Philo

Well... I just edited the file with a text editor and modified the transformation matrix. <Brick refID="0" designID="6260" itemNos="626001"> <Part refID="0" designID="6260" materials="1"> <Bone refID="0" transformation="1,0,0,0,1,0,0,0,1,-2.7999999523162842,0.24989999830722809,-1.2000000476837158"> </Bone> </Part> </Brick> is changed to <Brick refID="0" designID="6260" itemNos="626001"> <Part refID="0" designID="6260" materials="1"> <Bone refID="0" transformation="1,0,0,.5,1,0,0,0,1,-2.7999999523162842,0.24989999830722809,-1.2000000476837158"> </Bone> </Part> </Brick> There is a tutorial about this kind of transformation here. Sorry, it's in French...

Here is an example with a slanted torso. Rename the attached files as .lfxml to open them in LDD. skel2 is the slanted version, skel1 the straight one for comparison. But it's very difficult to do this kind of things with LDD (collision detection comes in the way), it would be much easier to do with LDraw system! skel1.txt skel2.txt

With matrix modification in LXFML you can do linear transformations of parts, eg. scale them (with different scaling in each direction), rotate, slant or mirror them. But you can't distort the overall shape (eg it's not possible to introduce a joint in the middle of a leg part).

I thought it was deleted groups, but maybe that's something else - anyway I had 2 files showing this issue, and Mike was kind enough to update his tool ;)

A new version is available, correcting a crash occuring during L3B importation when referenced groups were missing in file (this occurs after a group delete in SR3Dbuilder).

MLCad is not available on Mac (though it's probably possible to use it on a virtual machine), but Bricksmith is appreciated by Mac users...

Mike now updated the link!

Michael Heidemann has updated his MPDcenter tool. New features were added to previous mpd management capabilities: - Conversion of SR3D files with groups into multi-part Ldraw files. Each group is converted to a submodel, and MPDcenter tries to orient part in each submodel to ease further editing in MLCad. The first part encountered in each submodel is placed at origin of submodel with simple orientation (identity transformation matrix). If needed, using SR3D builder instructions miner, you can reorder parts in the group to choose this first part. (usage: start MPDcenter, then import -> import L3B as MPD). - Creation of part inventory of a model. This is very similar to LDView inventory function (and actually it uses LDView engine), but instead of relying on external images coming from Peeron, it locally generates thumbnails for all parts, including LDraw unofficial parts. (usage: start MPDcenter, open model, then Informations -> Inventory). ...and of course these new functions were added to an already capable tool that allows to merge MPDs, include unofficial parts to MPD, validate MPD files for inclusion in LDraw Official Model Repositary, etc... A complete documentation should be available soon.

Yes it does! A Li-ion charging circuit, TI BQ24123 or similar.

9V is clearly not enough, these batteries start charging at 9.5V. It is probably safe indeed to charge from 12V, but you can't be completely sure, some components in battery charger integrated inside the battery pack could be tightly specified... One problem is that these PSU are generally not well regulated when drawing too little current on all outputs (here your 5V output would not be loaded at all). Check this carefully... Yes. No problem about this.See this page, which includes information on NXT battery (EV3 one is extremely similar except for mechanical considerations).

Excellent! I guess you found my LDraw model of Firgelli's LA? ;) I'd be happy to know the difference between NXT and EV3 version!?Edit: maybe the auto-id resistor allowing EV3 software to identify properly the device used. I have to test this...

I tried that route but it doesn't work with reasonably sized resistors. V2 output current can reach 2A and over current protection (that causes problem here) is around 3A. We can say that the problem is that the V2 motor driver is at the same time very good... and a bit picky! Capacitor in M-motor has a value of the order of 1µF.

Nazgarot perfectly summarized the problem. Note that ONE M-motor on a V2 output is generally safe. Problem begins when you attach 2 M-motors on a single output.

Ah - yes... LDraw official parts now use utf8 so this should never occur. But Lsynth part have not been put on Parts Tracker...

If the lsynth parts (LS22.dat and LS23.dat are used here) are in your LDraw/parts folder they should be found without problem. They are regular LDraw parts, the thing that makes them different is their usage: while regular parts are used at 1:1 scale (possibly mirrored), LSynth parts are scaled anisotropically. Maybe that's what causes problem with Blender importer?

Isn't it an official/unofficial parts issue? Built in library of LeoCAD doesn't include unofficial parts, and even if you instruct it to use LDraw library, it doesn't manage the LDraw\unofficial\parts folder: you have to dump all unofficial parts in main parts folder (and LDraw\unofficial\p in LDraw\p)

Might be the rapidly changing highlights on baseplate studs that fools the encoder. Dark images are also more prone to compression artifacts.

(strong noise of jaw dropping on floor...) I have a hard time convincing myself that these are made from LDraw parts, especially those that I modeled (eg. the panels). They look so smooth!

That's also why I don't use alkaline either... A NiMH costs about the price of your lithium cell, and can be recharged many, many times ;)

Your are probably right, these batteries have a huge capacity - problem is that they are also VERY expensive ;)

Even the advantage of initial higher voltage doesn't last long - NiMH discharge curve is flatter. Have a look at this 250mA discharge curve: At higher current, another factor is internal resistance of the battery, the voltage of alkaline battery is below the one of NiMH from start of discharge... Mode details and comments here.

Nice find... many of the ideas have already been used in GBC modules, but the wood and metal implementation is great!. I think that the use of heavy, perfectly round steel balls does help, LEGO balls too often get stuck here and there... I like the Theo Jansen leg ball pump ;) Edit - The Cardan lifting mechanism is awesome too!

Ah, the old problem of unofficial parts that have been renamed... This occurs when the official LEGO part number is discovered after the part has been put on parts tracker, but before it became released as an official part. LDraw maintains compatibility (using "moved to" parts) when number is discovered after part is official, but no while it's still on parts tracker. There is nonetheless a way to check the history of such parts, using this kind of query (change the part number in url according to your need): http://www.ldraw.org...parts/u9113.dat x882 -> 70038 (shock absorber spring) u9112 -> 59143 (PF medium front) u9113 -> 59142 (PF medium back) 58124s03 does exist in official library, but it's a subpart in the part\s folder so it should be referenced as s\58124s03 (that's a bug in SR3Dbuilder, dunno if it has been corrected). About mirrored studs: this occurs when the "lazy" (*) model builder create one half of a symetric model and uses mirroring to create the other half. LDView has built in code to restore studs in normal orientation when this occurs (*) no offence intended. I know very well the length of time taken to create a big model!

Well, only the low resolution of the regular angular stud exist too, so it must have been substituted with a true cylinder at some time in the LDraw -> MODO conversion. Since the beveled stud can be easily defined mathematically (cyinder+torus+disk) it should be possible to get a non-facetized render. This is a more tricky issue, but I think it should vanish if you use a higher resolution torus. Not a big deal to change ;) How much thinning percentage is needed, according to you?

Pretty nice! What was the problem?