interesting possibilities, found a used system.

[cgi.ebay.com]

Its an Invision Si-2, the seller is a couple of hours drive North of me. I contacted 3d systems (you would be amazed how helpful people are when you call from a billion dollar corporation) and they quoted me $5 per in^3 ($0.32/cc) for the cartridges. How does this compare to the cost elsewhere? The salesman was trying to point me towards a more advanced system ($60K to start) and suggested that the older systems had reliability issues that have since been fixed, he may have been just trying to keep me away from a used system but its worth noting.

Ideally after running this thing until it falls apart we could disassemble it since it is a UV cured acrylic system. Or perhaps after a dozen or so printings start trying to refill it.

If anyone is interested in potentially going half (or third, or quarter) with me drop me a private message. I still need to research the relability question however.

Comments welcomed.

Mike

The thoughts and ideas expressed in this post do not reflect those of my employer and are intended only as communications between individuals. Any attempts at implement are at your own risk

Edited 1 time(s). Last edit at 09/12/2007 09:13PM by ohiomike.

I am unfortunately in a serious financial bind right now, otherwise I might have been interested. Anyway...

That machine makes me optimistic. 100mw UV solid state laser? That's the average power output for some of the homemade N2 lasers.

[www.3dsystems.com]

32L @ $0.32/cc ~ $10,000 vat price initial investment. I wonder if Fernando can educate us as to how much of a markup this is?

I'm on the other side of the Atlantic, so I can't join in on this, otherwise it's a fantastic opportunity! Imagine all the tests you could do....

100 mW laser? that's pretty achievable, right? I wonder what wavelength though. If it's in the short wavelengths like 253nm, 100mw is pretty potent.

32 liters of acrylic are around 24 kg of it, at this moment I can get the acrylic mix at around 20 to 40$ a kg, depending on photoinitiator, filler content and copolymers.... If I find a wholesale provider, this price may even be cut to half. Right now it would sum between 480 and 960$ for 32 liters of resin... heck, i guess they took that price, rounded it up to 1000$ and multiplicated it by 10, just because they like ones and zeros, specially zeros... scary!

According to the data sheet, the wavelength is 354.7nm.

Wait so if I'm reading that right, their fluid has a markup of 1000%??? That's crazy! Would you be able to use the standard acrylic in their machine, or do they put something funky in it to force you to use their stuff?

Kyle

By the way, I'm very interested in this on a provisional sort of time-share basis. Student == no money, but I need to get those darwin parts somehow...

354.7nm ? Isn't that excatly what the Air/N2 laser delivers?
That's actually looking very promising...

The only thing that would make the resin they use more expensive is the use of exotic photoinitiators. The one I know that works at 354.7nm is Mechelers Ketone and it's derivates. Those cost around 20

337.1nm actually.

Well at 354,7nm they seem to be using a Nd:YAG laser.

[www.rp-photonics.com]

[www.bigskylaser.com]

But Air/N2 laser could also do the trick.

I'd love to go in "halfsies", but I can't afford it. The most I could justify would be about a 5% investment. That'd be enough to maybe get a set of parts out of it.

If you want to take small investments, paid off in parts, then you'd be free and clear on the original, I would be willing to paypal you fifty dollars. Think you could find 18 other suckers^H^H^H^H^H^H^H^H customers?

I offered in a PM to put in $150 on that same sort of principle. Just 15 more, now...

Where did the 32 Liters figure come from, is that the size of a new cardrige???

How much did this machine cost originally?

I don't know about the original cost, but the original cut sheet on the machine is linked in the second post in the thread.

Dont know what this particular model cost orginally but the same level of machine today goes for $16K. I got a demo model offered to me for $12K during my search. This is the most basic of systems, anything more advanced jumps in cost to $40-$60K. Although I should say I got a functioning Stratasys 1600 offered for $3K but without software. I contacted Stratasys and they said they no longer support or sell a compatable software package for that system.


The major issue is that at $5 in^3 for build materials, I would have to charge a couple thousand just for materials to build a Darwin. But if I purchase the systems and try and refill the cartridges with homebrew, I risk destroying a thousand dollar machine.

Mike

How does the machine work? Does the resin just sit in the vat waiting to be exposed to the laser or is the resin pumped around or does it touch any critical parts of the machine? If it does, it's indeed a risky decision right now to go for a homebrew mix. At least as long as I can't certify a fairly long shelf life for the mixes, no uncontrolled polymerization and probably also find a solvent to remove any unwanted hardened stuff.

My understanding is that it is an inkjet design with the material in a cartrige and the UV source riding next to the nozzle. My issue is that if I over or underdo the viscosity or cure speed the result will be a clogged nozzle, which might rank anywhere from a minor annoyance to a major issue. I am simply to unfamiliar with the machine to guess. Partially cured resin would likley just need to be dug out of the port, but pot life becomes an issue unless I can figure out a way to flush the system after use. And of course refilling will automatically make an enemy of 3d systems and void any warrenty on the machine.

The other issue will be fooling the chip on the cartrige to make it forget that it is empty after I refill. Possible but not really my skillset.

The more I think about this the more convinced that mold making is the way to go with this or any other system. A set of molds would likley take $2K to make but then could potentially make a lot more parts for a lot less.

Mike

Mute point, the auction was cancelled.

Mike

Still interesting to talk about, though. I had a different view of the system, especialy since it was an older unit.

From my understanding of the system, it was a standard vat SLA. Vat of liquid and an X-Y galvo aiming the 100mW solid state laser from the top (looked to be CW? I'll have to check the normal modes). They improved accuracy by trading in for settling time, so it built slower. You also reduce your X-Y envelope so the laser doesn't have to play out to the edges and you get higher positioning resolution. You drop the vat a mere fraction, then play the laser over it. This baiscally forms an instant skin. Drop a skin's thickness, repeat. It takes time because of the extremely thin layers that get hardened each time. Needs UV blocking glass (probably polycarbonate) and absorption materials near the top to keep laser bounce from hardening parts it's not supposed to. A lot of the mass of the machine is there to keep it stiff and reduce the effects of exterior vibration on both the laser alignment and on the liquid pool surface.

I think this idea has a good deal of merit but I was curious where you got your figure for thousands of dollars for a set of Darwin parts? I tried to look at the STL files and add up the volumes for a set and came out to something less than 50in3 (not including support material)? Did I just mess something up in the calculation?

I would possibly be willing to buy an old stratasys (if I could find a good deal) and start making parts kits. I think that this would be the best way to jump start the project because judging from the forum comments I think quite a few people are in "waiting for parts" mode.

I called stratasys and they said a cartridge for the 1650 runs $425 for 116in3 material, $125 for 29in3 support and 10 foundations for $125....

Anyhow, I just wanted to keep this thread alive because I think it is a logical next step, even if someone can get there Darwin printing in the somewhat near future it will be a while before full scale replication could be happening and many of us (like me) don't really have the skills to repstrap an alternate machine.

I can't remember the exact volume of the parts, but that sounds about right. The reason for the 4-digit price for the parts is that we are quoting the price that a commercial RP company charges for their services, which seems to be about 10x the already-absurdly-marked-up prices that Stratasys charges for their materials. If you have a Stratasys, the parts cost will be closer to $300 or so (which seems to be supported by your info as well). The problem we are facing is that we as a group do not have one, and only some of us have access to one (at a university, for example).

thats not a bad idea... check this out:

[cgi.ebay.com]

the current bid is only $1000. if the RRRF were to acquire its own FDM machine, then it would be much easier for us to get parts out to people. there are 5 days left in this auction, so i'll be keeping an eye on it.

i think i *might* be able to front the money for it, but if people are really interested... please post here.

I was mistaken about the volumes, I had thought that the quotes were for just the materials not for a commercial printouts.

My next thought was that based on conversations and websearches I have located a manual 3d printing system, RapidPro, that takes a STL cross sections it and plots it on a vinyl cutting machine. RapidPro costs $900 and a basic automatic vinyl cutter costs $400. Assuming 50% usage of materials bulk vinyl would cost $1.50/in^3 so thats about $75 per set of parts. But you have to strip and stack the layers yourself (it uses locator pins to line up the layers).

$1300, but thats for a new and fully supported machine and it assumes that the resultant vinyl material will have the stiffness necessary to handle use. Or that the material can be resin treated to increase rigidity.

Charge $150 per set and payoff would be about 18 part sets. Assuming the material has no major issues.

Mike

The thoughts and ideas expressed in this post do not reflect those of my employer and are intended only as communications between individuals. Any attempts at implement are at your own risk

Edited 1 time(s). Last edit at 09/12/2007 09:14PM by ohiomike.

Seems like this one. [www.boxford.co.uk]

"RapidPRO imports a solid model and, using a special

Surely, until the RepRap Darwin is able to make its own parts, it is not worth building since you won't be able to make much else that is useful. After all, its parts are not particularly complicated objects. A soon as it can, parts will flood out exponentially. Why try to jump the gun by buying an expensive machine?

---

[www.hydraraptor.blogspot.com]

heres why buying a commercial machine is a good idea:

darwin is a beta design, meaning we're still testing it, and we dont know if it really is going to work or not. until we have a fully functional darwin that is actually printing real parts, we have a chicken/egg problem. what that means is that if we run into bugs, or have improvements that need to be made, we have essentially no way to make them (save to make them ourselves, beg/steal/borrow a university model, or pay $$$ to get them printed)

if the RRRF has its own stratasys machine, then we can provide parts (and modified parts) to researchers, at cost, until we finally get to a design that works well enough to replace the stratasys for producting parts. also, we might even be able to buy normal ABS plastic filament from a non-stratasys supplier, and use that. that would reduce costs even more, from $200-$300/set to <$100. if we can pull that off, then a major hurdle and barrier to reprap development would be eliminated.

Nowhere is it written that the STL's that you use to make parts for Darwin on a commercial machine like the Stratasys are going to be usable for what Darwin will actually be able to print.

Doesn't it make more sense in terms of both capitalisation and running costs to use wooden equivalents of the STL's like Joost is doing? Somebody with some good woodworking tools could put out a LOT of parts in very few man-hours for very attractive prices.

Zach,

It is funny you mentioned the ebay auction because that was what I was looking at as well.

Any idea if that thing could operate in a basement or a garage? I know you are supposed to keep the material free of moisture but that has a seperate compartment for storage that you could add a dessicant bag to.

My initial exuberence over buying it was tempered slightly when I saw that it was a lot bigger than I thought and weighs around 300lbs...

Forrest Higgs Wrote:
-------------------------------------------------------
> Nowhere is it written that the STL's that you use
> to make parts for Darwin on a commercial machine
> like the Stratasys are going to be usable for what
> Darwin will actually be able to print.

you're partially correct. but nobody is suggesting that we *have to* use the exact same design that we use on the stratasys to print parts with a darwin. this is about bootstrapping ourselves into a fully working design. of course being able to use the same STL's on both machines is the easiest route, and something we should strive for.

> Doesn't it make more sense in terms of both
> capitalisation and running costs to use wooden
> equivalents of the STL's like Joost is doing?
> Somebody with some good woodworking tools could
> put out a LOT of parts in very few man-hours for
> very attractive prices.

i dont think so. first off, you have to find a woodworker. secondly, its not replicable. the work is being done by a human. the whole thing that makes reprap so interesting is the idea that the production of parts is something you can replicate and automate. if we find one person to do it, then it becomes much harder to replicate. whereas since we are using a stratasys, all you have to do to replicate it is get yourself an equivalent machine and the files.

also, rapid iteration of the design is another huge benefit. if the woodworker makes a bunch of jigs for a specific design, and we change the design... everything is worthless. on a machine, its not big deal.

yes is a bit pricey, but if we want to really kick this project off, then this (or something like it) is looking like the best bet. i would much rather see a darwin jump straight to printing things, but i have a feeling we'll need to do a few iterations before it becomes ready for mainstream usage.

my $0.02.

People who can do basic woodwork are a VERY common in the States at least. Go into any Home Depot.

i understand that it is possible to do woodworking, and you're right on that. but what we're talking about here is producing parts that are:

a) identical. even the best woodworkers will find it hard to make parts that are identical every time, to 0.1mm. as for the worst... well lets not go there

b) in decent quantities. nobody wants to sit there and make the same part over and over again for us researchers. thats what machines are for. with a strat, we could crank out a set every 24-36 hours.

the price on this machine is very attractive. if its still at a good price when that last day rolls around, i'm going to be very tempted to buy it. theres no denying that having one of these solely for the use of the RepRap project, we would be able to make much faster progress.

Zach,

Looks like that auction is from an educational institute of some sort, perhaps if you emailed him and tried to see if you could work a deal for a low price in the interest of research the guy would go for it?

There are no bids now so he could still cancel the auction. Anyhow thought it might be worth considering but I am not really a "core member" so didn't want to solicit him in the name of the group or anything.

Have a good weekend.

Zach, I'd be willing to help with the cost of the Stratasys.