3.5 WIP: Hangprinter: Optionally allow up to 5 anchors/motors to be configured #601

jtimon · 2023-02-25T22:33:53Z

This is a rebase of a subset of #585 which can be seen in https://github.com/GloomyAndy/RepRapFirmware/compare/v3.4-dev...jtimon:v3.4-hangprinter-5anchors?expand=1

This is not complete yet though, because I just passed by the flex stuff. I hope it doesn't take me much to "complete" this PR in that sense.

The 5 anchor version must be the simplest to optionally support. 6 should be easy too, I think.
In 3.4, the first problems appeared with 7 anchors or more, because the config messages didn't fit. which I got some guidance for here: #585 (comment)
But I didn't do it because 3.5 was out and I needed to rebase.

Currently this shouldn't work for 5 anchors yet but it also shouldn't break 4 anchor hangprinters.
I didn't test anything yet for 3.5.

But it's ready for suggestion for changes from review or testing.

This PR contains the following PRs:

3.5 NOCHANGE: hangprinter: simplify things by turning unrolled stuff into loops 3.5 NOCHANGE: Hangprinter: simplify things by turning unrolled stuff into loops #597
3.5 Hangprinter: Generalize IsReachable() to any pyramid 3.5 Hangprinter: Generalize IsReachable() to any pseudo-pyramid #598

TODO:

Fully adapt to the changes from 3.4 to 3.5
(kind of optional) Adapt HangprinterKinematics::MotorStepsToCartesian to N anchors for better accuracy.

tobbelobb · 2023-02-26T11:15:59Z

Nice!

// Precondition: all the base anchors belong to the same plane,

I don't think we should add this precondition. Better to have an imperfect IsReachable() implementation than to place restrictions on users.

jtimon · 2023-03-01T18:59:46Z

I don't think it's imposing restrictions on the users. on the contrary, it's allowing more configurations. Not all possible configurations yet, but more than before. Nothing that was previously allowed will be restricted, only more things will be allowed. In future PRs we can allow more configurations, but I'm afraid trying to get it right for all possible configurations in one go is going to be very hard, specially since I want to scale this to up to 8/9 anchors.

Personally I'm most interested in the configurations with all anchors on top at this moment. So my plan was to focus on that and on adding more anchors. I'm not really sure there will be much interest in supporting pseudo-pyramids with not all the anchors in the same plane, to be honest.
Anyway, if you have an alternative to #598 that is more generic than this, all welcomed.
But at this point, personally I would prefer to focus on lower hanging fruits.

tobbelobb · 2023-03-01T19:41:53Z

Moved this comment to #598, here: #598 (comment)

tobbelobb · 2023-03-01T19:54:23Z

Anyway, if you have an alternative to #598 that is more generic than this, all welcomed.

Oh, I have been commenting on the wrong PR. This one is a ok 👌
It's the other one that changes IsReachable(). Sry.

jtimon · 2023-03-03T16:46:41Z

Well, it applies to this one too since this one contains the other one, but, yeah, I guess better discuss it there if it's specific to that.

tobbelobb · 2023-03-14T16:42:33Z

src/Movement/Kinematics/HangprinterKinematics.cpp

@@ -342,32 +402,33 @@ inline float HangprinterKinematics::MotorPosToLinePos(const int32_t motorPos, si

 void HangprinterKinematics::flexDistances(float const machinePos[3], float const distanceA,


This function must take an array of numAnchors distances that must be looped over

Yes, adapt everything in #569 to not using A_AXIS constants inside.
But first deal with HangprinterKinematics::StaticForces and the other changes in #562

tobbelobb · 2023-03-14T16:46:51Z

Only had my phone available so only a brief review. Noticed that flex compensation isn't yet rolled into loops over numAnchors. Need to finish that before we can test on hw

src/Movement/Kinematics/HangprinterKinematics.cpp

src/Movement/Kinematics/HangprinterKinematics.h

Otherwise it gets too deep. I think there's an option in eclipse to forbid this by project in format, one can config depness iirc.

@tobbelobb

This is a clean generalization from 4 to N. Pseudo-pyramid because the vertices in the base of the pseudo-pyramid don't need to be on the same plane. There's no base of the pyramid as such. Better documented and cleaner thanks to @tobbelobb on github.

…etups - Change from bool to class and introduce None mode for custom (experimental) setups (all allowed) - refactor out IsInsidePyramidSides - refactor out IsInsidePrismSides - introduce HangprinterAnchorMode::LastOnTop for prism-like setups with all anchors on top - introduce a T parameter in M666 to configure the anchor mode

This allows to print square pyramids instead of tetrahedrons, at the cost of an extra motor. WIP: not everything is adapted after recent changes. See the newest TODOs

jtimon · 2023-03-21T04:34:24Z

Only had my phone available so only a brief review. Noticed that flex compensation isn't yet rolled into loops over numAnchors. Need to finish that before we can test on hw

Yes, that's the kind of thing I mean by "Work In Progress" (WIP).
It means there are unsolved things To Do in the new comments of the PR.

But it's normal to have to rebase. When you were coding the flex stuff you had to focus on the flex stuff working and merging fine with the current branch at the time, which it did.

But if you do a PR on top of #597 but completing it in the parts related to flex which this doesn't account for, that would be awesome. But if not I will have to understand these new flex parameter at some point anyway.
With M669 and M66 there's a lot of parameters I still don't understand, which is alright as long as I don't break them.
But can I test myself if I broke them without understanding it myself?

In the case of the latest commit, I think yes, I think it's broken at the start part reserving memory dynamically.
The last commit breaks my setup, don't try in yours, try the previous commits. and on this last commit thank you for the review, keep it coming.

But when I break the wifi of my machine with the latest commit because I don't know what I'm doing with the memory...it gets me back to the times of opening the box for changing binaries in the SD or, worse, talkingto the printer by pronterface and an usb cable. Both mean opening the machine, sigh.

Just please nobody run this code on their machine unless you want to see an error and want to take out the sd card to reinstall the binaries, which is a pain in the ass.

Any tester interested in this PR should test the "previous" PR #603 which won't break your printer hopefully like this PR at this point.

tobbelobb · 2023-03-25T13:56:18Z

if you do a PR on top of #597 but completing it in the parts related to flex which this doesn't account for, that would be awesome.

👍 Will look into it. Since it's not merged yet maybe I can push directly into your PR instead of opening a new one? #597

tobbelobb mentioned this pull request Mar 1, 2023

3.5 Hangprinter: Generalize IsReachable() to any pseudo-pyramid #598

Open

jtimon force-pushed the 3.5-hangprinter-5anchors branch 3 times, most recently from cc7c485 to a2b32db Compare March 5, 2023 20:56

tobbelobb reviewed Mar 14, 2023

View reviewed changes

src/Movement/Kinematics/HangprinterKinematics.cpp Show resolved Hide resolved

tobbelobb reviewed Mar 14, 2023

View reviewed changes

src/Movement/Kinematics/HangprinterKinematics.h Outdated Show resolved Hide resolved

tobbelobb reviewed Mar 14, 2023

View reviewed changes

src/Movement/Kinematics/HangprinterKinematics.h Show resolved Hide resolved

jtimon added 2 commits March 20, 2023 19:05

HangprinterKinematics::WriteCalibrationParameters: flat return style

d227a9c

Otherwise it gets too deep. I think there's an option in eclipse to forbid this by project in format, one can config depness iirc.

Hangprinter: Loop around the anchors instead of naming them directly

e19697e

jtimon force-pushed the 3.5-hangprinter-5anchors branch 3 times, most recently from c626607 to d0d4bd3 Compare March 21, 2023 01:24

jtimon added 3 commits March 21, 2023 01:47

WIP: Hangprinter: Allow Hangprinter to be configured with 5 anchors

b6dacd1

This allows to print square pyramids instead of tetrahedrons, at the cost of an extra motor. WIP: not everything is adapted after recent changes. See the newest TODOs

jtimon force-pushed the 3.5-hangprinter-5anchors branch from d0d4bd3 to b6dacd1 Compare March 21, 2023 01:51

tobbelobb mentioned this pull request Mar 25, 2023

3.5 hangprinter: Allow custom and prism-like setups #603

Open

2 tasks

jtimon mentioned this pull request Apr 9, 2023

3.5 NOCHANGE: Hangprinter: simplify things by turning unrolled stuff into loops #597

Open

tobbelobb mentioned this pull request Apr 16, 2023

Generalize Hangprinter Anchors #606

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

3.5 WIP: Hangprinter: Optionally allow up to 5 anchors/motors to be configured #601

3.5 WIP: Hangprinter: Optionally allow up to 5 anchors/motors to be configured #601

jtimon commented Feb 25, 2023

tobbelobb commented Feb 26, 2023

jtimon commented Mar 1, 2023

tobbelobb commented Mar 1, 2023 •

edited

Loading

tobbelobb commented Mar 1, 2023

jtimon commented Mar 3, 2023

tobbelobb Mar 14, 2023

jtimon Mar 21, 2023

tobbelobb commented Mar 14, 2023

jtimon commented Mar 21, 2023

tobbelobb commented Mar 25, 2023

		@@ -342,32 +402,33 @@ inline float HangprinterKinematics::MotorPosToLinePos(const int32_t motorPos, si

		void HangprinterKinematics::flexDistances(float const machinePos[3], float const distanceA,

3.5 WIP: Hangprinter: Optionally allow up to 5 anchors/motors to be configured #601

Are you sure you want to change the base?

3.5 WIP: Hangprinter: Optionally allow up to 5 anchors/motors to be configured #601

Conversation

jtimon commented Feb 25, 2023

tobbelobb commented Feb 26, 2023

jtimon commented Mar 1, 2023

tobbelobb commented Mar 1, 2023 • edited Loading

tobbelobb commented Mar 1, 2023

jtimon commented Mar 3, 2023

tobbelobb Mar 14, 2023

Choose a reason for hiding this comment

jtimon Mar 21, 2023

Choose a reason for hiding this comment

tobbelobb commented Mar 14, 2023

jtimon commented Mar 21, 2023

tobbelobb commented Mar 25, 2023

tobbelobb commented Mar 1, 2023 •

edited

Loading