Open Desk

[jTr]

sorry - duped - see next post for what I meant to say...

[jTr]

Yep - got a bin full of those connectors on hand.
Have either of you committed to being a fabber for Opendesk? Just wondering if the concept is catching on well. At a minimum, I see making a couple items for myself. Just wondering if it is worth opening the door to being a local fabricator on their list.
(Once comfortable with this whole concept and successful at producing these items, of course!)

Jeff

[EricSchimel]

You might be thinking of the AtFab Open Cabinetm that's got two ply thicknesses. 12 for the dividers, 18 for the box.

[bill.young]

I wish it was that simple!

[EricSchimel]

jTr I am an OpenDesk fabber. I've gotten a reasonable amount of quote requests from them, not much has turned into actual work though. I think if you're going to really make their stuff at any reasonable amount you'll have to market yourself. I've had reasonable success doing that.

Marketing is easy, the big hurdle is learning how to make this stuff. The designs are good, there's just nuance in each one and if you mess it up you'll have loose joints and mis-fitting parts. Pay attention and you can make some awesome stuff.

[bill.young]

As Eric knows all too well I've become pretty curmudgeonly in my old age, but here are some thoughts to help deal with cutting imprecise materials using precise tools.

http://www.shopbotblog.com/2015/10/m...-with-plywood/

Bill

[EricSchimel]

Thanks to Bill, it's taken me a while to really think in the following way when I design stuff...

When I first designed this:

http://digitallyfabbed.com/#/maker-bench/ It was designed with the plywood thickess as part of the design. This means that I'd have to tweak the model for each sheet of plywood. Eventually what I did was build a parametric model so that I could punch in a plywood thickness and it would rebuild my model. Some OpenDesk (and other) designs depend on the plywood thickness being taken into consderation for them to work. In this table one of the most obvious areas where it's important is that stretcher across the bottom. For that to work it's got to pinch on those legs perfect. Too loose and the table wobbles, too tight and you'll be sanding a lot to get it to work.

As I've designed more stuff, I've tried my best to work around thickness. If you think of thickness of material as a variable you can't control, but the profile that the machine can cut out as a variable you CAN control, you can design differently... Take these sawhorses for example:

http://blog.sketchup.com/article/get...tchup-sawhorse

I designed these to be thickness agnostic. You've got to use any 3/4 plywood, and the way I did the joints is that they will work on plywood anywhere from .650 to .76" thick. You can even mix and match. All of the cutting is done zeroing off of the table, not the top of the wood so that your pockets always end up at the same Z height so they line up with the adjacent holes.

Desiging these sawhorses this way allows just about anyone anywhere with nearly any plywood to cut these and have them work the way I designed them to work.

Now, if you look really closely at a lot of the OpenDesk stuff there's a mixture bewtween these two design stiles. The AtFab range for example really depends on plywood thickness being dead on, or having a parametric model. Some of the OpenDesk tables have cleverly done pockets with adjacent "mouths" that slide into those pockets. Both the mouth and that pocket depth are things that the tool cuts out so it's under your control so it will work more often.

Hope that made sense!

[jTr]

Eric,
Good angle of discussion. Typical carcase construction in my shop is done as you describe- zero to the bed and let ply thickness fluctuations fall to the inside of the cabinet = outside dimensions are within a few thousandths of target size. Run of the mill built ins and gang installations such as kitchens go more smoothly with this methodology. I am simply using dados rabbets and pocket screws 95% of the time.

As for the Open Source 100k garages etc, they boldly go into this dicey realm of highly accurate joinery with an inaccurate product, the plywood. It seems to defeat the universality of execution.

Simply put: Thickness of tenons (tabs) must be established by machining with the cnc via pocketing, just as width of mortise (slot, pocket, etc.) is established by the cnc. This will reduce margin of error to a few thousandths vs. measurable fractions in ply thickness. Accuracy of this is greatly enhanced by Zeroing to the bed.

I understand there are those who have purchased machines and have zero drafting/designing skills, and a ready to cut plan is just the ticket. I read Opensource's forum enough to note this, as well as cries for help regarding things not fitting, which appear to garner little to no assistance.

Sometimes, I think we get overly excited about 100% digital fabrication, just because we bought this sexy new piece of technology and we feel we must prove it can do everything. However, if we have to tweak each design, we risk defeating the whole concept of "ready to cut". For those unable to generate their own designs digitally prior to buying a cnc, this must be incredibly discouraging.

Hopefully they can tweak a little more simplicity into this ambitious outpouring of optimism regarding the future of furniture production.
I wish them luck, as their hearts are in the right place.

Jeff

[EricSchimel]

jTr: Question for you:

So what do you think is the biggest hurdle in making OpenDesk stuff? Are the files hard to figure out? I know they send flat DXF files which can sometimes be difficult to decipher. Did you get a better understanding of what you were milling after you saw the SksetchUp 3D model?

[AlexRobinson]

I think it is pretty good