compucarve

[brianlanning]

I just received an email from sears. Looks like they have them in stock again. $1899.

brian

[Art Ransom]

As a FYI I am building a 4 axis mill that will let me route 20" by 128" and turn a wood column 24" by 128". Total cost $4000. see my site for details.

[Swingman]

Art,

Your website, and your philosophy, should be national treasures!

Ludwig Kieninger ... what a master! I know you do, but continue to cherish that association. I have a few in my life (music related, not woodworking) and what a privilege it is to work with the truly gifted!

May you keep up the good work, and the wisdom, for a long time to come.

[brianlanning]

I'm a software developer. I got into it because of my fascination with computers. But it leaves me empty because at the end of the day, I havn't created anything tangible. So I do woodworking at home as a hobby and love it very much. Because of these two big parts of my life, I'm very interested in the marriage of woodworking and technology. I like the legacy mill, but I'm sort of put off by the purely mechanical nature of that machine, as well as the price and limited capacity. So naturally, I invisioned a technological solution to this problem. I've just moved and I'm still completing my shop, building skills, shop tools, etc. I also have a large list of honey-dos to finish. Still, I'd love to spend time exploring this.

This carveright/compucarve thing has made me think about more than just turning spiraled columns though. I've been thinking lately about regular shopbot-style 3d carving. I've also heard that this new gadget has inputs for a 4th axis, but no attachment exists yet. I suspect they'll evetually created a sled that will do 3d carvings. We'll see what the capacity is, but it obviously can't be more than the thickness rating of the current machine. I've also heard rumors of larger machines in the works.

As a programmer, I have a very clear idea of what the software should look like, features it should have, and how it should work. The hardware is a bit fuzzier though. Although I have a sort of loose preliminary design in my head that involves stepper motors, threaded rod, captured nuts, and make-shift linear bearings. I'm fairly certain these components won't deliver the accuracy needed. But it is a starting point for experimenting with this.

If you or someone else were to come up with a kit with all the hard to find parts and directions, at maybe a $500 price point, I bet it would be very successful.

This is an exciting time in woodworking. I suspect the next ten years will see some wild new machines.

brian

[Morris Dovey]

| As a programmer, I have a very clear idea of what the software | should look like, features it should have, and how it should work. | The hardware is a bit fuzzier though. Although I have a sort of | loose preliminary design in my head that involves stepper motors, | threaded rod, captured nuts, and make-shift linear bearings. I'm | fairly certain these components won't deliver the accuracy needed. | But it is a starting point for experimenting with this. | | If you or someone else were to come up with a kit with all the hard | to find parts and directions, at maybe a $500 price point, I bet it | would be very successful.

There are no hard to find parts; nor is there any shortage of how-to information. The problem with the $500 price point is your desire to have someone else spend that amount and add a considerable amount of their own time and effort without charging you for it, so that you can then say the $500 price tag is more than you care to pay.

Time for some basic arithmatic - you're going to need two decent linear bearings on each axis at, say, $60 each (check your catalog of choice) - that's $360, which leaves you with $140. Then you're going to need a minimum of three micro-steppers at $140/3 = $46 or (better design) four micro-steppers at $140/4 = $35. I think you'd better go with the three-stepper version - that way you have a whopping $2 to spend on the controller, the motor mounts, the remaining linear motion control components...

Fortunately that "someone else" has a free shop filled with free tooling and has no need for groceries, telephone for customer support, clothes, medical care - and is overjoyed to spend his life gratifying the desires of people who want to have and be without earning and learning.

| This is an exciting time in woodworking. I suspect the next ten | years will see some wild new machines.

Absolutely. Question is: will they be worth buying?

(Hint: What percentage is worth buying /today/? Is the situation improving or worsening? Any idea why?)

-- Morris Dovey DeSoto Solar DeSoto, Iowa USA

[Lew Hodgett]

You won't even get a replacement part for $500, at least not me.

Start about $5K and you might have a chance.

Lew

[CW]

Careful what you wish for. I do woodwork at home as a creative outlet from my day job to. I run a shop full of CNC lathes and mills.

[Morris Dovey]

|

Possibly I was a tad harsh. On a web page at the link in my sig I posted photos of every step (except soldering up cables) in the construction of a 3-1/2 axis machine with a 1/4800" step size; and posted a drawing of the chassis to news:alt.binaries.pictures.woodworking before starting construction. The photo of the final step is still on the web page, along with a photo taken during the testing stage and another of the stepper controller guts.

I put in more than 100 hours researching the parts and will estimate at least another 500 hours in actual development activity. The cost of (all new) parts ran a little over $1400. None of the parts were hard to find. Identifying the most suitable parts and finding reliable sources at reasonable prices /was/ difficult.

While you may be able to do a better job in less time, I'll guess that you'll need to spend at least as much time as I did to do a comparable job. If so, you can feel secure in knowing that your time is worth $500/600hr or approximately $0.83/hr as a machine designer. Alternatively, you can allocate the whole $500 to hardware and feel insecure knowing that your time is worth butkus and that your "kit" has only toy value.

You probably wouldn't guess from the photos now on the web page; but the chassis of the machine is entirely of baltic birch. It was a genuine woodworking project.

-- Morris Dovey DeSoto Solar DeSoto, Iowa USA

[Leuf]

The machine I think he was referring to doing something similar to was not really a true 3 axis machine, but more like a printer. So sort of

2 axis, one of which only having a few inches of travel, plus the feed mechanism to move the wood through. Whether that actually reduces the complexity of the machine or not I don't know, but if Sears can sell it for $1900 then I'm guessing there isn't $1400 of parts in it.

Remember you are spreading that cost over all the units you sell, so if you could sell 100 machines development only accounts for 6 hours per machine. And if you were going to build one for your own use, well you had to invest all that time anyway so it doesn't really count. What really matters is how long it takes to build one now that you know what you're doing and have the prototype to refer to.

I've been thinking about trying to sell drum sander kits, but I wouldn't consider the time I put into building my first one as part of the cost. I was going to do that anyway, and the sander is paying for itself.

-Leuf

[Prometheus]

That is truly excellent, Art- hope I'll be making one of my own some day!

[Prometheus]

Yep. (Though I'd still like to make a home-brew cnc toy just for the sake of it someday- a waterjet cutter might be nifty, and help add some decorative metal panels to my wooden furniture...)

It's also amazing how theraputic beating the crap out of hunks of hot metal with a hammer is after running carefully engineered parts all day- and it's a cheap way to make turning tools.

Here's my rule of thumb- if it's for making money, a guy needs the best technology he can find. If it's for putzing in the shop, remembering the old ways and means of doing things is often more fun and satisfying. I'd much rather make a part on an engine lathe and a knee mill than use a zillion-dollar milling machine, but there's not a good way to compete in the market that way. Same thing with furniture and turning- only with that, I don't have to care about making money, so I stick with what is fun. If you ask me, watching a computer make something while you become a loader/unloader and QC fixture loses it's charm pretty quickly (though it's not a bad way to pay the bills!)

[Prometheus]

Harsh, yes- and a little out of character, too. But nothing you said wasn't true, and I hope you're not taking it back. Hell, just getting a little propane forge up and running took *at least* $500 if you count the research and fabrication time- and that was with free tech support from a guy who used to engineer heat-treating ovens, and the assistance of a very talented welder who worked at least one minor miracle for me (brass pipe welded into a cast iron elbow- I've had weld shops threaten to kick me out for even saying "cast iron".)

And that's just swirling gas and air in a big tube that will make cold stuff hot without burning the house down. I doubt a guy could build a CNC *anything* for less than that.

And we thank you for it- or at least, I do. I know how much work goes into anything like that, and it's truly appreciated when guys like you are willing to share. That same spirit of generosity has saved me years of painstaking research over time, and allowed me to do much more with my time than most people even consider.

So, a tip of the hat to you- and don't fret about telling a guy off. I have to listen to pie-in-the-sky crap from people who demand a Ferrari for the cost of a Pinto all the time too. It's nice that this group is so remarkably civil- but every once in a while, a guy needs a well-placed smack in the head to get him back to reality.

[Morris Dovey]

Leuf wrote: | On Thu, 18 Jan 2007 21:54:50 -0600, "Morris Dovey" | wrote: | || Morris Dovey wrote: || ||| || || Possibly I was a tad harsh. On a web page at the link in my sig I || posted photos of every step (except soldering up cables) in the || construction of a 3-1/2 axis machine with a 1/4800" step size; and || posted a drawing of the chassis to || news:alt.binaries.pictures.woodworking before starting || construction. The photo of the final step is still on the web || page, along with a photo taken during the testing stage and || another of the stepper controller guts. || || I put in more than 100 hours researching the parts and will || estimate at least another 500 hours in actual development || activity. The cost of (all new) parts ran a little over $1400. || None of the parts were hard to find. Identifying the most suitable || parts and finding reliable sources at reasonable prices /was/ || difficult. | | The machine I think he was referring to doing something similar to | was not really a true 3 axis machine, but more like a printer. So | sort of 2 axis, one of which only having a few inches of travel, | plus the feed mechanism to move the wood through. Whether that | actually reduces the complexity of the machine or not I don't know, | but if Sears can sell it for $1900 then I'm guessing there isn't | $1400 of parts in it.

From the published specs, I'm fairly sure you're right.

|| While you may be able to do a better job in less time, I'll guess || that you'll need to spend at least as much time as I did to do a || comparable job. If so, you can feel secure in knowing that your || time is worth $500/600hr or approximately $0.83/hr as a machine || designer. Alternatively, you can allocate the whole $500 to || hardware and feel insecure knowing that your time is worth butkus || and that your "kit" has only toy value. | | Remember you are spreading that cost over all the units you sell, so | if you could sell 100 machines development only accounts for 6 hours | per machine. And if you were going to build one for your own use, | well you had to invest all that time anyway so it doesn't really | count. What really matters is how long it takes to build one now | that you know what you're doing and have the prototype to refer to.

Absolutely true in every way - although the "if" should probably be set in uppercase bold italic. Throughout the project, I worked up a bill of materials and a vendor list so that I could offer the little monster as a product with (standard) parts. I did receive a number of e-mails expressing interest in purchasing copies at various (below-parts-cost) prices. I guess it must have left me with a raw spot for people who aren't willing to give/receive value-for-value.

Even having built one and having the part programs, a bill of materials, a vendor list, and a routing list on hand doesn't reduce either the parts cost or the labor content to anywhere near the level that people were willing to fork out for. FWIW, as soon as a machine moves from the "something I'm building for myself" into the "product for sale" classification, the rules change rather noticably.

I'd label your last sentence as "conditionally true" - because it's only so if the customer wants no further products or parts, /ever/. Product development/improvement activities /do/ have associated cost, and you're disallowing the use of sales revenue to fund those activities. I'm inclined to view that as (at best) "unwise".

| I've been thinking about trying to sell drum sander kits, but I | wouldn't consider the time I put into building my first one as part | of the cost. I was going to do that anyway, and the sander is | paying for itself.

Yup - I'm aware of your efforts and I think you did well. I'm just holding out for someone to offer a 45-pound complete, lifetime-guaranteed, dustless, self-feeding version with keyboard/digital thickness adjustment and drum/feed speed controls, and incorporating a paper clog prevention feature - all for $250

-- Morris Dovey DeSoto Solar DeSoto, Iowa USA

[Morris Dovey]

Prometheus wrote: | On Thu, 18 Jan 2007 21:54:50 -0600, "Morris Dovey" | wrote: | || Morris Dovey wrote: || ||| || || Possibly I was a tad harsh. | | Harsh, yes- and a little out of character, too.

I'm sorry for the out-of-character part. Got "caught in a crosswind" and forgot to bite my tongue.

-- Morris Dovey DeSoto Solar DeSoto, Iowa USA

[Swingman]

Gotta watch that chili in this cold weather, it'll do that to you! ;)

[Leuf]

I reckon a lot of that is a genuine ignorance of how much the parts for something like that actually cost. A little bit of education as you did in the part I snipped can go a long way.

Yup. Making the tool is sort of like making the woodworking project. You might start out thinking you're going to save all this money, but if you factor everything in it actually costs three times what it would have to just buy it. So you'd better either be getting closer to what you really wanted or better quality than what you could buy, or just really like doing it.

I would just look at as what I can sell it for minus what it costs to make has got to leave enough left over to keep it going and have enough leftover to make worth the effort. Something as complicated as a CNC is going to have more "upkeep" for lack of a better word, but maybe hard to quantify that cost until you've been doing it for a while.

I reckon there's a niche in there for a premade drum + shaft + pulleys

• bearings and accessory kit with all the other hardware, a lot of which I had to buy in packs of 25 when I only needed a couple. I don't see room for a whole lot of margin in it, but I don't have enough work to keep me busy all the time and the page I put up for the sander gets more traffic than my real site so it's worth looking at. Seems like everyone and their brother is selling router tables, but I could only find one drum sander kit. And having one now I can testify that everyone ought to have one. Not going to be much of a risk to put together 5-10 kits and see what happens, vs the $$$ you'd have to put in up front. Just got a few projects to get through first.

-Leuf

[Robatoy]

This one?

[purduephotog]

Perhaps a little off topic, but we just started our FIRST Robotics 'build' season (6.5 weeks from start to finish to design, prototype, redesign, build, program, and deliver) a functioning robot to perform tasks.

On the subject of costs: Our little controller PIC costs 459$. Each PWM stepper control costs (memory) 55$. Each motor costs 35$. Each victor (100 hz instead of 38hz) costs 60$. Aluminum is 10$/ft I believe.

Now you can pick quite a few things up on ebay to bring the cost down- it's amazing what I've seen go in terms of micro-stepper controls there. But it isn't a viable business to do that sort of engineering- at least, it's not sustainable.

Hit the 1000$ price point is probably more realistic... or having prepackaged 'drop shipments' from suppliers (AndyMack) might help.

But I am LOVING this description of the build process... and when you're finished I would love to see the offering of your research available for others to see. Who knows, I might finally get a larger garage and have space for that beast!

Jason

[Leuf]

like yours is somebody's copy/improvement of that one. Mine is a copy/improvement of a few sources, using regular sheet paper instead of spiral wrapping.

-Leuf