[MOC REVIEW] Akiyuki's Ball Factory

[bj51]

Ok thanks ! I'll take 100 then, just to be sure.

[DrJB]

About 4 years ago, there was a seller on eBay who had bags of 25 balls, each for $10.00 ... I bought 4 bags (2 are still sealed). So far 50 balls were 'plenty', though I agree that with the factory, I may have to open the other 2 bags.

[bj51]

I've finished building all the modules and I've found a problem with the instructions (PDF from Blackbird): on page 199, step 9, I can't see where the 1x8 tile goes.

Now I'm off to time the whole thing. Gonna be fun...

[Daniel-Technic]

I've finished building all the modules and I've found a problem with the instructions (PDF from Blackbird): on page 199, step 9, I can't see where the 1x8 tile goes.

Now I'm off to time the whole thing. Gonna be fun...

I may be thick but where do you find the instructions from?

I want to build it as I have enough parts but ball is other thing.

Daniel.

[bj51]

First post here:

http://www.eurobricks.com/forum/index.php?showtopic=117305

[Daniel-Technic]

First post here:

http://www.eurobricks.com/forum/index.php?showtopic=117305

Thanks

Daniel

Edited April 12, 2016 by Daniel-Technic

[Jonas]

Hi to all GBC fans,

thanks to Rebricker and Blakbird I was able to build my own copy of Akiyuki's legendary Ball Factory. Actually, it is not an exact copy, it is a slightly modified and eXTend version, and that is why I call it Ball Factory XT.

After I finished the standard version, I felt there was still some empty space on the left baseplate. I wanted to add something that would have made the ball path longer by adding several smaller modules. They had to fit into the given space without hindering the view on the brilliant Akiyuki's original modules. So, I added one longer slide, a watermill wheel, and a ball relay module inspired by NxtStep101's Mechanical Chain mechanism.

Here are two pictures showing the complete model:

and a video:

[jesuskyr]

Hi to all GBC fans,

thanks to Rebricker and Blakbird I was able to build my own copy of Akiyuki's legendary Ball Factory. Actually, it is not an exact copy, it is a slightly modified and eXTend version, and that is why I call it Ball Factory XT.

After I finished the standard version, I felt there was still some empty space on the left baseplate. I wanted to add something that would have made the ball path longer by adding several smaller modules. They had to fit into the given space without hindering the view on the brilliant Akiyuki's original modules. So, I added one longer slide, a watermill wheel, and a ball relay module inspired by NxtStep101's Mechanical Chain mechanism.

and a video:

Wow! Great^^

[wumsi]

Cool video! I noticed, bob is dropping the buckets not far enough on the wheel. if you push the 1x2 thin liftarms a litte bit more to the inside you could adjust that :D

[Jonas]

Cool video! I noticed, bob is dropping the buckets not far enough on the wheel. if you push the 1x2 thin liftarms a litte bit more to the inside you could adjust that :D

I have just noticed it. To fix it, it took me some time to re-learn - by try and error - what should I do. You seem to be an expert.

I think it is really necessary for me (or for the community, too?) to make a list of the most frequent troubles and their fixes. Now, I still remember most of the tricks from the building and tuning phase but I am sure that in several weeks I will forget them.

Edited April 30, 2016 by Jonas

[wumsi]

I have just noticed it. To fix it, it took me some time to re-learn - by try and error - what should I do. You seem to be an expert.

I think it is really necessary for me (or for the community, too?) to make a list of the most frequent troubles and their fixes. Now, I still remember most of the tricks from the building and tuning phase but I am sure that in several weeks I will forget them.

I'm not an expert by any means. It took me hours to make my ball factory run smooth, like you with alot of try and error :D

[Blakbird]

Wow, this is really a great addition. I like it so much I wouldn't mind adding it to my model!

[bj51]

Ok, so I showed the Ball Factory at the French Brick Expo last weekend, and here's my feedback...

It was a nightmare ! It woulnd't run more than 2 minutes without something breaking. Here's a sample of the problems I had :

• Balls get stuck in the bucket dumper,
  
• the chain on the second bucket conveyor would jam and tip buckets over, even though I reduced the size of the brick at its start to 2 plates,
  
• the ball picker would not release balls, or release them before the bucket, or stress the entire system which would jam and lose timing,
  
• the 8L axle with stop on the pusher mechanism would fall and balls wouldn't be pushed anymore,
  
• the long axle running from the bucket wheel mechanism to the ball return conveyor would disassemble near the bucket wheel, thus losing timing again.
  

I stood for the first 4 hours behind it, and then I just stopped caring and just turn it on for a few seconds, when people seemed interested, which they were a lot, unfortunately for me. It was such a mess that in the end, I had to manually help the ball picker go left and right or It would break.

Next year I'm trying other Akiyuki models, I hope they'll be a little bit more reliable.

[Blakbird]

Ok, so I showed the Ball Factory at the French Brick Expo last weekend, and here's my feedback...

It was a nightmare !

The ball factory has a lot of failure modes. You will notice that even Akiyuki himself does not use this module in his public layouts. His other modules are much more reliable.

[nychase]

Wow, this is really a great addition. I like it so much I wouldn't mind adding it to my model!

I agree, I think this is certainly worthy of being added to the real factory. Besides, you kept with the one motor concept and you really maxed out the space and kept everything interesting.

Consider this my vote to make it happen!

[Jonas]

Besides, you kept with the one motor concept ...

In the published version, the added modules are powered by an additional PF motor (M type). The reasons are just practical:

a) The main motor has already have large load. Why to stress it more?

b) The new modules do not need any synchronization with the rest.

c) Any additional module connected to the main power source would make the tuning (or re-tuning) of the critical parts more complicated.

On the other side, the added modules run rather smoothly, so the main motor would power them, definitely, if really necesary.

By the way, I measured the current consumed by the the main motor powered from a 9V train trafo and I was suprised it was only 0.3-0.4 A. I expected more. However, I did the measurment just by a common digital A-meter, so I can hardly say if there are any significant current peaks in those moments when the motor drives the critical phases, namely the bucket-wheel shift, bucket platform shift or ball picker movements.

I did these measurments because I plan to control the model from an Arduino board (equipped with an appropriate interface).

[DrJB]

By the way, I measured the current consumed by the the main motor powered from a 9V train trafo and I was surprised it was only 0.3-0.4 A. I expected more ...

Not to digress but the numbers seem low. To measure current, you need to 'break' the circuit and connect the motor 'in series' with the multi-meter. This is not an easy thing to do with the PF connectors. Have you tried comparing your numbers against philo's ?

Edited May 4, 2016 by DrJB

[Jonas]

Not to digress but the numbers seem low. To measure current, you need to 'break' the circuit and connect the motor 'in series' with the multi-meter. This is not an easy thing to do with the PF connectors. Have you tried comparing your numbers against philo's ?

I have a background in electronics, so I know how to measure current.

Here are some of my recent projects

http://www.brickshel...ry.cgi?f=562970

a here some historic ones when I was controlling Lego by an 8 bit microcomputer.

http://www.brickshel...ry.cgi?f=441799

Sure, I know Philo's excellent page but as I cannot measure the torque required by the Factory setup, I do not know the actual operating point on the torque-current curve.

Still, the values I measured seem to me much lower than I was expecting.

Edited May 4, 2016 by Jonas

[DrJB]

1. I have a background in electronics, so I know how to measure current.

2. Here are some of my recent projects

3. Still, the values I measured seem to me much lower than I was expecting.

1. My apologies, did not intend to be 'lecturing' ...

2. Thank you for sharing, very interesting. I also have interest in Arduino and Raspberry PI ... working on building an IMU (Inertial Measurement Unit) now.

3. Agree, not sure why values are so small. Hopefully Philo can see this and give his opinion.

[Jonas]

It is OK. I believe that our discussion can be interesting also for some other builders.

What we can learn from Philo's measurements is that a PF XL motor operates between these limits:

no-load current ... 80 mA

stalled current .... 1800 mA

The actual current value for the Factory setup depends on many factors, namely how smooth does each of the modules run. I also expect that the value may fluctuate in a larger range within one cycle (i.e. one bucket wheel shift).

Let us see if someone else come with his/her own measurements.

Edited May 4, 2016 by Jonas

[DrJB]

That is exactly my thinking. Doing a current/torque curve might be a bit difficult (especially due to damping/friction), but doing 2 key measurements (stall vs. no load) should be sufficient.

[Deepy]

Jonas, great development of the ball factory. Your model seems to run very smoothly, have you made any adjustments of the original model in terms of stability? I'm still trying to figure out how I can build my model more rigid. Additionally, do you have a brick list of the additional modules? It looks very good

Thanks

D

Hi to all GBC fans,

thanks to Rebricker and Blakbird I was able to build my own copy of Akiyuki's legendary Ball Factory. Actually, it is not an exact copy, it is a slightly modified and eXTend version, and that is why I call it Ball Factory XT.

After I finished the standard version, I felt there was still some empty space on the left baseplate. I wanted to add something that would have made the ball path longer by adding several smaller modules. They had to fit into the given space without hindering the view on the brilliant Akiyuki's original modules. So, I added one longer slide, a watermill wheel, and a ball relay module inspired by NxtStep101's Mechanical Chain mechanism.

Here are two pictures showing the complete model:

and a video:

[Blakbird]

I found the torque requirements for the ball factory to be very low. When running it by hand with a crank, I could barely feel any resistance. I had no problem running the whole thing on an M-motor. With that in mind, I don't think it would be an issue to gear the extended modules to the same motor.

[Jonas]

I found the torque requirements for the ball factory to be very low. When running it by hand with a crank, I could barely feel any resistance. I had no problem running the whole thing on an M-motor. With that in mind, I don't think it would be an issue to gear the extended modules to the same motor.

You are right. The additional load is not that critical. Yet, when I run the whole setup by hand (without a crank) - usually when tuning some parts - I really feel moments when the drive need to overcome significantly larger resistance. (I mentioned them in my earlier post.)

[Jonas]

Your model seems to run very smoothly, have you made any adjustments of the original model in terms of stability?

You know, in video everything looks better than in reality. One publishes only those shots where everything runs well.

But seriously, now. I have remade the spiral lift tower, recently. Its construction was the weakest point in the whole setup and whenever I touched it accidentally, something got broken.

I succeeded to remove the small gap between the tilted platforms and the rest of the tower and I made the two columns much stronger by adding additional (but almost invisible) support. I need to test it in a longer run and then I will send some photos.