Auto-Calibration - $100 Bounty and Bragging Rights

[St Monkfish]

I have a fast solution to bed leveling, non software, specific to delta models, Cartesian systems are a different problem, i would need the full geometry and calculations for positioning, I’m not sure I have all the variables accounted for, but I’m doing the mod to an Orion delta today, if it works you owe me cash! love to talk about it, contact me

[Jimustanguitar]

I have a fast solution to bed leveling, non software, specific to delta models, Cartesian systems are a different problem, i would need the full geometry and calculations for positioning, I’m not sure I have all the variables accounted for, but I’m doing the mod to an Orion delta today, if it works you owe me cash! love to talk about it, contact me

I hope it works! If you need a Beta tester, hit me up!

Make sure to post it to the forum when you're finished!

[St Monkfish]

Wow that was fast, thanks
I Need some help with the connections,

I want to add an led to the end stop switches but I'm not shure where to tap in
At the Rambo pin connectors the + pole is open ( what is the voltage?) how is it triggered by the switch?)
Or
At the switch there is an open leg ( I think "NO")

If I tap the +(5v?) pin run wire to led - 480 ohm resistor - center leg (NO) of switch , then when the switch is triggered the light comes on

Does this work ?

[Polygonhell]

There is a Marlin branch https://github.com/RichCattell/Marlin that has a real attempt at auto calibration in it, it uses some sort of iterative algorithm to converge on the appropriate geometry, including individual tower position and orientation offsets.

[lordbinky]

Wow that was fast, thanks
I Need some help with the connections,

I want to add an led to the end stop switches but I'm not shure where to tap in
At the Rambo pin connectors the + pole is open ( what is the voltage?) how is it triggered by the switch?)
Or
At the switch there is an open leg ( I think "NO")

If I tap the +(5v?) pin run wire to led - 480 ohm resistor - center leg (NO) of switch , then when the switch is triggered the light comes on

Does this work ?

Off the top of my head, I think the switches are normally open, otherwise the induced voltage spikes that were triggering the endstops wouldn't have occured so easily. If so you can put the LED in-line on one of the end-stop wires and it should work fine without messing with the TTL logic on the board. Throw your multimeter in-line and work the endstop to check the current through the switch to size your resistor properly.

[St Monkfish]

Actually the Orion is using the switch in the normally closed position with just the signal and neutral, on the Rambo board the pin connector has a + pin I assume is 5v, I don think its triggered by the switch, if the switch has both NO/NC without mixing it should work, at this point I'm gambling because I don't have the data.....

Anyone got stats on this ?

[St Monkfish]

We'll.... That didn't work.........

Mod version 2.6, keep it simple stupid,.....taping the + pin on the Rambo board limit switches was a 20/20 dumb idea because the + pin is for high side logic whereas the - pin is for low side logic.......meaning No voltage, it's some high/low digital signal.
The fix is to just tap a 12v line off the power core, reset the plasma grid and tune the flux capaseters.....making sure the polarity of the switches is such that when the switch is triggered the LED sees ground (-) NOT THE SIGNAL PIN...no i didn't fry a board, I wasn't getting a closed test circuit .... Switch polarity....

Annyhoo the LED circuit works great, tomorrow I test the homing and leveling solution.......

Since this Isent a total fix, even though it will make the rest of the solution less complex, I will accept partial braggers acclaim and a spool of glow in the dark filament from Matterhackers, not lol...... Fer raal dude...

[JohnStack]

Videos, full docs, and if it meets a couple of folks standards (including mine), that $100 might be yours.

At a minimum, you're going to get electronic applause. I have the for you now...

[Eaglezsoar]

A spell checker is also recommended.

[notarat]

Let's say we set a standard that's fit for aeronautics:

More points on the plane = better geometry or map of the bed surface. Having the ability to set the number of points during what I would call Bed Surface Geometry or a "Bed Surface Map" would be great! (This is based on the assumption that there would be a number of high and low spots.)

Different temps may affect a bed's surface. How would that be accommodated?

Other points:

The firmware would have to accommodate a "Bed Surface Map". This would likely have to be configurable in the firmware.

And I want to take this time to say thanks. I didn't throw it out there just to see if someone could do it. I put out the reward because I think someone might take on the challenge and get it done. In no way does the bounty match up to the amount of effort this would take. /jws

I'm a total n00b, especially with Deltas (My Rostock Max is still inside the cardboard box it was shipped in), but I'm wondering if someone couldn't use a MS Kinect to, in essense, radar map the surface and feed those values to firmware/repetier?

I've seen the MS video on 3D Printing and how they are integrating Kinect to do things like scan users to 3D print avatars of the users so, if it's able to do that maybe it can be used to supply the mapping information...

dumb idea, I'm sure...I've not had my coffee yet.

[St Monkfish]

Who said it is misspelled, and spellcheckers can cause more damage then fat thumbs by selecting the wrong word and this IPad loves to finish my sentences and interject strange words....infernal devices anyway!

[geneb]

It's not really dumb. It's more along the lines of, "Let's contract NASA to build us some 12" tall water-powered rockets."

From what I've read on the deltabots list, the auto-calibration is nearly a done deal with the Marlin firmware, but a) I'm not using Marlin on a bet and b) No one that I know of has yet tried it with a Rostock MAX.

g.

[lordbinky]

The degree to which the print surface is warped itself being warped is typically the lesser of other build deviations. Ie, differences in the lengths of arms, angle and distance from a the center point for the towers, angle of the towers vertical alignment, etc. At the expense of computation power, if you account for things had variables for each of the arm lengths, you would get better compensation on the movements and positioning than just having an external map of the bed.

If you map the bed with the effector itself, you kind of wrap all those variables together and just compensate the Z height based on the X/Y position. Still, at the cost of computation and memory.

[dtgriscom]

If you map the bed with the effector itself, you kind of wrap all those variables together and just compensate the Z height based on the X/Y position. Still, at the cost of computation and memory.

Wrapping all these issues together and dealing with them at the same time would be a BIG bonus. Be aware, though, that this would only be accurate at the bed surface; the compensation would get worse as you increased the distance above the bed.

The computational downside should be pretty small, and the memory impact (unless we're collecting hundreds of calibration points) should be minimal.

Dan

P.S. Congratulations on your 666th post! Any demonic plans?

[Captain Starfish]

Waitasekint.

I get trying to alter the print alignment to cope with a planar, non-horizontal bed - picture a perfectly flat glass sheet that's propped up with a matchstick on one side. You skew "vertical" for the print and all the right angles stay right and the part is fine. MatterControl doesn't do it automatically but it does make the process quite simple.

If your bed looks like a golf course instead of a pool table, though? What exactly are you compensating for? Trying to warp a flat model fit the bed is going to leave you with a warped part. You end up having to go back to a planar base which is established by 3 points in space. Providing that the nozzle doesn't embed itself in a high spot of the bed protruding through that plane, of course, what's the point in going for eleventy billion points that you're going to end up doing some arbitrary mathematical reduction on anyway to determine what three points will - ie the orientation and offset of the base plane.

However you do it, a non planar bed needs to be dealt with using extra/less fill - exactly what the extruder will achieve anyway on layer one by simple virtue of the fact that the top of layer one (ie the path of the nozzle) IS moving around in a plane.

[dtgriscom]

If your bed looks like a golf course instead of a pool table, though? What exactly are you compensating for? Trying to warp a flat model fit the bed is going to leave you with a warped part. You end up having to go back to a planar base which is established by 3 points in space. Providing that the nozzle doesn't embed itself in a high spot of the bed protruding through that plane, of course, what's the point in going for eleventy billion points that you're going to end up doing some arbitrary mathematical reduction on anyway to determine what three points will - ie the orientation and offset of the base plane.

In my (very limited) experience, the primary problem with bad calibration is printing the first layer. Unless "Z==0" is within a fraction of a millimeter of the right height, either the nozzle is too high and I get poor adhesion, or the nozzle scrapes the bed and the extruder feed jams.

Yes, after that it would be nice if each succeeding layer were exactly planar, but if it's consistent enough to adhere well to the previous layer then I don't think I'd care if things were very slightly warped. Yes, if we're really dealing with serious warping then I'd call it a problem, but a problem with a clear solution: buy a flatter bed.

BTW, an alternative to adjusting the Z height would be explicitly increasing/decreasing the feed rate to properly fill that first layer. This wouldn't be automatic, as there is (or should be) very little elasticity between the nozzle and the extruder feed: you'd have to have the software do the adjustment, and it couldn't adjust for a whole lot. But it's something to try.

[Captain Starfish]

BTW, an alternative to adjusting the Z height would be explicitly increasing/decreasing the feed rate to properly fill that first layer.

BAM

Now that is a great reason to have an accurate topographical map of a bed.

I take it all back. Just got to get the MatterControl, Repetier, etc, etc, etc guys to write us a new bed filler plugin to use the map!

[JohnStack]

I suggested this to Simplify3D as well as compensation stuff for cartesians they say they support but don't really - and the response was like I came from another planet. Even though I do, it's impolitic to remind me.

[lordbinky]

I don't think the issue most people have is with beds that look like a golf course, it's more along the line of the mysterious Z0 perimeter problem most are fighting with and wanting resolved with autocalibration.

So when the effector thinks it's actually at Z0 at some 250mm from the center, it's actually .2mm higher, you just compensate for whatever geometry issue there is by correcting the software that Z0 is actually at Z-.2. It's correcting what the software thinks it's location to compensate for it's real location. Applying this to movement is the terrifying part. You can either map the bed at a sufficently high resolution mapping that you minimize computation differences (tradeoff : high memory, low computation) or you can have a mapping with few points and interpolate the Z offset ( low memory, high computation option). This gets even worse when you factor in the hardware in which the computation is being done on. So the interpolation algorithm is going to be modified to account for the rounding errors and differences in the direction the head is coming unless you want inconsistant offsets that will result in odd patterns (at best) in prints.

It's better to perform a series of test points from the bed and interatively get the correct values for the settings of the printer, as from what I understand the marlin firmware is doing. I'm curious if they used a method similar to what I outlined or if it's completely different, I'll have to look into it sometime.

[dtgriscom]

I suggested this to Simplify3D as well as compensation stuff for cartesians they say they support but don't really - and the response was like I came from another planet. Even though I do, it's impolitic to remind me.

Perhaps their dismissal was related to the division of labor. For this to work, the slicer would need to know about the printer's geometric problems/parameters, but that's probably best handled in the printer firmware itself. Maybe there should be a new G-code specifying that we're printing the first layer, and the extrusion should be scaled by the ratio of the actual Z altitude and the specified Z altitude. (Standards are for extending and customizing, right? Just ask Microsoft!)

P.S. My goal for the day: use impolitic in casual conversation.

[mhackney]

Captain Starfish, it's not that your bed looks like a golf course. It's that the path of the nozzle above a perfectly flat and level bed can travel in a non-planar surface if your machine is not mechanically perfect - tower position, tower leans, arm length, delta radius, etc. You can either fix all the mechanical problems (preferred but sometimes difficult) or compensate for them in software. For instance, I had an issue with my Z tower when I first assembled my Max last year. The symptom was that the nozzle would raise above the bed when it was between the Z and X towers but traveled in a pretty good plane everywhere else. This created problems putting the first layer down since the nozzle rose above the bed at that position and was too far to get adhesion of the plastic bead. (the problem I described is greatly simplified, the movement was "odder" than that but you get the point). In my case, my Z tower was leaning ever so slightly toward the X. Turns out that the opposite effect was happening on the other side, the nozzle would get too close to the bed. I spent a couple of days tweaking until I got that out. I know that my effector movement is not *exactly* planar now and it actually changes day to day. Maybe heat and humidity affect the tower positions/lean or something. But it is usually good enough. The sort of calibration we are talking about has 2 components:

1) actual calibration - by taking a series of measurements at strategic spots over the bed, the delta arm length and delta radius can be determined. Test, compare, repeat until you are within some small margin. The strategic spots are at the center and at the base of the towers.

2) movement compensation - once the delta parameters are calibrated, if there is any non-planarity in the nozzle movement other mechanical issues are at fault (tower lean, non parallel delta arms, etc). Most of the time these are within reasonable limits but even a slight variation, along with daily changes described above, can cause enough variation to cause problems getting the first layer to stick well - especially over large areas (i.e. large parts). So, the compensation takes many measurements on a grid (there are other ways to do this) say at 1cm increments and creates a topographic map in Z. This is NOT a map of the bed. It IS a map of the nozzle/effector movement. With this map, you can add or subtract the ∆Z from the known actual Z (which by default could be at X=Y=0). Now, when printing, when the nozzle is at a give X, Y position, the firmware can check to see what the ∆Z is and compensate for it. The simple approach is to just use the same ∆Z for all points within a square cm, a better approach is to extrapolate between neighboring grid points. The simple case works surprisingly well since we are just trying to tweak out very small (<1mm for sure, .5mm if possible) non-planarities.

Interestingly, you could actually do 2 above in gcode. If you had the topo map of ∆Zs you could post process the gcode and insert explicit Z movements for every X, Y movement. I've done simple tests doing this and it works pretty well.

[lordbinky]

Good point, my thinking was stuck in firmware but you can totally do the compensation as a pre-process and avoid real-time firmware calcs and the software/processor limitations.

[mhackney]

Consider this, you could use this same technique to generate sinusoidal layers (you'd want to start flat and gradually develop the sinusoidal deposit) or many other interesting variations...

[Eaglezsoar]

Why don't we go to the set of Startrek, get some of the self replicating nano bots, train them to walk on the bed and transmit the coordinates, after
mapping the whole bed they can line up on the glass binders until needed again. We may have to buy them some heatproof booties, I hear that
their little feet have trouble with heat.
If we get this to work, Mhackney can go back to writing in English.

[mhackney]

If we get this to work, Mhackney can go back to writing in English.