ABL device for 2.4 : need advice !

[YaaJ]

I'm lost among ABL systems.

First of all : no Tap, at least for now.

Want something that does not care about the PEI coating thickness, so the steel sheet can be flipped : this discards inductive and capacitive devices (if I understand correctly). Unfortunately. Was about to order a Beacon probe... This is sad...

Wanted to adapt (redesign) an optomechanical device. Unfortunately, the opto is too large and will not fit. Discarded.

Have an old BLTouch v2, with a cracked ear. There's a mod for a BLTouch (replacing the Omron probe). I'm currently using a BLTouch on my bed slinger, in a 60+°C heated chamber ; no problems (it's far over Antclabs specs !). But the BLT is on the modded Stealthburner side, not behind the hotend ! Maybe could use some insulation, the kind of material we all were using in the old days for the hotends before the silicone socks invasion. Did someone test this mod ?

The Klickies... So many variants that I'm lost. Classic, NG that still requires glue (read some comments), and Unclicky I don't underrstand how it works...

I built my own Klicky variant, with auto Z offset. It's been working really great until one of the magnets decided to move away. Didn't rebuild it, and re-attached the BLT. Spent far too much time designing that and writing macros to get everything working with auto Z offset, don't want to reinvent the wheel for the 2.4 !

Also had a look to the Klicky PCB ; out of stock. Euclid : out of stock. These devices have an advantage : screwed magnets.

At this point, I have to decide. BLTouch or Klicky-ich.

Smallest magnets with a center hole I can source on Amazon are 8x3mm, with M2 countersunk hole. But no polariry in the descriptions. And need the two variants obviously. Found a comment about one product, it seems they come paired. Any experience with these magnets ? Isn't 8mm a bit too large to build a Klicky mod ?

What would you recommand ? BLTouch mod, Klicky, Klicky NG, Unclicky ?

[atrushing]

1 hour ago, YaaJ said:

Want something that does not care about the PEI coating thickness, so the steel sheet can be flipped

This can even be an issue with contact probes as the textured side often needs a different squish factor so the Z-offset will still need to be figured out.

I have had good luck with the Biqu MicroProbe. It's very small and more accurate than the BL Touch. It also doesn't stow after each probe so it is quicker than the BL Touch as well.

[YaaJ]

38 minutes ago, atrushing said:

textured side often needs a different squish factor

Oh yes, forgot about this one... I very recently installed a textured PEI sheet on the bed slinger, and used it 3-4 times only.

38 minutes ago, atrushing said:

doesn't stow after each probe so it is quicker than the BL Touch

The BLTouch doen't have to stow after each probe. Perfect leveling, I don't even multiprobe. Always been happy with the BLT. But had a 3DTouch that was a disaster ! Again, never behind the hotend at high temps ! My printer.cfg :

 

stow_on_each_sample: false
probe_with_touch_mode: true

 

 

38 minutes ago, atrushing said:

Biqu MicroProbe

Didn't think of it...

Jumped into the BTT Git, found SETPs :

Then went to my favorite 3D printer shop (where the LDO kits come from) : available and inexpensive (a lot cheaper than the out of stock Amazon items) !

Just have to put some other things in the basket, due to P&P. Meanwhile, I printed the BLTouch mod (while you were answering...). Some more plastic for the failed parts and abandonned projects grave.

I think the problem is solved !

Thank you !

 

Edited December 12, 2023 by YaaJ

[YaaJ]

Found this video :

The probe is working upside down. Never seen this before (I'd like to find a teardown !) ; already seen the printer in maker's fair videos, the "belts" are inresting (can't remember their name, but there are topics on Reprap Forums).

This probe could be great as a sensor in a Z-stop for auto Z offset. Microswitches being hit or miss. Got genuine highest grade Omron, gold plated (verified after cracking one open) that are one order of magnitude worse than some I unsoldered from the cheapest Reprap endstops I could find ! (did test one, 4 left I didn't test). The very best being salvaged ones from a old worn out Microsoft mouse (a pair of Kailh)...

Of course, already ordered ; at first I put an order for two (one for the Voron, one for the bed slinger), and then removed one (the BLTouch working great). Of course, I found the video after I ordered and had the idea... I don't know if Klipper can emulate a second servo for the deploy/stow command to be put before and after probing the nozzle...)

[atrushing]

4 hours ago, YaaJ said:

Found this video :

Cool vid! It's interesting being able to watch through the glass as the probe triggers.

I can agree with his comments on the IR probe. I have one that works great on my V0.1 on smooth and textured PEI but three others from the same vendor that only have good accuracy on a matte PEI surface. I've even tried custom firmware but that just made it worse..

[TitusADuxass]

20 hours ago, atrushing said:

I have had good luck with the Biqu MicroProbe. It's very small and more accurate than the BL Touch. It also doesn't stow after each probe so it is quicker than the BL Touch as well.

I've got one of these for my v2.4 but have never been able to get it to work with my Manta M8P.

[YaaJ]

It will be connected to a BTT EBB36. Should be there on friday, hopefully I can test it the same day ; currently wiring the toolhead and getting everything routed (it is very different from the Voron design, even more from the LDO wiring guide, and takes a lot of trial and error with special parts design). Also have to slighly modify the X carriage design provided by BTT.

[claudermilk]

20 hours ago, atrushing said:

This can even be an issue with contact probes as the textured side often needs a different squish factor so the Z-offset will still need to be figured out.

I have had good luck with the Biqu MicroProbe. It's very small and more accurate than the BL Touch. It also doesn't stow after each probe so it is quicker than the BL Touch as well.

It sounds like your probe solution is resolved. I had Klicky (OG, then NG, with a brief fling with Unklicky) which worked until I pitched it for Tap when that came out.

For the issue @atrushing mentions, I have some macros and printer variable defined in which I tell the printer which plate is currently in use. Then print_start looks at the variable & applies the pre-defined offset. Works pretty well I think. Check the macros_plate.cfg file in my printer backup GitHub.

[atrushing]

2 hours ago, TitusADuxass said:

I've got one of these for my v2.4 but have never been able to get it to work with my Manta M8P.

I am using one with a Manta M8P. I just plugged it into the Probe port and followed the instructions for Klipper macros in this video.

[YaaJ]

After RTFM, it appears the Microprobe does not emulate a servo the BLTouch way. It simply needs a logic command (HI or LO) on the enable pin ; activate and deploy, or deactivate and stow.

This means that it could also be used as a nozzle sensor for Z endstop/auto Z offset ; simpler as there's no PWM servo signal involved ; I have been using this plugin with great success : https://github.com/TypQxQ/KnobProbe_Klipper/tree/main, first configuration was a auto Z offset for the BLTouch, then used it again for a homebrew Klicky (with dockable_probe by cloakedcode, pull request pending). Of course with a different set of macros, the example being for a tool changer, the Tnn commands made no sense. To be tested. Building a nozzle Z enstop shouldn't be difficult, at worst it will require the help of a spring if it is too sensitive or triggered by the 6mm pin weight... The test/tuning doesn't even require a motherboard or a µC ; just 5V, a resistor and a multimeter.

Yet another project inception. This printer project will never end. Started 3 months ago...

[TitusADuxass]

14 hours ago, atrushing said:

I am using one with a Manta M8P. I just plugged it into the Probe port and followed the instructions for Klipper macros in this video.

Thanks, I'll try it that way. I've trying to get it to work over the SB2209.

[claudermilk]

Printer projects never finish/end, they are just abandoned... 

[YaaJ]

Found a mistake in the BTT mod for the Mircoprobe.

They forgot a fillet. Shame on them.

DHL should deliver tomorrow. Not sure I'll sleep at night. Too excited. And also some ASA.

Edited December 14, 2023 by YaaJ

[YaaJ]

Received the Microprobe. It's a v2. It's cute. For now, no electrical test. Just tested it against the 5mm Z probe pin weight, the Microprobe being upside down. Not being energized, the detent requires more force : it could be used as a high precision Z endstop in a auto Z offset setup, without the help of a spring. Most likely. But could be a repurpose for the small springs that come with the BLTouch for height adjustments and/or crashes prevention.

The tip is not too sharp. Similar to a BLTouch one. Should not catch the low spots on a textured bed.

Only criticism for now : wirings is crap. BLTouch's come with silicone wirings, the BTT microprobe comes with stiff PVC (?) ones. But the tiny JST connector on the probe is the same ! (in theroy 1.25mm pitch if it is not a 1.27 knockoff, but doesn't matter for a 5 pin connector). One can just unplug the device and replace the probe, not touching the wiring itself, or repurpose a salvaged cable.

Currently printing the STEP files, after a minor remix (the missing fillet + M2 heatset inserts for the X endstop)

Thanks again for this suggestion. Hopefully it will fit the needs. At least it started two more projects (remixed X carriage + Z endstop). The 2.4 does not thank you, but on my side, I do...

When I test it as a Z endstop, I'll open another thread.

[YaaJ]

Quick and dirty testing of the probe (as a probe) :

- it looks like stow/retract is level shifting HI/LO LO/HI (rising and falling, not just voltage) ; doesn't work with a switch for example, probably requires a clean signal (sounds legit)
- there's a large distance between the trigger point, and the point on non return (the pin remaining retracted) ; +- 3mm ; great for a Z endstop !
- works fine upside down (we already knew that according to the video)

The models in the repo do not match the v2. Will soon upload better models, the device itself, and slighly better X cariage for the 2.4 + SB + E3D compatible toolheads (same nozzle height) ; already made a prototype

Promising.

Will take some time for the Z endstop experimentations : other priorities (the-project-that-never-ends, it's its new name), and also have to order another probe.

@TitusADuxass isn't there another [probe] section in you printer.cfg ? For an inductive probe for exampe ? Ask me how I lost 3 hours trying to figure why QUERY_PROBE was constantly returning "TRIGGERED" ; the inductive probe was defined by default, and forgot to delete it...

Edited December 17, 2023 by YaaJ

[TitusADuxass]

On 12/17/2023 at 4:45 AM, YaaJ said:

@TitusADuxass isn't there another [probe] section in you printer.cfg ? For an inductive probe for exampe ? Ask me how I lost 3 hours trying to figure why QUERY_PROBE was constantly returning "TRIGGERED" ; the inductive probe was defined by default, and forgot to delete it...

Aaah! Yes, I have TAP set up via the SB2209. I wonder if that takes priority or is read later?

[YaaJ]

Pretty sure.

On my side I've been realizing the pin wasn't configured as expected because whatever I do, the probing pin was always "TRIGGERED", and the DMM was always showing 0V instead of a 3.3V pulled up input. Then asked myself the question : "what the hell is pulling this input down ???". Had a look to the schematics. Found a optocoupler for a inductive pobe sharing this same pin... Jumped to the end of printer.cfg, found a section for the inductive probe ! Commented out and bingo it worked !

(was using the default config file, as I just started with the electronics, and motors/endstops are not even connected)

Edited December 18, 2023 by YaaJ

[selbitschka]

I just love my Voron Tap. As long as the nozzle is clean, I have a script to clean it after each print so it’s ready for the next one. I love most about it the ability of it to work with any bed surface. Since it uses the nozzle as the probe it doesn’t matter what the bed surface is made out of you end up with the same z offset and accurate.

[TimGehres]

If you want a reliable and easy to use Z sensor for any surface just get the ChaoticLab's VORON CNC Tap v2.0 for about $45 and be done with it. You will be very please with the results.

BTW, the way, I set the Z Offset initially using the test patches to get it close and then print the first 10 layers of the CT3D.xyz Benchy from https://www.thingiverse.com/thing:763622 on glass to visually determine the correct Z offset. The following pictures are prints at 0.10 layer height with Z offset at -0.800 for the top Benchy, -0.700 for the middle one and -0.550 for the bottom one. With a layer height of 0.20 the offset is -0.550.

You will have to run you own test to determine the Z offset for your machine.

49CA4EAE-28F1-403D-AEAB-39D1FF6AD9FF_1_201_a.heic

[TimGehres]

Sorry for the confusion on the last post regarding the CNC Tap Z Offset adjustment. I mistakenly referred to the patches, which are actually used for Flow Ratio testing.

For the Z Offset adjustment, I set the z_offset: to zero in the printer.cfg file and then home all axis. Then, use a 0.10mm feeler gauge I move the TH down until the gauge will just slide in with only a little friction. Then I take the Z value displayed in the Dashboard - Toolhead as absolute Z value and  subtract that value from 0.01 to get the correct z_offset. For example in my case, at .10mm the reading was .700mm. So, ( .10 - .70 )= -0.600mm for the actual z_offset. This gets you close. Then I use the Benchy found on https://www.thingiverse.com/thing:763622 to refine the z_offset. as described in the above post.

[YaaJ]

6 grams vs 60, no added overhang, no play. Perfect integration (custom bracket). For now, untested. Depending on the results, will order 3 more.

[WINEDS]

Just a heads up.  Hopefully you don't have issues with the heat set inserts pulling out like I did.  There is an open issue on GitHub which I linked here :

 

https://www.printables.com/model/675632-voron-sb-toolhead-square-nut

Edited December 24, 2023 by WINEDS

[YaaJ]

I am aware of this mod ; but a properly printed Stealthburner does not have the issue.

This Stealthburner has been printing a 0.2 and many other parts, for hundreds of hours, and has absolutely no problems as you can see. It is a worst case : a V6 block (large block, very close to this area) + 50W heater + active heating and up to 65°C chamber last summer. As you can see, there's nothing wrong.
Had the problem with my very first Stealthburner, because the parts were not perfectly flat, and had to overtight the screws. Reprinted it, problem solved.

(it's a mod I made for my bed slinger, were the motor plate and the cariage were printed as one unique part, hence the layers orientation)

The best solution would be heatset inserts on the other side with longer screws, but the longest ones being 50mm according to ISO, it is not an option.

Moreover, when swapping cardridges as I do, square nuts being only 1.75mm thick (= 2 threads), wear will quicly be an issue. Brass inserts are 4mm thick, and the Stealthburner is designed so the 50mm screws use the whole insert thickness (= 5 threads).

[YaaJ]

On 12/24/2023 at 9:32 AM, WINEDS said:

Just a heads up.  Hopefully you don't have issues with the heat set inserts pulling out like I did.  There is an open issue on GitHub which I linked here :

https://www.printables.com/model/675632-voron-sb-toolhead-square-nut

Just discovered this on Amazon France.

M3x55.

It could allow for mounting the inserts on the back side, and solve the problem if it occurs :

https://www.amazon.fr/gp/product/B01N68JLPE/ref=ox_sc_saved_title_1?smid=AQ1IBDB6G2RRD&psc=1

Ordered, will see if a mod can be made.

Edited January 24 by YaaJ

[WINEDS]

I don't think you can't mount the inserts on the back side of the x carriage.  The horizontal bolts which hold the 2 carriage parts together interfere.  Thanks for the link though.   I have canbus installed now via SB2040 and have ordered the companion CNC carriage for the IDM scanner probe.  So my square nut x carriage mod might be redundant for me soon.

I like your combined cw2 motor plate and X carriage by the way.  Looks neat.

 

 

Edited January 25 by WINEDS