Has anyone purchased and used BL Touch for Mesh Leveling? I want to know your experiences.
Printable View
Has anyone purchased and used BL Touch for Mesh Leveling? I want to know your experiences.
I've now got a VERY nice printer with a big bed. It has a BLTouch probe. I'm getting all geared up to bring up the devel_UBL branch of Marlin on it. I'm going to be using it as the development platform for devel_ubl going forward.
(It is more likely than not you don't know what that is.... This branch is going to bring all flavors of Auto Bed Leveling to all types of printers. Right now we are still missing the n x n Grid based. But it doesn't matter. What is there works so well, it doesn't matter. And the foundation of the system is a High Resolution Mesh leveling scheme.)
I'm not overly impressed with the 'Repeatability' of the BLTouch probe. But it is plenty accurate to bring up the UBL branch if you so desire. And really... the work EPatel did to create the Mesh Bed Leveling technology is impressive. I left as much of it in place at the lower levels of the code as possible (just out of respect for his efforts). But if you are thinking you might want to check out Mesh Bed Leveling.... For sure you want to bring up the UBL Marlin branch!
There are some people getting good results with BLTouch probes. My impression is a lot of people are having trouble with them. If you are trying to make the decision whether to do a BLTouch or a Servo with a probe leg... My recommendation is go with the servo and probe leg.
Yes. That is true. It is one of those things where in hindsight it should have been called something else.
As I go digging on the various dark corners of the internet, the BL Touch seems to get a *lot* of attention in terms of setting up the firmware. There are reams and reams of yack about this or that setting in this or that sub-sub-sub version of Marlin or some other firmware flavor. I see very little information on tuning the probe it's self. There also seem to be a number of pretty shaky looking mounts out there. Based on servo arm stuff, any mount that is not 100% stable will have issues. The next issue is information in general about just how far from the hot end *any* bed level probe can be. Closer is obviously better. I'd bet that one axis is more important than the other on this printer and less important on another printer design.
No, none of that is really news. It does explain to some degree why we all seem to be a bit confused about just how well this gizmo does or does not perform ....
Truth in lending: Yes, I have one on order. I'm on my 2,873 version of the extruder and mount. Only about 10% of those have made it to the "print it and see" stage :) .... at least that's the way it feels right now.
Bob
Bob, if you load up the RC-8 (Release Candidate #8) of Marlin, there is an easy to configure BLTouch option. You pretty much just tell it you have a BLTouch, and give it the correct Offsets from the nozzle and you are done.
https://github.com/MarlinFirmware/Marlin/tree/RCBugFix
Since RC-8 is a few days(?) old, what's been posted over the last 12 months has been struggling with a variety of RC-X issues as well as basic "this is my first bed level" stuff. Hopefully the word will get out on RC-8 since it does indeed take care of roughly 95% of the stuff that has been posted with about three lines to modify.
Bob
And incidentally... RC-8 should be almost equivalent to RCBugFix. At this point in time, it probably doesn't matter which one you decide to try. But the RCBugFix branch should evolve into the 'Stable' branch if we don't find too many new bugs.
I believe so... But I don't know the answer for sure. It mostly depends on how well RC-8 goes.
Hi
Well, it would be a really good idea to get a stable release out there. There have been a *lot* of changes since the last one.....
Truth in lending: I seem to have errr ... a piece of firmware of my own .... that ... errr ... should have been finalized back in June .... just a few more tweaks. It's not like lives depend on it ... ummm ....errr .. they do ... YIKES !!!
Bob
I may purchase one or two of these units and attempt to implement Bed Leveling/Mesh Leveling over my sabbatical. I will most likely also switch firmware's from the Makerfarm default.
I hear you! And in fact... In general, new features should not be added to a 'Release Candidate'. But the argument being made by the people that will be supporting the new 'Stable' branch is they want things (and especially the code) cleaned up before it goes 'Stable' because they will have to live with it for a long time.
So, while I'm philosophically opposed to adding new code or even changing code that is working in a 'Release Candidate', I do see their point. The good news is these people are starting to feel good enough about the current code base I think Christmas is a definite possibility.
to respond to the original post - My brother and I installed a BLtouch on his 10" I3v. we set it up like a servo and used the angles given by BL touch's directions. It worked fine. it is screwed to the underside of the x carriage with some spacers to get it at the right elevation in relation to the nozzle. I programmed the reset angle into a button on pronterface and it works very well. This is a version of marlin compiled more than a year ago, not sure which one but nothing current.
Thanks for your input!
I have been manually leveling my bed for 3 years now and never had a problem. My concern with ABL/Mesh Leveling is the fact you are relying upon a servo and its movement to ensure you get repeatable results. That is why the BL Touch intrigues me because the only part that moves is the pin which theoretically should provide better results.
I will also have a month off of work to get everything dialed in.
Actually... I think it can be the opposite of that. If you design your probe leg such that when it is straight down, the switch lever is a little bit off center what will happen is the probe leg will get a slight rotational force (toque) applied to it when you are actually depressing the lever arm. If you make it so the probe leg hits a stop when fully deployed, you just tell the servo to go a few degrees past that point and turn off. From then on, all probed points just gently press the probe leg against the stop.
With that kind of a setup, you can get better results with a micro-switch than with an inductive probe. The micro-switches trigger at the same place every time.
Hi
At least the last time I went through the servo leg approach, simply finding a servo that was temperature stable was a major chore. I wound up with a collection of servos that drifted between 30 and 90 degrees as the temp changed from room up to 100C on the bed. Even with the one I finally used the change was around 10 degrees between an 80C bed and a 110C bed. Not a lot of fun.
I have been running my Pegasus for about six moths now. I find that it needs a re-level roughly every month or so. Each time it's off by at least 0.05 mm on a couple of points. Since that's half of a 0.1 mm layer ( or 1/4 of a 0.2) .. it's a big deal. Yes, I could over-extrude the bottom layer and make things less sensitive. I don't like ugly bottoms on parts .....
Bob
not yet. I'll learning about it
That was my concern with the servo arm and is the exact reason why I have not used it yet. I saw a video of the BL Touch the other day and loved the idea and design. While I have a month of of work, I thought I might as well try it and get it dialed in. If you use Nylon Lock Nuts on your bed, manual bed leveling should stay "level" for a while like you mentioned. Now if I switch to ABL/Mesh leveling, would you replace the springs on the bed with spacers?
I can't do this because I need to have springs on my bed. (Any error in the software being developed can crash the nozzle and I need a little bit of give.)
But if it were not for that, I would mount the bed with no springs on a rigid plate. And if you are thinking about bringing up a Mesh bed leveling scheme, This is the Cat's Meow right now: https://github.com/MarlinFirmware/Marlin/tree/devel-ubl It is a development branch (as opposed to a 'stable' branch). But it works very well.
> What leveling scheme should I use and what is the best stable version of Marlin for this?
It depends on what your need. The various bed leveling schemes all offer different advantages. The UBL (Unified Bed Leveling) branch is trying to bring the best of all the different techniques to all printer types. It is trying to make it so you don't have to pick and choose. Right now, it still doesn't work on Delta printers but that is its primary focus. It will work there soon enough. That is what it was designed to help. And so far, it doesn't do the Grid Based Leveling with a Least Squares fit on the sampled points. It will do that soon enough also, but the truth is, it isn't even needed.
Seriously... I went to to the Dollar store and got the worst piece of glass I could find. I found a picture frame and when I held it up to the light I started giggling. I got a piece of glass that has huge ups and downs on it. And one side is thicker than the other side.
But here is the important part: I can get a perfect first layer across the entire piece of glass every time. It doesn't matter where I put the part, or how it stretches across the glass. I get perfect adhesion on the first layer every time. And that piece of glass I'm printing on sucks!
Yeah... I apologize. Please grab the current contents of the https://github.com/MarlinFirmware/Marlin/tree/RCBugFix branch. For all practical purposes that IS RC-8. We just haven't checked with each other and all shaken our heads up and down. Trust me... That is so close to what RC-8 will be, it won't matter.
Hi
Well ok :) That's what I decided to do. I'm in the middle of rewiring the 12" with a BL Touch and upgrading the firmware.
(I *do* seem to have a pices of code at work that either is or isn't the release version depending on which heads shake which ways ...).
The real question is weather I'm going to do enough with any of this to go through the whole "set up Eclipse" nonsense. I find the Arduino tools great for simple stuff. They are a bit of a pain once you get past simple stuff. Eclipse is a bit of a pain on any first setup ... who knows what I'll do.
Bob
Hi
I am setting it up right now on a 12" Pegasus. Once I get that running I will transfer it to the 10" Pegasus. The BL Touch is mounted to a custom extruder I came up with. It also holds mount and duct for the cooling fan. Once they both are done, they also will have dedicated +5 supplies and Rpi's to run them remotely.
I'm not real sure how much use my firmware will be to others. I'm running a SSR on the heated bed and the filament cooling fan. I have the bed offset a bit to the right of the location MakerFarm recommends. My bed limits don't work very well on a "standard" printer. They are great for doing stuff that is a bit larger than the bed. Another part of that trick is the thinner power supply that I use.
Once I get it all done I'll toss it up somewhere.
Lots of fun...
Bob
Hi
You may not want to compare to what I have right now. For some strange reason it does not realize the heated bed exists.....I have not quite yet made it to the BL Touch. I suspect I have something *REALY* obvious goofed up. I dropped back to the original firmware and the bed still is ok there. It's not a broken wire somewhere.
Lots of fun !!
====edit
Well, so far: Both RC-Bugfix and RC-7 have the "missing bed heater" issue. Roxy's branch for bed leveling (and the low memory warning :) ) seems to give me a valid bed temperature. Back to figuring out what's missing where.
Bob
Hi
Ok, here's what I did:
Grab RCBugFix, define the display (RepRap Discount Smart Graphical), upload the code to the Ramps.
Reset Ramps
There is no graphic on the display for the heated bed
Try to set bed heat ... no go (from serial, or from sub menus).
Try to run PID tune on the bed ... thermal runaway, glass bed shatters into many pieces. Why did I try that? Terminal stupidity is my only excuse :)
===
Repeat with everything in sight. I only get a bed heat graphic when the screen reads " UBL6 FT-2020 ready." All the other new branches have the issue with the bed heat on my machine.
I'll pound on it some more this evening. Firmware usually works best after you have a few beers and yell at it a lot ....
Bob
devel_ubl was forked from RCBugFix on July 4th. The BLTouch stuff is not in devel_ubl yet. (And that message "UBL6 FT-2020 ready." is from the UBL code base)
Hi
After a bit of "quality time" digging through the code ... errrr .... during the day ... I have a few ideas what may be going wrong. The working code is indeed the UBL code branch. I was looking at it mainly to see what did and did not work in terms of the bed heat issue.
=====
So, yup, pilot error. The bed thermistor defaults to zero in the "normal" branches. It defaults to something rational in the UBL branch. I was compiling it and looking at it *before* digging into the thermistor tables. Obviously not the best approach !!!
Bob
And just so you know... The Configuration.h file from the UBL branch is different than the default Configuration.h file from the RC-7 and RCBugFix branch.
The reason they are different is UBL is considered a 'Development' branch. The Configuration.h file in the UBL code is very literally my Configuration.h file. I know it works (at least on my hardware). And the theory is, if a user starts with that file and makes the minimum number of changes to accommodate their machine, they have the best chance of the code base still working for them.
It may be time to go back to the standard and default Configuration.h file in the UBL code. Things pretty much work. (at least on Cartesian Printers. Delta's still have not been brought up on that code base yet.)
Hi
I would certainly encourage folding UBL back into the main branch. I have a cartesian printer so that may bias my recommendation.
Bob
Roxy, I have a silly question for you. My Printers are now running the RC8 Marlin 1.1 RCbugfix, from Dec 2016.
I see the #define bltouch, do I just have to uncomment that and update the spacing from the nozzle? Or do I also have to play with my starting G-Code as long as uncomment the BL Touch?
This hasn't been discussed lately (in the last few months). But the plan was to get RC8 promoted to 'Stable' and then no more features would be added to it. Only bug fixes. And at that time, the plan was to then make the UBL branch the main Development Branch and use that to absorb any new features to the code base. Eventually... (maybe in 6 months???) it would start a Release Candidate phase and that would ultimately result in a new 'Stable' branch.
The big problem with 'folding' UBL back into the main branch is the other bed leveling schemes would disappear. UBL pretty replaces the existing code and feature set. So it is probably less controversial to just let it evolve on its own until there is no argument that it handles every case.
I just ordered a BLTouch, for use with my 10" Pegasus. I've got a Titan extruder and a blower fan, and I cannot find one BLTouch mount that will work with my setup. It seems like every Titan mount has the BLTouch right where my fan would be. (Fan mount: http://www.thingiverse.com/thing:2074787) Anyone have recommendations for a mount that I can add, or another compatible blower mount that includes BLTouch?
Hi
I simply redesigned the whole extruder and X carriage from scratch ....
Bob
I've been struggling to get my BLtouch to work on my 10" Pegasus. Anyone have a config I can use?