OzShan

I am getting close to the end of this long term build and it's time to share some pictures and a bit of the story. Even before I had finished putting together 10231, I decided I wanted a Crawler to go with it. Being a Technic fan it had to at least drive around and lift the launch platform and shuttle. Those two basic goals spawned a project that has lasted a little over 2 years so far. Some ideas have stuck around since their inception, others were a bit optimistic (like building a peristaltic pump and hoping I could find a way to control the pneumatics hydraulically). February this year marked the 50th anniversary of when the two crawlers went into service, so recently there has been extra motivation to finish. The base equipment; - 16x M-motors (drive) - 4x L-motors (pneumatic jacking and leveling) - 4x IR Receivers (V1 as the V2s do not like driving multiple m-motors on a single channel) - 4x NXT servo motors (steering) - 4x RCX rotation sensors (measuring jacking level between truck and chassis) - 2x NXT bricks (one master and one slave. The master communicates with the Android Tablet and coordinates itself with the slave. Programmed in LeJOS) - 1x PF IR-Link sensor (link between master NXT and all PF motors) - 2x PF Battery boxes (with thermal overload removed) - 1x Android Tablet Future add-ons - Accelerometer (automatically detect the crawler is on a gradient and adjust the leveling to suit) Bricksafe folder is here: http://www.bricksafe.com/pages/OzShan/Crawler Firstly, a couple of my favorite reference pictures; The build itself started with the trucks, thinking that the pneumatics and LAs would dictate the scale. First proof of concept - build a coupling to give height, pitch, roll and yaw to the truck. The pneumatics need to be on their own gimbals too. The reinforced 2x2 rounds slide and rotate in the 4x4 macaroni's. It is on the limit of what will hold together without glue, but it does hold. The two 1x2 technic bricks at the base of the 2x2 column are helped a little by a string (not pictured) which runs up through the 2x2 rounds with the axle. Initial prototype of the drivetrain. I would have liked a higher ratio but there was just no room at this scale. When the gearbox was married with the truck chassis I had to juggle positions, so you will see in later pics the crown gears are facing in not out Best laid plans.... Marry studded and studless they said. It will be easy they said... Showing what will eventually be the steering between chassis and truck. The guide tube and pneumatic cylinders are all on gimbals with the pneumatics coupled together. The average height is preserved during any tilting. With prototypes sorted, it's time to bricklink some parts and quieten down the colour scheme! (thank you 42030 for providing 5L thin liftarms with axle hole in LBG color) You can see the relationship between "guide tube" and cylinders here. The pneumatic system was overhauled too many times to remember but this is what it arrived at. It is all controlled by the direction of the motor. Running forwards drives the pump. When running backwards, the lobes operate the pneumatic valves in series, letting small amounts of air escape each rotation. This lowers the chassis in a slow and controlled manner. The motor can be turned on or off and run in either direction at any time due to the valve timing. Early attempts with PF Servo and NXT servo just couldn't park the valve reliably and after a few operations I would hear a slow leak. I have been trying to keep up with the LDD but it's hard to stay motivated when I know I'm just going to have to suck it up and move over to LDraw if I want to include all the motors, pneumatics and LAs Works so far; (I'll make the files available if anybody would like them). I found LDD essential in the early days to plan ahead and simply find parts, but later on the build overtook it. Original 'box' pump. 1x PF XL motor, 4x 6L pumps running at 90 deg to each other. Very smooth but bulky. Flatter attempt in the same vein. The truck itself. The final design for the height control modules. I hope you enjoy the build so far. More pictures to come of chassis, steering, leveling and interior details. I'll leave it to others to decide what 'theme' it belongs to .

Thanks. There are so many competing factors going on here, it is one of the things that drew me to the project in the first place. The biggest challenge was the fact that the chassis is basically hollow, so supporting everything mechanically and keeping the look was a battle. So yes, lots of things are on the limit of what plastic can tolerate :) In terms of speed, yes that is going to be a problem. I like that it has some good speed but it really should be slower. There is no space in the drive train for higher reduction or extra stages in the gearing. The speed it runs at is going to make locating the MLP on launch supports difficult. I will be re-visiting the worm drive because that did give a much higher reduction. The thrust of the worm gear against whatever is supporting the axle generated a lot of Lego dust though. Not really a long term solution, I'd be forever replacing bushes and beams as they get drilled out. Glad everyone is enjoying the developments :)

Haha, I recognised the avatar :) I'm sure you'll have a lot of fun.

Thanks guys it's been a long road :) Mahjqa, to be more specific, the drive sprocket and road wheels are not exactly riding up off the treads, but they do click against the sides. From your picture the example in the top right would show the same situation. The treads get laid on the ground at a larger radius than the actual circle. As the truck rolls over them they are pushed sideways across the ground towards the center of the turning circle, then pushed out again before leaving the ground at the rear of the truck. A quick sketch; It's still driving well. I'll take the tracks off after the show and see how it's all holding up (and how much Lego powder has been created in the drivetrain )

Hi, yes it has been a while . After sorting the problems with the drive it took ~6 weeks for Bricklink to catch up. Drivetrain is working well now. The tracks themselves have some issues where they catch and pop when cornering. It is because they are not being fed straight on and it tries to climb off them. Not much can be done there except reduce the turning angle and not turn as sharply. Another small update: Still lots of tweaking and minor improvements but the major parts are there now :) For those going, it's on display at the Gold Coast Brick Event this weekend (can't be there myself unfortunately).

Looks awesome so far. I'm also keen to see final pricing :) Nice how you've kept #1 aside. Interested in numbering the first 100? Will you have to lay a bucket up each time because of the ports around the teeth, or is it possible to make two male patterns with which to make subsequent silicon moulds?

Surprised at that side shot. Wouldn't have thought they were so similar. Looking forward to seeing more of the styrene process, thanks for sharing.

This is fantastic. It looked good before but now that it's painted it looks properly finished. Looking forward to seeing it mounted for a test drive.

Fantastic! The gap at the end of the sequence really is tiny in the scheme of things. Great to see something so different.

Battled with this one for a while. I've tried NXT and power function approaches. None were as compact as I wanted, only one of my attempts so far has been reliable. Lengthening the pneumatic arm is a great way to improve sensitivity, but the trade-off is speed. I'll share a few of my attempts, I'm not afraid to point out where they went wrong :) The single valve approach suffered from reliability issues returning to center. Both NXT and PF servo motors struggled to return without any deadband that left the valve a open a fraction. My preference is two valves. This one was based on Sariel's autovalve but uses 4 valves for one function. It is designed to have separate central pump and tank supplying it. All it does is allow small amounts of air into or out of the cylinder depending on motor direction The next iteration used only two valves but with the addition of its own pump and was a little more bulky. Drive the motor forward and the cylinder extends, run in reverse and it retracts by indexing through the two valves in series. Only one valve is open at a time though so there is no issue with valves sticking open and draining too much. It worked OK and the 4 piece linkage working the valve is kinda fun to watch but the spring return meant that the mechanism had to overcome the spring and the valve resistance making it work more than 2x as hard. Final iteration is similar but it doesn't need the spring return and runs a lot smoother. If someone does have something that is compact and provides a controlled release from the cylinder I'd be keen to see it. I like the ideas ^ using a restrictor to limit how fast the air can escape. You could also restrict the flow by weaving/wrapping the hose between pins or axle and bushes to crush it a little. Then you wouldn't have to keep track of that one hose you put the pinprick in 6 months ago

Great review Jim. The shots are just fantastic. Thanks for sharing :)

Really enjoying this thread. I've also been using the poor mans remote of shutter on a timer. Maybe slightly off topic, but I did find that leaving the camera on the tripod when taking videos reduced the shaking significantly. Doesn't weight much but adds a lot to the rotational inertia :)

Sure, have a play around with this; http://www.bricksafe...rTruck mkII.lxf After trying the worm drive I have a different set of problems. It is generating a lot of Lego Dust at each end of its axle at the half bushes because of all the thrust. This is only from a few minutes testing so I'm not keen to keep going this way. Another issue is that I cannot run beams side to side and that allows just a little flex when the load comes on and the worm gear is catching on the Pole Reversers. It generated a lot more separation pressure so rather than limit the number of holes visible and use short 1x2 bricks, full length beams keep everything from separating now. My bad for sacrificing rigidity for looks. I had to significantly modify the steering arms to get the changes to fit and that made them a lot more flexible and the list of cons gets longer. I do like that it is so much slower but that's about it. Going back to the original bevel gear approach (which is a little closer to reality) and making a number of modifications has helped. No more gear or part separation, right to the point where the motor stalls on a new set of batteries. The m-motor axle is now properly supported and the gear ratio ended up the same 1:6. The new bevel gears seem to hold together much better than the 24t crown gear. Even with a little play they have not skipped at all. A productive couple of days after all .

Thanks for the encouragement, I guess it really isn't finished yet . I can't really swap out all the m-motors but I have another plan. After having a good look around the truck I can see a little flex in a few places and the 8t mounted straight on the m-motor with just a 2l axle is just not going to last the distance. It wasn't something I wanted but the area is so tight. Overall ratio would drop from 1:6 to 1:24 which I'm looking forward to. Pity about the friction and backlash but lets see how it goes. It might even get me out of another problem where at full steering lock the speed difference between the inside track and outside was too high and the inside track power dropped below the minimum required to keep it moving. The irony is that during the test runs of the first real crawler they had massive drive and bearing issues and were picking up pieces of melted and fused bearings off the crawlerway after they self-destructed

Great build so far. Just read through the whole WIP and a couple of things hit me; 1- You had the guts to tear it down and start over when you weren't happy (and do it publicly) 2- You are pursuing a realistic chassis, probably at the expense of extra grief. I don't see you taking the easy way out anywhere! Looking forward to seeing the rest as it develops.

Well done on a fantastic creation. Nicely presented too. The colour scheme is right on the mark. Deserving of HOF and technic tag :)

It's finished, but.... It doesn't drive so well any more Last weekend while it was trying to level itself and drive, one battery box cut out then the second and as each recovered it tore the drivetrain apart a little more before I shut it down. Somehow I lost track of the modified battery boxes and they had migrated off into other models. I tracked them down and thought I'd be ready for round two but there is more damage in the drive and it is still binding somewhere (probably more than one place). It might need a bit of a re-think in some areas where the gears are separating too. Looks like it was engineered a little too close to the edge because when one thing failed problems popped up everywhere. That's all going to take a while to fix so for now I'm going to leave this with the latest build pics and a big thank you to everyone for their kind comments. On the NXT screen (this end is the master) the three columns are "Height set point", "Actual Height" and "Difference". Cheers, Shannon.

Really nice tow truck and so many features! I really like the cutaways too. It shows just how much work has gone under the hood :)

I got to 5000 in LDD and that was for a bare frame chassis and the trucks (there is 700 in each truck). Since then then I've added - walkways - all interior details inc floor - roof panels I'm guessing about 7000 now but it's hard to tell Are you using NXG? Im using Lejos, and NXC before that. They seemed fine. There is about 100ms between channels so you notice a difference between channel 1 and 4 starting. Haven't noticed anything between red and blue. I'm using combo commands that write both red and blue together. Any of these help? I'd also like to extend a big thank you to Philo and all his work here: http://www.philohome.com It was my starting point for this project and it helped marry up PF and NXT (and RCX).

Thanks, I'll keep digging around. Instructions aren't really the focus for me, but steps might be the answer.

Thanks again for all the comments. Haha, well that did hold things up for about a month and I worked on it in two halves for a lot longer... I left out a detail pic of the PF bay last time. Here you can see the 4 PF receivers and the NXT IRLink transmitter. The two receivers towards the back are driving the height control, the two to the front are driving the traction motors. I've put together a small clip of the height control system. I'd really appreciate any comments or advice here too because you guys generate such great looking material. Moving from LDD over to LDCad has been a trial but I'm really starting to like it. Looking forward to trying out the flex tools soon. Cheers, Shan.

Moving from LDD to LDCad at the moment. Took a while to find some things but I'm really liking now. It really is a great piece of software. I was looking for this kind of select tool. In the meantime how do you suggest to select multiple parts? Apart from Ctrl+click.

Really cool write up, thanks.

That's a great looking truck. Fantastic details too.

Glad you got the issues sorted out. The landing gear and its sequence is my favourite.