Could The CHEAPEST 3D Printer Bed Probe Also Be The BEST? Everything You Need To Know About Klicky
Vložit
- čas přidán 10. 06. 2023
- Is it possible that one probe could be both the best performing and the most economical? The Klicky probe might just be! In this video I examine this novel stow-and-go bed probe and provide a full walkthrough on assembly and firmware configuration, using the Troodon 2.0 3D printer as a test case.
Join me on Patreon: patreon.com/user?u=80746241
Watch the full Troodon 2.0 series: • Troodon 2.0
Buy the Troodon 2.0 (Formbot): www.formbot3d.com/products/tr...
Buy a supercharged Troodon with Stealthburner pre-installed (Tiny Machines 3D): www.tinymachines3d.com/produc...
Buy my 3D models: ygk3d.com/collections/digital...
YGK3D GitHub (Troodon 2.0 RepRap Klicky Configuration): github.com/YGK3D/Troodon-V2-R...
YGK3D GitHub (Troodon 2.0 Klipper Klicky Configuration): github.com/YGK3D/Troodon-2-Kl...
TeamGloomy GitHub: github.com/TeamGloomy/Troodon-V2
Klicky GitHub: github.com/jlas1/Klicky-Probe
PCB Klicky GitHub: github.com/tanaes/whopping_Vo...
Klipper Auto Z: github.com/protoloft/klipper_...
RepRap Auto Z: github.com/pRINTERnOODLE/Auto...
pRINTERnOODLE GitHub (Voron 2.4 RRF Klicky): github.com/pRINTERnOODLE/RRF-...
Copy & Paste Snippets:
Klipper Klicky Install:
cd ~/klipper_config/
wget raw.githubusercontent.com/jla...
unzip Klipper_macros.zip
Klipper Klicky AutoZ:
cd ~
git clone github.com/protoloft/klipper_...
./klipper_z_calibration/install.sh - Věda a technologie
IMPORTANT: @5:07 when I show the order of wires, I have mixed up the labels for power and ground. The ground wire is blue and the power wire is brown. The correct ordering is shown below, from left to right, with the connector facing towards you (as in the video).
Klicky PCB: Power (Brown), Signal (Black), Ground (Blue). Extruder PCB: Power (Brown) Ground (Blue), Signal (Black).
The detail of this is amazing! Thanks for putting in all the effort.
Thank you!
thanks for including all the links to everything makes following/duplicating the work soo much easier.
What a thorough breakdown on implementing a clicky probe.
I really love your video style.
Thanks for featuring my Repo!
Hopefully everyone can see just how easy and effective RepRap/Clicky setup is.
*pRINTERnOODLE has entered the chat*
Thanks for your efforts and for providing these resources back to the community!
Very good video! I dont have a Troodon, and I didn't plan on getting one, but I can tell alot of effort was put into this video
Thank you very much!
Tank you for this another great Tutorial on Troodon v2.0.
Not very simple for the beginner to understand (frenchy guys😅) but step by step it s work ! thank you man
I love printing TPU on glass. It's very reliable and there is no mess (glue stick, etc.) or fuss (removable flexible print surfaces, etc.) Klicky is clever but the one bed probe option you didn't mention is using the nozzle to probe the bed and using force sensors under the bed to detect the nozzle contact. This seems to have all of the advantages and none of the disadvantages. The bed and nozzle can be preheated so they're directly probing in their heat expanded state. The actual nozzle and bed height are probed relative to each other without introducing the error of a Z offset for the nozzle relative to a separate probe. There is no added hardware on the hot end to add weight that decreases print speed and no need for the complexity of a dockable probe sensor. Klicky is clever but I'd rather add force sensors under the bed so I can directly probe with the nozzle.
yep force sensors under the bed is optimal, especially when you want to keep the toolhead as light as possible for fast corexy printing.
I built one. they degrade badly in a heated chamber. For strain gauges the full bridge Wheaton strain gauges is only glued to the compliant mechanism with super glue which melts the moment the chamber gets near 60c. And the ones in the bambu labs uses Piezo discs. which can't handle prolonged contact and susceptible to creep. I have moved on to using Beacon. being able to do a high resolution scan of the bed in 9 seconds is better than clicking on the bed. and printing on glass with TPU yeah I have lots of chipped and damaged glass beds now 😅
@@ShinobubuAre there no glues that survive both the “High” temperature and the temperature fluctuations that can be used with staring gauges?
@@ShinobubuBeyond the glue issue, doesn't the nozzle often have plastic remnants stuck, causing the measurement to be off by that plastic layer thickness? In this case, a Klicky probe has an advantage in that it has no plastic remnants on it. But all that docking, being mounted away from the nozzle on a piece that can flex, etc, would seem to introduce error. Then again, plastic remnants remain on the build plate, as well. Maybe if we took many measurements, we could average out most of the error. Or could one combine two methods to get even better accuracy?
a led optical radar reflection amplitude voltage would be the simplest, you also get the current distance, of the print bed and the print head, well light attenuates over air at inverse of the square distance. ironically it can be detection of the usual illumination led light, add a second led to the print head.
I am curiouse if this would also work for a LDO Voron 2.4 RevC. I used the RRP configurator for the first steps and want now to include the Klicky probe. But most of you configuration files are a little bit different from mine :D
I have run tests between a few different types of probes over the years and the difference in accuracy and repeatability was not very much and they were all capable of giving me a good usable mesh, I have stuck with the BL Touch on some machines and stuck with the stock Pinda on my Prusa MKIIIs
Yeah, it seems that they’re all pretty comparable in terms of accuracy and repeatability. As compared to BL Touch, Klicky has the advantage of being cheaper. I haven’t heard of an Auto Z implementation for BL Touch but it seems feasible. The advantage over an inductive probe is that it measures the bed directly so it can be used with all bed surfaces. And with the addition of Auto Z you won’t need to re-calculate Z offset when changing nozzles or build plates. But this also relies on having a pin end stop, which not all printers have.
have you tried the beacon? it's $80 for a reason
@@angrydragonslayer No I have seen it in operation though, I am happy with what I am using but the next printer I build I might use one
Is it a simple way to add filament sensor on Troodon 2.0 print head (without need to unmount cable chain)? As fillament sensor is far from printhead and if filament breakes somewhere in middle of tube, printer stil prints - air. Thank you.
Wow great video.
Thanks man!
This was already done, it's called the Euclid Probe. Heavily documented and well supported, same magnet idea as well. Someone said the third wire isn't needed, but it definitely is. GND, +5/12/24V, and Signal. Having a dedicated signal wire means it's possible to detect if the probe hasn't been successfully deployed/picked up at the dock. Having just +5V/12V/24V means you have a floating signal when the switch is open, which can pose problems with certain boards not configured with an internal/on-board pulldown.
Horseskat, you 100% don't need three wires to do what you say, in RRF with conditional g code you can intergoate the switch state and know if the switch is attached or not, you can do all of that with no need for anything other than GND/signal wires, you can also use rgb leds to change colours according to whether the probe is attached or not and also displays when either the probe (and or) the z stop is triggered all with only two wires to each switch (well at least those who know what they are doing can do all that. And "elicud" is an over priced idea stolen from an idea by annex engineering, you can make your own collectable probe literally for cents
Current project for me right now what was the spacer for the auto z endstop 3mm more? I might just order a 30mm but it will take some time to come I bet. Great guide as always.
I believe that’s what it was. I ended up finding a suitable size ball bearing and dropped that in before the pin, but a printed spacer would work too.
Thanks!
It's a very informative video. Excellent work! Here are a few constructive criticisms.
1. After three errors in the dialog, I suggest a re-record instead of onscreen corrections. I know it's extra work, but for something this information-dense, you add a lot of unnecessary extra confusion.
2. I would consider separating Kilipper and RepRap configurations into different video chapters so people can skip what they do not need. Going back and forth between the two firmware versions gets confusing.
3. Adding an overlaid outline that shows all the steps with highlights on each step as you complete them would be excellent. I’ve seen you do many onscreen graphics in previous videos; it is within your skill set.
4. Include that same outline as a PDF with live links in the video description. It would make a big pedagogical difference.
For me, these suggestions would be challenging as I am not as skilled of an editor as you are. You have the chops to add these and further improvements.
P.S. Slow the cuts a bit when showing screenshots of code replacements. We can pause, but consider the shot you go to; perhaps that B-roll can wait.
Thanks for the feedback and suggestions!
I've heard the rubber booty needs to be removed from the actual probe tip area or it can mess with readings.
How big is the pin spacer
You are a legend
Any plans to do a session on going to TAP for mesh level on tge Troodon 2
Yes. That’s something I’ve considered for a future video.
@@ygk3d Im.makimg myself crazy deciding on building a Voron or going Troodon it's a big step up from my Ender 3 v2
I wrestled with klicky for about week and went back to the inductive probe. It just wouldn't complete a bed mesh no matter how slow I went
I can’t figure out why I do probe calibrate save config I get an z offset must be specified??? Huh? Why klipper why what does your error mean
Hey! Wondering if you would be willing to install the enraged rabbit carrot feeder on this machine. I would love to see it. I would even pay for access to this content behind the Patreon paywall :)
I have one in hand and am planning to document the build process. It’s quite involved so might be a while before it sees the light of day..
For this money I've got a knock-off B:-Touch probe that uses an electromagnet and Hall effect sensor. The only problem with it was the fact I had to make my own cable and replace connector on Ender 3 V2 board. And I had to design my own mounting piece as the available design didn't quite match the lenght of probe...
The DIY version of the Klicky probe is less than $10 in parts, so cheaper than even a clone BL Touch.
I might be missing something, but your wiring on the XH2.54 connector to the PCB clicker board does not seem to be correct. On the Github page , it states, "Double check wiring, make sure you've connected your wires correctly to the probe. With the probe installed in the toolhead, they should be Ground / Signal / +V going from back to front. " This means your ground and +V should be switched.
Thanks for catching this. I had mistakenly interchanged the labeling of ground and power. I have issued an update in the pinned comment.
Still figuring things out. My switch_offset: 0.246,but it was too low. I had to baby step it to -0.493. Do I subtract the -0.493 from the 0.246?
You can't look at the absolute value of the z offset to determine the change. You need to determine the difference between the original proposed z offset and the final value after baby stepping. For instance, if auto Z gives you a z offset of -0.732 and you baby stepped it to -0.493 (raising Z), you should add -0.732 - (-0.439) = -0.388 to your switch offset. This will make the switch offset value smaller and therefore make the final z value further from the bed.
Probaby yes , do it and do a test but not over yuor bed ,yuo dont want a crash, i put a cardboard box and i home my printer over that when i setup my probe( bltouch clone, 3dtouch), in this way , i nknow if something go wrong i cant damage anything, always put yuor hand in the off swithc
@@ygk3d Ok,I'm trying to calculate those numbers to get the -0.388 and I'm nowhere near it. Then I have the way I need to adjust different numbers if needed.
fun fact it was originally called the quickdraw made by an annex member, a voron member accidentally took it then someone else proper took it, added a magnet and called it klicky
Don't forget about the euclid probe. Klicky also was based off the euclid.
I gotta say my Chaoticlab CNC Tap for my Voron, is amazing, I don't have to do anything, I hit start, I have an ABM macro, and it goes, perfect layers, over 500 hours on it now, I did have one issue where the rail was coming loose, but loctite fixed that right up! Deviation on it is around .003 at 100 cycles. around .0004 at 10 cycles.
Precision Piezo is still better. Nothing to remove. Just use the extruder nozzle as the probe.
Offset -1.118 is outside the configured range of min=-1.000 and max=1.000 I keep getting this error I'm gonna guess i missed something
Try increasing max_deviation in [z_calibration].
I got it working i think lol when you do the z calibration I folloew the video but I can't see if the probe is still attached the the nozzle im gonna says its not when your setting the offseet and jogging the nozzle down to the bed.@@ygk3d
@@redline3036 It's not attached at that step.
I think I know whats going on your using the plunger style switch and im usnig the version that doesnt have the rubber plunger. Also i have the revo six which the hotend is 62.4 not 62 so im trying to figure out where to change those settings correctly klipper is fun but it takes alot of time to figure stuff out for sure..@@ygk3d
You know, we did this 10 years ago now with an allen wrench. Look up Self Deploying Allen Key Probe for a guy who did it recently, but this is an overpriced version of something people did long ago.
It was designed to build yourself with just the cost of the switch and a couple magnets (and some spare wire). If you've built a printer, you probably have all of the parts on hand. The purchased version is just available to make it easier for beginners or those too lazy to make the probe assemblies themselves.
Edit: corrected typos
@@jettachamp26 exactly. It can be done for less than $10 with the DIY version.
Ah, now you need to look at the unclicky probe. It is even better.
I’ll check it out. Thanks!
@@ygk3d I'm partially joking but it is an interesting project. Basically the clicky probe but without the micro-switch. The project boasts a much more accurate reading and more durability.
On a more serious note, there is a variant of the klicky-probe )I forget what it is called, you should find it as a sibling project of the unlicky probe) that uses a two part probe body and I find it works much better on my voron than the default clicky probe. It also assembles easier.
You don’t need 3 wires. The extra wire are just for the led.
Right. But that’s one of the main advantages of PCB Klicky.
Why (for RRF) Has this been configured in such unnecessarily complicated and downright messy way? I am genuinely interested in the thought process that someone concluded that things needed to be like this. I can catagorically state and prove that you need 99.9% less of the mess.
I have implemented this feature using a dockable probe and a pin z endstop, with ONE, I REPEAT ONE 64 line macro that requires NO conditional g code, the macro saves the auto defined z offset in the config_override.g folder using M500 P31 after it has done what's has been told to do.
It makes it look like some klipper fan boy purposefully created this mess to make RRF look overly complicated.
Piezzo and load cells still seem simpler and more precise
Prusa, Bambu, Creality have all moved to nozzle probing.
Ultimaker has been doing it for years
There is a small problem in your code:
In unklicky.g are some lines which occure problems:
G1 X60 ; wipe off
G90 ; absolute
M98 P"/macros/Klicky/moveto/clickstage.g"
That means after unklicky the head will move 60mm right and than again to this coordinates. That means it would take the probe after putting it in the dock.