Philo

AFAIK no, it's not possible I confirm that PF handset is compatible with EV3, handy if you need more than one channel.

Sure

Just for the challenge, I made a version with connected arms. Get it here! The trick was to attach the two arms together first, with only one ball joint connected. Then adjust progressively the two arms angle and rotate the ball joint till it fits properly the other ball joint (using hinge mode). But definitely the ball joint is always a nightmare to adjust in SR3D...

OTOH even with retail version you can save files with sensor data. Program is a bit less obvious, but pretty feasible (and actually more flexible).

Ah - thanks! I corrected the wrongly formed links in my original post. Note also that I have prepared stickers for EV3 panels. These panels + stickersare also available in Complete unofficial parts archive, but they are done using texture map so you also need texture images, and they show properly only with a handful of LDraw applications. Get texture images and see more usage details here.

It's a single blink. More details here .

V1.5 has a firmware that corrects some bugs of v1. You can recognize it by LED blinking at startup.

Actually both, included in the motor driver chips. Indeed, all use auto-resettable thermal fuses.

NO!!! contact is not done with metal but with relatively high resistivity conductive rubber. Note that NXT touch sensor uses a true metal-to-metal switch, but there is a 2.2kOhm resistor in series...

No, that won't work! RCX touch sensors have an internal resiatance around 500 Ohm, much too high to light a lamp or allow a motor to run (and as pmichiel suggested, you might even fry your sensor, burning the conductive rubber!) I just tested 10 RCX sensors, all of them work fine.

Your disappearing parts problem is probably caused by "Draw to selection only" mode (it's toggled on/off with the icon looking like a pencil with line and spark, near fine grid icon). With this option on, MLCad draws only the parts placed above the selected one in text list. (edit) I had a closer look at your screenshot, I confirm that "Draw to selection only" icon is pressed... Otherwise, I completely agree with legolijntje, SR3D is the way to go, especially for Technic constructions. Yes, you may find it quircky sometimes, but once used to it, it's very efficient. Having to press enter for each parts may be a hassle, but it's nothing compared to placing and rotating parts at odd angles in MLCad!

Mmhhh... I am not well equipped for that kind of measurement... I noticed too the noise of EV3 motors, that's what lead me to thing that it could be a lubrification issue. But then I heavily tested only one motor, it could be a problem specific on this one... I have not tried yet to open the motors, but anyway it is not possible to open them fully without breaking the red hub (like the NXT motor). BTW, you can see a reverse-engineered schematics of EV3 motors internals here: http://bberry.perso.neuf.fr/swcaddb/Lego/CDEV3pack.htm

Measure of stalled torque is very imprecise, and a 3.5% difference in "normal" torque is nothing considering the overall precision of measures - especially since the high torque of EV3/NXT motors stresses a lot my test setup. More significant is the low energy efficiency of EV3 motor compared to NXT, but I can't really find a reason. Poor lubrification in EV3 motor?

This is indeed the coolest 3D printing for LEGO I've seen so far! The black one is nice, but I kind of liked the raspberry look of the first one ;) Your blog is extremely informative, thanks for putting this up!

On the other hand, what you could do with built-in software was extremely limited. Have you considered programming it using NQC through BricxCC?

As I said, 12V might be too much (but I don't think so). To be on the safe side, you might just add two silicon diodes (1N4001..7) in series, they will drop around 1.5V off the 12V suppply. There is no problem using a higher current rating, a supply only provides what it is asked for!

Any DC power supply with at least 700mA and in 9.5V to 11V range should be OK. Charging "intelligence" is in the battery itself. But there is no charging below 9.5V (voltage too low), and going too high might be unsafe for some components in the battery - though 12V is most probably OK. At least, charging circuit in the battery (a TI BQ24123) is 16V rated. 8 hours charging time is definitely too long, as this battery charges in about 2 hours: Note that this curve was done at nominal 10V, current is proportionaly lower at 12V - but the 500mA rating of your supply might nonetheless be the cause of your problem at the end of the charge. Some more details here.

Something like this?

Well... I tried it, but it is of very limited use. You can only measure the energy (and current, and voltage) going in and out the small NiMH (200mAH!!!) battery included, Nice to learn about energy harvesting and such, but you can't use that for general purpose measurement. I'd rather recommend Mindsensors' Current and Voltage meters for that.

Early Mindsensors devices simply used holes in PCB with tight fit around axles (4.7mm diameter?), and used axle pins to attach to LEGO parts. Remains to try to control it using NXT/EV3 BT...

This video is kind of a teaser... ...for the building instructions and programs that I plan to publish when the EV3 set is released... One thing for sure, EV3 color sensor and IR sensor are really great! Enjoy! Philo

What is perfectly legit is to download and install the blocks needed to use them! You can find the gyro block here. And since they made the blocks available on retail website, it clearly means that the edu sensors work with home edition... (but I have not explicitely tested)

SORT3R has now a home on the web! See here for building instructions, programs, etc...

On my website, but I'll announce it here...

I already switched firmware on my EV3 from edu to retail and back a few times. The main difference (don't know if there is any other) is that Edu firmware contains an on-brick datalog applet. If you have Edu software, it allows to reinstall Edu FW...