt cheap to compete
Bl clones also fail. Check this out:
ruclips.net/video/HvpxfCd0zcI/видео.html
@@slip1959 I have to admit I dont like the word clone, its used to much to describe a fake product, or an illegal copy of a product .
To be fair these things are cheap enough to support the original creator really so let us try to do that or well end up being dictated to by the corporations and the little guys and gals out there just wont bother to make these brilliant add ons anymore.
How unfortunate this has such major drawbacks. I was able to follow this fairly well (the code stuff is still scary to me) and the price was right. Such a shame.
I'll stick with the EZABL, I'm not a fan of mechanical levelers.
Yeah I was gonna say, that's not a very accurate system to measure the bed.
add a Servo to Remove the Magnet
Wait for a china copy... 😂. Got the geeetech bl clones, and no issues.
I think enabling MULTIPLE PROBING will fix this
Can you do a video on this with the sidewinder X1 and thank you
It lv8729 stepper drivers are installed on MKS Gen L?
ReactorTV
You can use any that fit the configuration (I forget the name). Just do a search on them. Your best bet is 2208 or 2130, although those are the more expensive options. I have the 2130 drivers installed on my Gen L board, and haven’t had any problems as of yet.
Testing them out. Print fine, but not as quiet as TMC. Wondering if I'm missing something. Video coming.
@@TeachingTech Maybe they will show the best results with 32bit motherboard and 1/128 microstep? I do not think that a 8bit motherboard with higher than 1/32 microstep is good idea.
Well sold me I'm Getting a bl touch
Why can’t I make his vids full screen on iOS?
Those photo detectors are just not accurate enough for this type of use.
you could probably 3d print an insert for the photo detector to make the "view box"? narrower to make it more accurate, i think i might try and make my own, I have everything needed except the magnet and magnetic ball bearing. Both should be easily bought from local stores. Dont have any for of ABL yet so why not try, it should only cost me a few $ to make.
It's more a matter of a repeatability than accuracy. One want to have sensor triggered on the same height, rest is a matter of z offset srtting. In this solution metal rod have short guidence and it's wobble. That can be a source of inconsistence.
@@LoosiuFlying I'm working on a version now that should have very little if no wobble, based off parts i have laying around, ill just have to get some magnets ($8 for 10), a magnetic ball bearing ($10 for 216) and a steel M3 bold/threaded rod.
It should just be a matter of getting the clearance perfect. loose enough that the bolt will drop and move without catching, tight enough that it guides the bolt correctly.
My Opinion instead of an optical sensor you can also just let the pin rest on a washer and if the pin is pulled up, it brakes electric contact with the washer, should be very repeatable and accurate
@@badstep495 you would have to either have 2 pads that it bridges, which would cause inaccuracy, or you would need to attach a wire to the pin, restricting its freedom of moment or have some sort of sliding contact, that would again restrict its movements.
The optical sensors I have here have a very narrow beam area and will be inside a 3d printed black housing to absorb any refraction. i will also be painting the pin black to help as well.
I am looking for a deflection gauge so i can measure repeat ability.
Could you explain how to do an RGB logo frame like the one you have on your background wall? Great videos. Thanks a lot!
He did a video of it already. ruclips.net/video/kLtnRwV2jRk/видео.html
Send me over the translated manuel. I'll finish the translation. Thanks
"buys €10 bltouch from China"
Works better than that, and is still 50% cheaper
@@TheRybka30 welcome to China
@@TheRybka30 Indeed, here is the original, note the date, who did the stealing? ruclips.net/video/UPeNtwEEcww/видео.html
@@m3chanist Yeah, those were all the rage back in 2014. My friend and I made one as well. Though it used a servo that swung a magnet on top of the metal probe to retract it.
But With the BL touch!
We're adding formal support for this in Marlin shortly. Watch out! It will ensure a minimum height for probe deployment. You'll still have to tune movement limits and/or the magnet position so it doesn't drop the probe near the end of the gantry. While the side opposite your endstop is an option, some people like to put their camera trigger there. github.com/MarlinFirmware/Marlin/pull/14101
Thanks, Scott. Appreciate all of your work.
Seems a very inelegant solution since it has to home each time. Yeah it uses less wires but if you're adding a touch sensor you may as well pay the few extra quid for the Creality kit.
BLTouch 50€ & QuickInstall & Works 99,5% Perfect // Touch Mi 20 € & Print Adapters & Configurationchaos & Works like sh#t
Can't touch this.
I'm so glad you you got the MC Hammer reference Michael. :)
huh. reminds me of xiaomi. imagine a xiaomi printer!
10 minutes of installation video just for you to essentially say "this shits garbage". Creative way to get your viewer retention I guess, as if you opened with that most people wouldn't have watched the installation. RIP 14 minutes.
There is a fair chance that I have an issue I've caused so the instructional part might still be worthwhile. I never clickbait or do annoying things for clicks or retention.
3D printed bltouch
So cheap iduction sensors are still the best option.
From a repeatability test I saw, the best option is a simple microswitch. Key points: Needs to have a repeatable positive stop position. Remove the metal tab carefully. It's cheap and provides very accurate results. The down side is in solving the automatic deployment issue. You could integrate a servo or something like that, but there in lies the complexity of making it full auto. Semi auto would be a simple magnetic mount that would have a couple stop tabs that keep it positively and repeatedly located.
That's one weird looking sensor.
Leave it to the French to name a product in such a way as to guarantee a person think of them as perverts... Awesome.
Get the touch mi 2.0 called touch mi daddyyy
Of course France named it "touch mi"
I am French and I cannot support this thing
...this is literally a scuffed bltouch
Love english comment ... Don't know what you doing wrong but i've got a touch mi since 1 year and i get the same value with a bl touch (original). The manual is on the product page ... drive.google.com/file/d/1zAGvagyjzEX3QtdQGDpWTvXLcYBuwUqO/view
Touchmi, tee hee hee.
i can make this in 3 dollar..
P L E A S E stop saying its name
Its creepy
Your video needs to be more scientific in its testing no standard deviations or minimum amount of movement for trigger very poor job overall glad I’m not subscribed I didn’t learn anything
seems quite crappy to be honest