RepSpider Alpha Build

[jmccartney]

Documentation of prerelease RepSpider reverse engineered from Sho's sketchup mesh. More to come soon.

[sho]

One difference between RepWalker is the diameter of Hole and Axis of joints.
I though, it's easy to use(get) "5.0mm drill" than 5.2mm.
So, I decided diameter of hole as 5.0mm at first and adjusted axis diameter.

But I tryied Rainer's Cring (for 5.0mm axis) to this 4.6mm axis and it was fit well.

[jmccartney]

The main change has to do with the gear/leg holder part of the chassis. I printed them in two separate pieces and removed the spacers so I could print them flat.

Doing that makes it much easier to print, but then causes a secondary problem where the two gear/leg holder parts separate at the top because they are not attached. I am just going to glue mine, but a solutions is probably needed.

[jmccartney]

Off the top of my head the other thing that bugged me was the indent on the long leg ligament. It seems pointless from a functionality standpoint, but it requires to match them up for aesthetic reasons which to me is not worth the trouble. (I noticed I have one backwards which is annoying).

Picture attached.

[jmccartney]

I have added a mount for a cheap ultrasonic ranger as "eyes" for my spider. I also have bluetooth up now. I am not terribly interested in having a "remote" control spider at the moment, I am playing with more autonomous behaviors that can be monitored using BT. I am using working on finite state machines using the SM library which has been a lot of fun. I have the BT working just fine in a concurrent SM without a loop and it works perfectly.

[steamboat]

Hello Sho and James,

do you think you could provide me with a set of improved Spider stl's next week? That would give me time to print it before the Maker faire.

[sho]

Hello James and Rainer,

James
Please see this photo and see inside legs of your RepSpider.
The Shape of "LinkB"(you show me on the picture) need to avoid conflict to "MotorBase".
The eyes on RepSpider looks cute !

Rainer
I attached the stl of 3 type of large parts for now.
You need each parts 2 copies.
2* MorotBase3.stl
2* BaseA10in.stl
2* BaseA10out.stl
By the way, I sent you the assembled RepSpider with motor and battery with the package !

[jmccartney]

(06-26-2014 12:10 PM)sho Wrote:  Please see this photo and see inside legs of your RepSpider.
The Shape of "LinkB"(you show me on the picture) need to avoid conflict to "MotorBase".

I guess what I am saying is to me, it doesn't look like a conflict. There is enough clearance. I will try to flip one around take a picture. I am just thinking reducing unique part count would be good..

[jmccartney]

(06-26-2014 12:10 PM)sho Wrote:  2* BaseA10in.stl
2* BaseA10out.stl

At some point we should discuss the pro's and cons of unique parts. I find it much more difficult to explain/ keep track of one of these builds when all the parts are unique and have to be oriented in a specific manner. Your "BaseA10" part is a good example of this. If you remove the 3 spacers then that part can be used for either side, and either side can receive the spacers....this is partly a matter of preference to. Keeping things symmetrical and flat allows for much easier builds and assembly for the end user.

I guess I am advocating: "less unique parts" > "total part count"..

[steamboat]

Hello James,
I even held lectures about part design for manual assembly at the University of Hannover for 4 years...

As ever there is no best solution for this I think. Because we do not have mold costs and don't generate costs with each part number in the inventory or for spare part keeping it is not the goal to focus on standard parts here - but always a good goal!

I have this order in mind when designing 3D printed parts.

1) sturdy design
2) easy to print
3) easy to assemble without instructions

Regarding assembly it is important for me
1) to have symmetrical parts:
-> make them symmetrical even if it is not necessary.
- Add a hole for instance even if this is only used in one or the other orientation
- this saves time during assembly because you don't have to think about orientation and you can't do anything wrong.

2) to have non matching parts where no math is allowed:
-> take the hip and shoulder axles of the RepWalker for instance
- it needs a different axle length for this what can't be detect by the "worker" easily - so I gave this two different parts a different screw distance to make it impossible to do something wrong.

3) "Over-design" asymmetrical parts:
-> if parts have to be asymmetrical make this easy to detect
- For the RepWalker I did this with the spine-left and spine-right...

This just in short...