[GBC] The Akiyuki Project (2010 - 2020)

[LucyCol]

This set http://shop.lego.com/en-GB/Maze-21305 also comes with 4 orange balls and it would appear that you can order them as individual items from here wwwsecure.us.lego.com/en-gb/service/replacementparts/sale

You may need to adjust the links for you own country.

[DrJB]

True, but those balls are now as expensive as anywhere else. Close to $1.00 a piece.

[Courbet]

Hello everyone,

Wrong colors, and wrong... planets :

A MPD file will follow next week, to keep Blackbird busy !

[bj51]

Wow Courbet, that's very cool and so soon ! Glad someone did that, I love this module !

[djm]

I posted a comment to Akiyuki's new web site with a link to the first page of this topic. In his reply, he confirmed that he knows about about this project. Kudos to Blakbird.

[Mogwai]

Whow,that`s fine ... Courbet the next Akiyuki ?

[dr_spock]

True, but those balls are now as expensive as anywhere else. Close to $1.00 a piece.

Usually happens when instructions for something popular is made available, some parts become more expensive due to demand. There are other ball options to fit one's budget.

I posted a comment to Akiyuki's new web site with a link to the first page of this topic. In his reply, he confirmed that he knows about about this project. Kudos to Blakbird.

He did ask for Akiyuki's permission beforehand...

[Blakbird]

Wrong colors, and wrong... planets :

A MPD file will follow next week, to keep Blackbird busy !

Wow, that was fast work. You are certainly keeping me busy all right. I've been working on Technicopedia updates lately which have caused me to spend a bit less time on this project, but I am still working on it.

I posted a comment to Akiyuki's new web site with a link to the first page of this topic. In his reply, he confirmed that he knows about about this project. Kudos to Blakbird.

I love his new website. I actually tried to use the comment form there as well, but couldn't get it to work. It kept timing out. I'm glad you confirmed that he is pleased with the project.

[Blakbird]

I am making good progress on the Crane Loader. I have draft instructions for the crane, switch, and motor and have collected the parts and started building. This is a really giant pile of parts! I am about 30% done so far. I've been improving the instructions as I go.

One thing I noticed in Akiyuki's 2012 video is that he uses two trains. One waits on a switch track until the other passes. This would be good in a layout because the crane module is pretty slow so this would double your throughput. Courbet, any interest in whipping up a file for the switch track just for completeness?

[Courbet]

I can build a second train and give it a try, then quickly make a file if it's working.

[Courbet]

Here is my take on the siding track for Akiyuki's GBC train:

- the central part is the most accurate to Akiyuki's design. The external parts are left crude, modelled from the older video and maybe already wrong (the SNOT 2x2x3 slope is my guess)

- this will not work without a smooth axle joiner into the driving ring on the train. When I began working on Akiyuki's GBC Train, that was my first choice but when I tried to make a loop with the crane loader, I found out the train would not go all the way and would stop in a middle of a curve. Then I changed for an axle joiner with ridges and it worked much better - but it won't with the siding track, unless maybe using a very worn out one.

- in my experience, it also works better with a longer siding track, so I added a straight track to the design

- even so, I may have missed something because it doesn't always work

Edited December 1, 2016 by Courbet
delete attachment

[Courbet]

Here is a MPD file for the Planets module, the last one of Akiyuki's flawless designs.

The only thing I could not deduce from his material is the area beneath the "black hole" turntable, including the gearing from the main shaft so I improvised there using one more gearbox part. Otherwise, I believe the rest to be quite faithful to Akiyuki's original.

GBC_Planets.zip

[Blakbird]

Here is a MPD file for the Planets module, the last one of Akiyuki's flawless designs.

The only thing I could not deduce from his material is the area beneath the "black hole" turntable, including the gearing from the main shaft so I improvised there using one more gearbox part. Otherwise, I believe the rest to be quite faithful to Akiyuki's original.

Thanks for another file and more work for me to do! Did you see the photos on Akiyuki's blog? He has one from the bottom that should help with the gearing for the turntable.

[Courbet]

I couldn't have done that much with the video only. The bottom view was indeed very helpful but still I couldn't decipher that area mostly hidden behind the outer frame in the picture - and the all-black build didn't help.

[Blakbird]

I couldn't have done that much with the video only. The bottom view was indeed very helpful but still I couldn't decipher that area mostly hidden behind the outer frame in the picture - and the all-black build didn't help.

It also did not help that the picture could not be enlarged. Sometimes you can click on the picture to get a bigger version, but not this time.

[Blakbird]

I have good news and bad news. The good news is that I finished the instructions for the Train Module with Crane Loader. If you build everything, it is 262 pages of instructions and >4500 parts. The bad news is the module has has lots of problems.

The first problem is traction. The train wheels do not have enough friction to enter the module with enough inertia to engage the mechanism. They slip very easily. I solved this problem by removing the hard rubber bands from the train wheels and replacing them with some rubber bands my daughter uses to make bracelets. They are 1000 for a couple of dollars. This completely solved the problem. Now it has enough traction to stall the motor if you stop the train.

The next problems are with the motor module (see picture below).

• In the orange highlighted area you can see that the wheels are not constrained. They can slide along the axle. They need to be midway between the clutch gear and the side rail to fit in the tracks, but this also means they can (and do) slide around and screw things up. We need a way to put in a 1/4 stud spacer to lock them in place.
• In the yellow highlighted area you can see the bevel gears that drive the wheels. These gears are not braced so they can (and do) easily skip. Ideally, there would be a gearbox housing around them. But at the very least, the train side rails need a cross brace to hold them together. When the white rails push apart, the gears skip. There is nothing solid holding the rails together, and it is not obvious how to add something without interfering with some part of the mechanism.

The next problem is on the crane. The yellow wedge is supposed to push in the axle on the train and put it into neutral. However, on many occasions the axle rides up over the top of the wedge instead. When this happens, the train does not go into neutral and may even derail. The wedge cannot be made higher without interfering with the trigger. I tried putting a 1/2 bush on the end of the axle but then it was even worse.

The final and most serious problem is power. The train barely has enough power to move the crane, and in many cases it just stalls. I checked all the joints for friction and do not find any issues. In fact, the whole thing runs very easily by hand. Perhaps much of the drag comes from the knob wheels, but in any case the crane does not reliably work without enough power. I tried changing to an L motor which provided plenty of power but caused even more gear skipping in the train bevel gears.

I had a crazy idea for how the train motor could be given a little help. The axle highlighted in red below connects the timer/trigger mechanism to the crane and transmits power. A differential could be added here in series with this axle, and a second motor could be attached to the diff ring gear. If a pole reverser was used at the end-of-line buffer, this motor would only run when the train was present and would provide some extra power, then shut off. If the second motor were unpowered, the diff would allow the train power to simply bypass it.

I am hesitant to release the instructions until these problems are solved. On the other hand, it will be hard for anyone to offer solutions without building it first!

[jesuskyr]

• In the orange highlighted area you can see that the wheels are not constrained. They can slide along the axle. They need to be midway between the clutch gear and the side rail to fit in the tracks, but this also means they can (and do) slide around and screw things up. We need a way to put in a 1/4 stud spacer to lock them in place.
  

for a 1/4 stud spacer, what about http://alpha.bricklink.com/pages/clone/catalogitem.page?P=11010#T=C

[Mogwai]

@Blakbird you write : interfering with the trigger. I tried putting a 1/2 bush on the end of the axle but then it was even worse.

I think to all the Modules (unloader/loader/crane loader/switch) Detail with the shock absorber add a plate on the groundfloor to make it higher.

to all the bricks ,not only the yellow wedge.i can not try this , at the moment on assembly and home only on weekend . I hope you understand what i mean

[Blakbird]

On 5/16/2016 at 4:20 PM, Blakbird said:

....The good news is that I finished the instructions for the Train Module with Crane Loader. ....The bad news is the module has has lots of problems.

On 5/17/2016 at 2:48 AM, jesuskyr said:

for a 1/4 stud spacer, what about http://alpha.brickli...age?P=11010#T=C

That's an expensive spacer! The ring is less than 1/2 stud but still more than 1/4 stud in width. I might give it a try if I can find a couple of them.

I spent most of last night working on the problems and I managed to solve most of them although my solutions may not be for everyone.

For the problem with the switch mechanism hopping the wedge, the problem was that it took too much force to switch the transmission. I changed to a smooth axle joiner which solved the problem but then had a but too little friction and would slip out of gear. I rotated the smooth axle joiner 90 degrees to use the "tall" side instead of the "low" side against the driving ring and now it works perfectly. Not a single problem with this since.

The skipping of the bevel gears and the lack of power are really the same problem. If the power requirements are reduced, then the gears don't skip. I managed to fix it by adding a motor. I realize this is no longer faithful to Akiyuki's design, but it work. I put a motor on the back of the crane module on the axle that passes over from the hopper. This drives the crane much nearer to the power requirements instead of so far remotely.

To turn the motor on and off I used a pole reverser. A little 4-bar linkage at the end buffer is impacted by the train which starts the second motor. A mechanism connected to this is hit when the trigger fires. When the trigger hits the switching axle on the train, this axle pokes out the other side. I used the force of the trigger on the other side of the module by picking up the motion of this switch axle to shut the switch off. There is also a stop to prevent the pole reverser from ever going backward, which is bad.

The train motor and the secondary motor are still mechanically coupled. If they were not connected, the crane would get out of sequence with the trigger. The secret is to run the second motor just fast enough to help but not fast enough to actually backdrive the train. Once I got it all set, it works perfectly. I ran it for about 1/2 hour with no problems.

I would still like to build some conduits to hide the ugly wires which are messing up the beauty of Akiyuki's module, but otherwise I am pretty happy with it. I am going to release the instructions soon. I don't think I will include my changes in the official instructions, but I can provide details if anyone is interested.

[DrJB]

Hat's off to you for pulling all of this together. As much as I'm happy to see instructions available, I'm not that excited about your instructions causing BL prices to go up and depleting the supply ... can't be always happy I guess.

[Blakbird]

Instructions for Akiyuki's Train Module are now tested and complete and are available here. This Module is really made up of a bunch of smaller modules. I've listed the parts lists sepately for those who don't want to build the whole thing.

• Motor (157 parts)
• Switch (167 parts)
• Unloader (324 parts)
• Siding (106 parts)
• Loader (603 parts)
• Crane (3046 parts)
• Complete Set (4569 parts)

Most of the credit for reverse engineering this module goes to Courbet who did the work making an LDraw file with instruction steps. I had the easy job of formatting the instructions.

The Train Module is really a series of smaller modules, not all of which are required to make a layout. At minimum a motor, loader, switch, and unloader are required. The crane loader can be swapped for the simple loader. If a siding is added to increase throughput, then an additional motor is required. If you want to build everything then there is a loader, crane loader, siding, 2 motors, switch, unloader, and connecting tracks. The instructions show how to make 3 different layout options, but of course there are others. The images below show how many of each sub-module would be needed for each type of layout. The last layout uses the most parts, but still does not include the loader because it uses the crane loader instead.

 

 

This is a really complex module, perhaps even more so than the Ball Factory. Unlike the ball factory, relatively little timing is required. The crane itself is self timed because every mechanical motion drives another. The triggers need to be timed to fire at the right point in the cycle. This is pretty straightforward. There is some additional timing in the "timer" of the crane module at the worm gearboxes, but if you orient the parts as shown in the instructions it will work correctly.

As has been elsewhere discussed, I had problems with the power requirements of the crane. Anyone interested in adding an extra motor to solve this problem can see my previous post on the matter.

[Courbet]

@Blackbird, I was going to suggest you the switch to a smooth axle joiner, I'm glad it solved your problem.

Strangely, I did the exact opposite to solve mine (train stopping in a middle of a curve).

Also, when I tested the crane loader lack of power was not an issue, the M motor was only suffering a bit with a full load of balls.

What I spend most time on was the tuning of the trigger mechanism. Did you find a way to tune the 2 gearboxes quickly ?

[Blakbird]

Also, when I tested the crane loader lack of power was not an issue, the M motor was only suffering a bit with a full load of balls.

What I spend most time on was the tuning of the trigger mechanism. Did you find a way to tune the 2 gearboxes quickly ?

Who knows what is going on with my power problems. Obviously it works in Akiyuki's video, although he does mention the M motor really strains on his blog. I did check all the sources of friction and found that it took very little to run the crane, so I don't know what the problem is. If people can make it work without my extra motor, that would be best!

Edit: I've added a lousy iPad video which shows the operation of the module using the changes I made.

[Blakbird]

On 5/17/2016 at 1:19 PM, Courbet said:

Also, when I tested the crane loader lack of power was not an issue, the M motor was only suffering a bit with a full load of balls.

What I spend most time on was the tuning of the trigger mechanism. Did you find a way to tune the 2 gearboxes quickly ?

It bothers me that I don't have enough power when your video shows that your module works fine with the same design. Tonight I am going to go a friction hunt, isolating each system until I figure out where the problem is.

By the way, I also have a draft ready for the elevator and planets!

[Courbet]

@Blackbird, good luck in your investigation.

For the Elevator, the 100cm string from the 42042 Crane is just the right size and length. I used it without any knot, using Axle 3L with Stud to tie it to the elevator cars.