OctoPrint plugin...

[geneb]

FYI, the least-squares delta calibration from David Crocker (dc42) is now a plugin for OctoPrint - I'm still testing it and it requires unreleased firmware, BUT it appears to work based on my one test.

I'll let you know more when I've got more testing and it's ready to be hammered upon.

g.

[DerStig]

Wait what? and it will work with the HE280 accelerometer? Oh my god this could be the solution to my problems.

WHERE?!?!?!?!?!? Where can I get this and how do i run it?????

[enggmaug]

My googleing did not return anything interresting.

The Octopriny pluggins seach neither.

Can you please share a link ? I'm also eager to give this a try with my FSR setup.

[Noircogi]

I don't think it's generally available yet. I don't see it as an available plugin in octoprint or on the web page.

[geneb]

It's not released yet and when it is, it'll only work with the HE280 and Repetier Firmware.

There's a few outstanding issues that need to be resolved before it's generally available - and it'll require a specific version of the firmware in order to work.

g.

[DerStig]

It's not released yet and when it is, it'll only work with the HE280 and Repetier Firmware.

There's a few outstanding issues that need to be resolved before it's generally available - and it'll require a specific version of the firmware in order to work.

Any idea of when it might be available? No matter what I do I'm still low between the X and Z tower and I know that the build is perfectly square

[geneb]

No idea, sorry. The primary issue right now is that you can only use it one time - Octoprint stops communicating with the printer and we're not sure what end the issue is on.

g.

[enggmaug]

it'll only work with the HE280 and Repetier Firmware.

Both in short and long term ? How much of an effort would it be to port it for repetier with FSR setup ?

[crocky]

It's not released yet and when it is, it'll only work with the HE280 and Repetier Firmware.

There's a few outstanding issues that need to be resolved before it's generally available - and it'll require a specific version of the firmware in order to work.

g.

Watching.... Patiently.

[Jimustanguitar]

it'll only work with the HE280 and Repetier Firmware.

And other hotends retrofitted with the accelerometer board, right?

[geneb]

it'll only work with the HE280 and Repetier Firmware.

And other hotends retrofitted with the accelerometer board, right?

Yes, as long as it's the SeeMeCNC fork of Repetier that's running the printer.

Getting weird results on the end stops right now - I think it's due to the firmware going into thumb-twiddling mode occasionally.

g.

[626Pilot]

Subscribed.

[mhackney]

Gene, would the SeeMeCNC knife or spoon (or maybe their spatula) work as well as the fork or does it have to be the fork?

[lordbinky]

Gene, would the SeeMeCNC knife or spoon (or maybe their spatula) work as well as the fork or does it have to be the fork?

A knork, they would have a knork.

[geneb]

*sigh* You guys kill me.

The good news is that I've gotten it mostly working. The issue we're facing right now is that the more you run it, the lower the max Z gets and we're not quite sure why.

Having never touched JavaScript before last Friday, I now understand WHY I never touched JavaScript before last Friday. *facedesk*

g.

[wcogen]

DerStig wrote:
"Any idea of when it might be available? No matter what I do I'm still low between the X and Z tower and I know that the build is perfectly square"

This is exactly the problem I'm seeing. I'm really looking forward to a solution.

[lordbinky]

DerStig wrote:
"Any idea of when it might be available? No matter what I do I'm still low between the X and Z tower and I know that the build is perfectly square"

This is exactly the problem I'm seeing. I'm really looking forward to a solution.

Just in case, but double check the arm lengths and the widths on the carriages and platforms need to be the exact same. Small differences here cause the effector to tilt and wreak havok with software calibration, I learned this the hard way when I made my own mag arm parts for the umbee.1 mm difference between the carriage and platform width for the arms made the hotend very noticibly not perpendicular at the edge of the bed.

[lordbinky]

I'm looking forward to knorking smoothieware to work with thus. Squee!

[626Pilot]

The issue we're facing right now is that the more you run it, the lower the max Z gets and we're not quite sure why.

I might be sure!

That sounds similar to something I had to compensate for in my Smoothie fork. Changing endstop trim also changes Z height. Make sure you are adjusting for the changes you make to the endstop trim, either by adding or subtracting it. Otherwise, the trim settings will walk around, taking Z height with them.

Maybe this helps you:

Code: Select all

    // Normalize trim (this prevents downward creep)
    auto mm = std::minmax({ cur_set->trim[X], cur_set->trim[Y], cur_set->trim[Z] });
    cur_set->trim[X] -= mm.second;
    cur_set->trim[Y] -= mm.second;
    cur_set->trim[Z] -= mm.second;
    set_trim(cur_set->trim[X], cur_set->trim[Y], cur_set->trim[Z]);

BTW, my condolences re: having to touch JavaScript, the most annoying major programming language out there.

[dc42]

BTW, my condolences re: having to touch JavaScript, the most annoying major programming language out there.

The original implementation of this algorithm is the C++ version in RepRapFirmware, so no need to touch JavaScript if you don't want to.

[626Pilot]

The original implementation of this algorithm is the C++ version in RepRapFirmware, so no need to touch JavaScript if you don't want to.

I want this for my Eris, which has a RAMBo.

[geneb]

The JavaScript version was the shortest-path solution for me as my C++ Fu is pretty weak. (I saw the TEMPLATE keyword and ran screaming.

626Pilot: The code is originally from dc42's web-based calculator, which (AFAIK) is functionally identical to the code that results in near-miraculous levels of calibration on Duet-equipped printers. Ryan suspects the issue may be how I'm applying the results of the calibration. (This is confirmed, I just heard that he's got it working properly and once some cleanup on the firmware side is done, it'll get released.)

g.

[mhackney]

It's true, the calibration in dc42 firmware is the best there is available for delta printers in both speed and results. The new grid leveling takes the delta experience to the next level - it's in the latest dev branch and I've been using it for a few weeks. Any oddnesses, idiosyncrasies, etc remaining after calibration are eliminated with grid based compensation.

[stough]

Awaiting releases of plugin and firmware with baited breath...

[DerStig]

When will this magical firmware be available?