Help wanted with 3D printer selection

More information is good. From my ownexperience printsover the 20cm range are going to face soem warpage due touneven cooling that 20cm range just fits in the sd3 range however if i were you youwill want larger build areas over and above that to accomadate slightly larger parts. I'd recomend you look at the medal max maybe that has a build area of 9 X 12 X 8 if i recall correctly but some assembly is required its on the lwo end of yoru budget then you can spend extra building an enclosure to help retain heat

also you might want to invest in a new hotend one that can go up to around 300 but if you do that you will want a holder thats not printed plastic as well the plastic could melt.

that will give you a range of printign options from 140c on up to 265 which is what nylon prints at

every printer is going to require tweeking in all honesty. at 150 cm the sd3 is perfect and your budget would allow for 2-3 printers and enclosures that way youwould never be without a printer unless you broke everyone.

heh why am i haveing flashbacks to the original star wars mmo where you build a tool to make a better tool to make a better tool .... lol some peopel may get the refrence ...damn you sony for ruining it

As much as I've come to dislike the company, I think the maker bottom replicator 2 (not X) is right up your alley. It will be pretty much ready to go out of the box. It only uses pla but it is supposed to be pretty easy to get the same results every time. And with any FDM printer, you can put a pause in the g code insert whatever you want in the part, then finish the print.
Depending on print size, an sd2 expert might work. It's a lot less money. Can print abs or pla, and has a heated build platform and an enclosure. Software will be harder to use. But not something you couldn't overcome.
I'm not overly impressed with the cubify. Small build platform, proprietary (expensive!) filiment cartridge. But it does have very user friendly software.

Look at Ultimaker as well, since it is in your price range.  They have a reputation for being best in quality and reliability.  Avoid the cube, they use the printer ink model, with very expensive filament in chipped cartridges.  Also I've heard that there is far less control which is great for simplicity and usability, but settings are not one size fits all.

Also, Makergear's Mosaic has a great reputation. I only know that I've liked their groovemount hot ends & plastic supplies.

Makergear's M2 is a great machine and has a fully assembled option at $1750. The ultimaker is good machine, but outside the USA (if the is a concern of yours).  It is also only available in kit form.  So the M2 wins there.

But I also second the idea of having multiple Solidoodle 3's if you are going to need to pump out a bunch of prototypes (prints can take longer then you might think) and you can have the peace of mind of a backup machine if one goes down (don't undervalue this if you have deadlines to keep). 

If you MUST print different colors at the same time, the Makerbot Replicator 2 or 2X are the most viable options for you.

An ABS capable printer might serve your needs well, as you can print your designs in parts and weld them together using a variety of methods (this can satisfy your sensor needs). Plus it can be easily polished and cleaned up using different acetone techniques.

But no matter what 3d printer you get, someone is going to end up learning more about the mechanical nuisances of the machine than they might currently think.  Make sure someone is up to that job or you could have some expensive paperweights or sub-quality prints on your hands.

Do you really need to print the whole object all at once?
Can you assemble the printed objects from smaller pieces?

You could, for example, create the objects so they interlock as alignment for gluing together.

If you print something out of ABS, nylon or PLA it'll be plenty strong enough to grasp as long as it's not hollow and thin wall and the temperatures involved do not weaken it.  Keep in mind how strong an egg is and it's both basically hollow and thin walled.

Though it is out of your price range as a purchase, laser sintering could produce metal objects plenty strong enough (you can have this done for you online).  So you could start with 3D printing to verify the design and then send it off to be made.

However, let us think outside the 3D printing cube (pun intended).  You might be able to sculpt the objects from ShapeLock, InstaMorph (whatever we call it this week) pellets after you boil them down and let them solidify a bit.  Then you end up with an entirely solid mass unless you introduce air bubbles or voids yourself.  If you want to say make a hollow ball, make 2 pieces press the hollow into both and then assemble them with whatever you want in the middle.  This has the advantage of not requiring more skill than playing with clay and you don't even require a kiln.  Plus you can easily melt it down and reform it any time you like.  Want more of them?  Stick the end result in a 3D scanner and there you go.

Just to give you an idea.  I've seen robots that weigh 100+ lbs drive around on 3D printed gears.  The question of strength can come down to the mechnical knowledge of the designer of the printed object.

Also if you really need strength and you design strange shapes you could pour it yourself out of aluminum using some homebrew casting.  For anyone that's never handled liquid metal though, that's probably very trial and error.  (Yes I realize that this environment he's describing does not lend itself to operating a forge, but you can melt aluminum on a barbeque).

The problem I see from working with a variety of printers is that most things that print ABS, PLA, or nylon in your price range will likely produce a surface texture that is characteristic of 3D printing.  You could use acetone to help with that but that's a subtractive process and again depending on how you do it will impact your ability to reproduce dimensionally identical models.

As others have suggested the closest solutions will be stereolithography or machines that use a clay base.  A stereolithographic machine (even the cheapest which are in KickStarter now) is out of your price range.  The price for that accuracy is durability.  If your object is mechanically designed properly you can alter the structural durability.  I have plenty of examples that were printed using either process that will withstand basic handling but are not suited for making gears or other wearing parts.

Again I see little stopping anyone from starting with a basic printer and then after verifying the design sending out the files for production.  Initially I would recommend making a few samples to get an idea of the durability of the process.  Start with truss and voids.  I am very confident that you will find a way to produce what you desire though it might not appreciate being dropped or being submerged in liquid.

Another option I can offer instead of removing material to smooth the printed surface with a solvent like acetone.  You could cover the surface in a building primer, an autobody glaze or rubber like paint (bedliner, PlastiDip, that stuff they show on TV, or LiquidLatex).  Any of these would fill the voids between the layers on the surface of a printed object and at the same time add resistance to liquid intrusion.  The impact on the dimensional accuracy would only be limited by your control on the process unlike removing material with acetone in which you will have the impact of the material you printed to consider.

I know what you want is possible to save you time and money perhaps my most useful suggestion is get samples till you find one that appears to suit your needs.

As an aside, casting in copper or bronze is just as easy as aluminum and given the amount of copper floating about I would be surprised if it interfered with your magnetic sensors.  If it did there would be no way to put the printer or anything else with electricity in the room.

Thanks for the reference to Miicraft.  I hadn't seen that one yet.  It's got kind of a tiny build area however.
Build size (footprint): 43 mm (X) x 27 mm (Y) x 180 mm (Z)

Where as the Form1 is about 100mm x 100mm x 165mm.

Above you mentioned that your average print will typically be about 150mm x 150mm x 150mm obviously you can break things up into pieces but still that's something to consider.  You may want to inquire with these companies if they can offer you any solutions to how to part your models down.  I can see parting down solid geometric objects in the slicing process, but more complicated shapes might require some tinkering on a part by part basis.

Also this is something to consider.  FormLabs hit a hiccup with 3D Systems suing them for patent infringement.  As far as I know the early orders are slotted to go out sometime in April 2013.  So you might want to consider that it could be an issue for you or not.

The last SLA system I worked on was definitely not something for the home user and it was not cheap to operate.  I've noticed that these projects have been clever to avoid addressing some questions about the output they can produce.  I hope that when they hit the shelves they really present the market competition they appear to be.