[GBC] Akiyuki Ball Factory

[Berthil]

22 hours ago, dunes said:

This was by far my most complex technic built ever. Had fun and was really exhausting.

Glad you liked it :)
I hope the BI was not exhausting. I tried to build in logical steps and keep them fun and challenging.
Challenging like a puzzle but not in such a way that parts are hidden or one has to take apart stuff 5 steps later to fit something.
Also at the end of every BI Part components are finished for a better sense of achievement before going to the next :)  (no half way 'cuts').

[Berthil]

Render 381 of 1200 renders to get to a 20 second 4K movie of the Ball Factory V3. From the top bar you can see this will take 435h / 24h = ~18 days and its 6 days underway. But as the model gets complexer also rendering time goes up so probably it will take about three weeks. Most likely some part of it will turn out of sight in the video as it will turn twice. I hope not too much otherwise I have to start it again smaller or without turning.

 

Edited May 24, 2019 by Berthil

[rskamen]

@Berthil a few months ago, someone, maybe you, posted a link to a company that sells 3mm tubing in bulk. Not sure what thread it was on. Do you know what company that might be? It might have been in response to your entrance and exit ramps on the Invisible Lift.

[t0mster]

Hi rskamen

I ordered this one: https://www.aliexpress.com/snapshot/0.html?spm=a2g0s.9042647.0.0.70d64c4d2deEWj&orderId=900130519619801&productId=32778220810
But its not delivered yet. What i have red on forums is that you need 3.2mm (not 3mm) for a good fit.

[Berthil]

1/8 inch at McMaster Carr for U.S.A. 1/8 inch outer diameter (they don't ship to Europe),
Flexible Nylon Opaque Tubing on this page in different colors:
https://www.mcmaster.com/catalog/125/168
e.g. black:
https://www.mcmaster.com/5548k81

and modular Germany 3,2mm outer diameter:
https://www.modulor.de/en/abs-round-tube-opaque-coloured.html
Brass insert from Modular, recommended to use with these rigid hoses otherwise they knick instead of bending as they are quite hard.
Very good for straight applications though.
https://www.modulor.de/en/round-rod-brass-oe-1-5-l-1000.html

Edited May 25, 2019 by Berthil

[Ankoku]

Brass would be a bit hardcore.

[rskamen]

Thank you @t0mster and @Berthil. Very quick response. Yes, the McMaster is the site I was looking for. I am back home now and can begin sorting and ordering the parts I need to rebuild my version 2 Ball Factory into the version 3. Can anyone comment on the stability and reliability of version 3 as compared to version 2. The build looks so much more solid and less flexible, which has to add to a far better performing GBC. I am going to completely document and video my version 2 before I begin to dismantle it for parts. If I had the room to store it, I would leave it intact, but I don't have that option.

[9v system]

i used this stuff on my ball shooter run and is perfect

https://www.ebay.com.au/itm/1-8-Hose-Flexible-Nylon-Yellow-Ryken-Pneumatic-Air-Line-Tube-Per-Meter/281843046827?hash=item419f2799ab:g:X1cAAOSwhcJWJdX5

[Jonas]

Hi Berthil. I am progressing slowly but I have almost completed the raw portion of the build. The next portion will be tuning.

What is not clear to me, is the loop-back exit. If you look at the picture, the 2 hoses are positioned too far from the left ramp exit and there is no barier that would direct the rolling balls to the path made by the hoses. (The other exit seems to be OK.) In your video, where you show the loop-back, there is a brick built barier that seems to work well.

[Berthil]

@Jonas It seems the hoses need to be further back, I will correct. I might have forgotten to show the brick build barrier which is needed to stop de balls. I will check and add it. Thanks for noticing.

[Chipstertrain]

Hi Berthil

Just a slight observation if you are amending the instructions, there is a blue 3l pin added in step 40 which floats until step 42 which is where I think it should be added

Thanks again for all your work, I'm thoroughly enjoying this, albeit very slowly, build

Chip

[Berthil]

@Chipstertrain Thanks. New versions of part 1, 3 and 4 with corrections have been uploaded.
A 1x10 DBG plate and 1x6 DBG tile have been added to the parts list.

Edited May 29, 2019 by Berthil

[Jonas]

Berthil, you are superfast and your instructions are excellent. You have been able to do in one person what large designer teams in TLG do in a year.

Allow me another small comment. When you make an update of part 2, look at black 6L axle added in step 167 (the axle to rotate the black lever). This axle passes only through holes and hence it is not fixed in its position. I myself fixed it in step 168 by adding a bush (e.g. 3713) in between black Axle and Pin Connector Angled #1 (32013) and black Pin Connector Plate with One Hole (32529).

And a question to the large basket wheel. Its stand (4 vertical 13L liftarms) seems to me like 'hovering' just one brick high above the tiles laid on top of a 2x6 brick. Yes, it is fixed in place by the 4 red axles but one needs to move the stand back and forth in order to align the holes in the stand and the platform. I added another 2x6 brick on top of the first one and now the stand sits on the tiles and the holes are aligned perfectly. 

And before I end this post, I add one personal (lame) experience. I spent almost one hour to find out why the large wheel - when added to the platform - did not fit with the rest of the system. I searched for mistakes here and there, just to realize (and remember forever) that the wheel stand is NOT symetrical and that it must be placed in one specific direction. If you add such a notice to the BI, it may help some other builders.

[Berthil]

@Jonas thanks again! I will check the Wheel alignment but it should be so that the black crossbeams (with 1x1 connectors) between the 4 13L liftarms should rest on the 1 x 5 thick liftarms already connected to the frame. These 1x5 liftarms are also guides when placing the wheel. But I will check, may be something got shifted during the last fix, in the finished model in Studio it is sometimes hard to see with all these parts surrounding it :)

I will check the rest and update/improve accordingly.

Edit, I checked Wheel position with 1x5 Liftarms positioning/rest and this is correctly placed in the model.
I will update the BI if it is not clear enough. Part 4 needs to be updated anyway, I see now that in current PDF all callouts somehow have been converted to pages instead of callouts.

Edited May 29, 2019 by Berthil

[Jonas]

Thank you, Berthil. Now, it is clear. When building the wheel stand, I did a minor modification (because I was missing one specific element) and this caused the above mentioned troubles. I apologize for a false alarm. 

[Berthil]

1 hour ago, Jonas said:

Thank you, Berthil. Now, it is clear. When building the wheel stand, I did a minor modification (because I was missing one specific element) and this caused the above mentioned troubles. I apologize for a false alarm. 

No problem. Everything is there for a reason (except the tiles) :)
Part 4 has been re-uploaded, quite some work to correct all the callouts again.
Meanwhile an iOS update stopped the two week rendering so I have to start over.

[Ankoku]

I feel as though @Berthil has already spend more time on the instructions than he did on the digital model creation itself. This is why people are generally not keen on producing instructions.

One thing people like to be able to do, is spin the model round a bit, to get a better view of some aspect of it, so it may be worth doing a bunch of hi-res renders at random PoVs, so people have a chance of seeing things from different angles. Especially underneath.

[Jonas]

The build is complete. I spent almost all week-end to make it run. Unfortunately, the run is not smooth, yet. When driven manually, I can feel significant resistance at some phases. I even had to replace the clutch gear by a standard one.

I tried to locate the source by switching off individual modules and it is evident that the biggest sources of resistance in my build are all the lever-driven mechanisms (bob, ball picker, basket shifter). I checked the sliding axles (by Berthil's axle tester), optimized the position of the pivots, etc. but there is almost no improvement. When driven by motor, one can hear that it roars irregularly. I must take a break, now and try it later.

 

[Berthil]

Your red elastic band seems very stressed. Try putting less stress on it by placing the 1x2 thin liftarm on the long axle lower.
It that is not possible (because Bob does not get low enough) stretch your elastic band by hand up to its limits a few times or use on older one that has been stretched.
Or run it without the elastic band when running manually, it's only there for the high speeds.
When listening at the sound and looking at timed components also have a look at the bucket emptier timing and see if it doesn't push against something that can't move there.

I also see a hesitation in the ball picker when moving to the basket. I can't see why as you didn't film the back. I don't think that causes the problem but might want to check that out. Sometimes lubricants cause a stick-slip effect so I used no lubricants on the long axle sliders anywhere.

Not a cause of the problem but you might want to bend up the hoses right at the switchable ball exit so balls get away faster there, at the end of the video you have a line-up there.

Other than that it seems to run well, I wouldn't call it a fail. Nice job!

[Ankoku]

If you don't get the ball picker correct, that can cause a lot of strain. Fixing that from the v2 instructions back to how it was originally in Akiyuki's version made a huge difference. Obviously, Berthil followed those changes, but it shows the potential for the amount of strain the picker can produce.

Never had an issues with Bob. Shifter was all the way wrong, fixing it back to the original completely changed it, so it is hard to say with that one.

[tismabrick]

@Berthil

Just noticed that on rebrickable (https://rebrickable.com/mocs/MOC-24605/Berthil/akiyuki-ball-factory-new-style-v3/#bi) there‘s instruction part 2 v1.6. But if you download it it says v1.5. Might just be a labeling thing though...

Cheers, Thomas

[Berthil]

3 hours ago, tismabrick said:

@Berthil

Just noticed that on rebrickable (https://rebrickable.com/mocs/MOC-24605/Berthil/akiyuki-ball-factory-new-style-v3/#bi) there‘s instruction part 2 v1.6. But if you download it it says v1.5. Might just be a labeling thing though...

Cheers, Thomas

Yes, that's just labelling, I will correct it the next time, thanks.

[Jonas]

Thank you, Berthil, for your advice. Although I think, I have tried most of your hints during the tuning. The run is still too irregular with several movement 'hesitations' as you call it.

I have made another video, this time with more detail and at varying speed. If you find time, please have your expert look at it and maybe you discover something. I have made some minor modifications but I dont think they are critical. You may notice I changed the shape of the warped rod that controls the basket unloader. 

What is a still puzzle for me, how did you manage to clear the rolling path towards the out-box. In my video I show that there is so little space under the second chain (which is hanging down, moreover).

 

[Ankoku]

Couple of things: Have you made sure all your axles aren't sticky and all the rotation points are nice and loose?

[Jonas]

Good point, Ankoku. I just cleaned all the sliding axles by soaped water, but no improvement. Rechecking all the rotating points will be more laborous as it means disconnecting all the modules and re-synchronize them after that.