Noob recommendation request

[cwm9]

I'm an inventor and wish to own a CNC mill to prototype and test ideas.

I am not a machinist and have never run a mill before. I am, however, a Physicist and have no problem reading books or instruction manuals.

Can you please recommend a good option for me <$2000?

The intended use is to produce prototyping plastic resin molds/diy plastic injection molds.

Requirements:
Able to mill aluminum/plastic
Complete or close to complete hardware package (no software) with minimal assembly time (not a built from scratch)
Compatible with OTS software like, perhaps, camworks

Preferences:
Precision (rigidity + resolution?) is more important than job speed
Able to work on parts at least 5"x5"x4"

Ideally:
UPS/Fed-Ex-able

Thanks so much!

[Stepper Monkey]

Extreme precision costs, and there are very highly accurate machines in that size range available for semiconductor/lab/etc. work costing $10-30k or more. For <$2000 precision is going to be relative, but can still be very impressive considering.
Don't know if they will be good enough for you, however. What are your expected tolerances?

There are only a few mill packages in that price range and a couple of them won't really fit your bill. The two best candidates are the Sherline and the Taig. Given those two the Taig would I think be the hands down favorite around here overall, and my recommendation as well. They are pretty much both designed squarely at the envelope, materials, and spindle speeds you will be needing. The Taig is both uglier and generally more capable all around, but both are well-built, simple to operate, and with a lot of support and accessories available.


One mill available in the same price/size range you might run across would be the MaxNC, which would not serve well for what you want it to do, and they tend to have a history of some rather disconcerting quality and design issues as well. Best avoided.

Another would be the X2 or other Chinese mill conversion, a heavier-duty if somewhat coarser design than the other choices. It is capable of doing heavier cuts and in harder materials than the other choices, and handling heavier/larger tooling, but as such also has much lower spindle speeds and several other design details geared toward that purpose that would not really recommend it for what you propose.

There are a larger number of choices available if you can go above $2000, especially useful if you want the mill for more than one dedicated purpose. A Taig is a really good choice for your budget, but there are also much larger and more capable mills for a relatively incremental increase in cost. Being an inventor, you will likely find other uses as well for cutting larger items, steel items, etc. once you have a mill at hand. If you have the space and the budget can stretch, you might want to consider it.

[cwm9]

Thanks for your reply...

I don't have any specific tolerances in mind at the moment. My statement was more of, "if there are two in the same price range, I'd rather have the more precise of the two than the faster of the two." I suspect my needs are pretty easily met. Is .01" precision asking too much? I got the impression that all the units out there offered much better than this. Really, the desire for precision is more of a "what might I want to mill in the future" thing than a specific need right now.

$2k isn't a limit so much as it is a comfort zone. I could go up to $3k before my fingers started hurting...

I saw the Romaxx WD-1 and the IMService Servo units... These seemed to be in my price range. I wasn't sure if these were any good, or if they could even be used for aluminum... Also they're in my size range (tabletop).

The mills that you mentioned, are those all CNC mills with everything included?

[High Seas]

cwm9 --
Given the workspace you have defined - seems like a desktop cnc router might be an ok solution too. There are a few out there in that price range - IMServe advertises on this site - and at least one of the members (Widgetmaster - as I recall) has a similar tabletop router he sells too.
Might try the question in the other threads - or try a search on "desktop cnc router"
Welcome Aboard, and
Jim

[Stepper Monkey]

A small desktop router may indeed be perfectly acceptable for you as they are much cheaper. They are also not meant for as heavy of work, so the aluminum moldmaking may come to be an issue. You will have to determine that balance.
One word of advice in any case; Though the routers can generally be just as accurate as the mills, if lighter duty, the spindles are a world apart. The router spindles tend to be much cheaper, much faster, and much sloppier than you need - usually by an order of magnitude on each count. While the usual hand router, a dremel-esque tool, or laminate trimmer choices all make perfectly acceptable and even optimal tools for most routing tasks, they do not work for your needs at all. You will need a much more accurate spindle with a broader array of collet choices.
This isn't a problem, it can be a perfectly workable configuration for you with a small desk router, but you need to be aware of the difference. Depending on choice of spindle, this may also remove much of the cost benefit of a router over a mill, which is another thing to take into account.


If you are feeling either really adventurous or somewhat patient, there is a new supplier of router-type machines that look to be exactly designed for your purposes. They are at www.vivaek.com, and they are cheap and appear to be an impossibly good fit for what you need. One catch, no one has tried them yet! I have ordered some component parts from them and have been happy with the quality and price of those, but I don't think anyone here has sprung for a whole mill package yet. I liked the looks of it enough I'm confident - I'm now ordering one for myself to try out - but it's a complete crapshoot as to how good they really are until one of us actually tears one apart and sees what makes it tick.

[blades]

Really, the only turn-key or near turn-key game in town (that I'm aware of anyway) is going to be vivaek, as stepper monkey mentioned. This gantry type mill has a literally unbelieveable price of $2355. http://www.vivaek.com/cnc3002.html It's really difficult to get a grasp of the actual physical size. It appears to be a sizeable machine, but it only weighs 210 lbs. I would question it's robustness, or apparent lack-of, considering that a little X2 weighs in over 150 lbs.

Here's a turn-key converted X2 mill, but it's priced @ a rather hefty $3500 (not worth it, imho)
http://www.syilamerica.com/product_X2.asp

[cwm9]

Wow, first, thanks for everyones' wonderful help!

It would seem the choice is between a Sherline or Taig or Sieg or similar then... sigh, I can see this isn't going to be a off the shelf as I was hoping. Oh well, that's life! Guess I'm going to have to become a CNC geek.

How much would a Taig Micro Mill run me if converted to CNC w/servos? w/steppers? (I understand stepper motors can loose steps?) I see there's a microproto product that says it is a CNC micromill... http://www.microproto.com/MMDSLS.htm Is that really the same mill with the same capabilities only CNC?

I'm a little confused because the microproto site lists the MicroMill DSLS 3000 price as $2995, where as the Taig site lists a 2026ER CNC mill price as $2155... but they look like the exact same machine! eh?

...

The Romaxx and IMService machines are both ruled out for being desktop routers. (BTW, what's with the desktop router manufacturers like IMService claiming 0.0004 mm step sizes when they don't use ball screw drives, and even if they did repeatability is only on the order of 60x that?)

As to the Vivaek, I went and read the thread... everyone wants to know if anyone has tried them yet. I did some research, and here's what I found:

They appear to be importing this:
http://www.cnlike.com/en_ProductShow.asp?ArticleID=117

Vivaek has internal photos up at: http://site.vivaek.com/cnc_pictures.html

The Vivaek comes with ball screws. I don't know how high quality they are, but the manufacturer web site states .01mm/.0004" The Sherline and Taig come with leadscrews, which I understand to be less accurate and not to last as long.

Is the difference between a mill and a router unit that the mill has a head driven by a belt and a router unit is directly driven by a motor?

If that's the case then I assume the Vivaek is really a desktop router (it actually says that on the manufacturers website), and the ball screws may be... inappropriately accurate compared to the rest of the build?

Regardless, I think "tried and true" is better for me than "buy and hope..."

[Crevice Reamer]

Ballscrews are for rapid motion and low backlash. Acme lead screws are actually extremely high in resolution because of the higher thread count per inch.

For example, if your stepper motor is set at full step, that's 200 steps per motor revolution. A Ballscrew with 5 turns per inch will give very fast motion and 200 X 5 or 1000 steps per inch. That's a resolution of .001.

The same setup with a 20 TPI Acme leadscrew would be slower in motion, but have 4 times the resolution because there would be 200 X 20 or 4000 steps per inch.

The disadvantage of lead screws is that they WEAR so fast and have to be tightened and adjusted frequently. Then the wear over time happens at the point of most use.

CR.

[Stepper Monkey]

You have definitely been doing your homework on CNC! There are a number of logical if incorrect assumptions newbs invariably make when looking at all this stuff for the first time though. A lot of it is pretty confusing. Hope this helps and doesn't make it more confusing!

First off let me say that ballscrews are nice for a number of reasons, but accuracy isn't necessarily one of them. Ballscrews can be less accurate than an acme screw or have higher backlash. I favor ballscrews usually, but one isn't always inherently superior over the other, they both have their place. In your context, ballscrews might be nicer but not a big issue to worry about much either way.

The stepper versus servo issue is a complex one, but it is not cut and dried. For small mills doing this sort of work steppers and servos have such little to separate them sometimes it is almost academic. Different story on a Bridgeport sized machine, but for this it's almost a wash. Steppers can even actually have the advantage, and in your case probably do. "Lost steps" is more a classic sales bogeyman than an actual problem in your context. So is the argument that there is a reason all the "big mills" do it that way, and so by extension paint steppers as some sort of half-measure - the reason the big mills use servos has more to do with the fact they are well, big. Notice that most small high-accuracy devices use steppers regardless of cost. Scale matters here.

The difference between mill and router usually is more a generalization about their intended function than anything. Routers just generally run toward larger envelopes, faster running, lighter materials, and to hold less tolerance than a mill. Routers often are of the gantry type and mills usually aren't, but that isn't even always a good measure to make the distinction.

Linear ball races are a form of linear bearing, and don't have anything to do with ballscrews. In your example the IMS axis actually uses a traditional screw, but runs back and forth using a simple ball-bearing containing tube on the linear shafting, as opposed to a traditional solid bronze bushing.

Finally, resolution is just as much of a joke as you suspect. It is simply the distance traveled for each incremental step. Take the screw pitch times the microstep increment and you have it. Carve a thread onto a stick with a pocketknife while drunk, and as long as you just rotate it a really small amount each step, you get resolution in the millionths! As you can tell, it doesn't mean jack, but it sure sounds impressive. Dealers often use resolution numbers when the real numbers that count - accuracy and repeatability - don't come out looking quite so hot....

Hope this helps. A lot of the answers might be different if you were looking larger, higher budget, production environment, etc. but for what you want there isn't any real cost/benefit for going fancy. The differences are minor at best within the envelope, materials, and work you need to accomplish.

[cwm9]

Wow, stepper monkey, that was REALLY helpful!

Thanks.

What exactly does everyone mean by an "acme screw"? I get that the acme screw has more threads per inch, but is there an actual "acme" screw company? The only acme I know of is the Warner Bros. "A Company that Makes Everything" gag...

That explains the Vivaek: it's probably designed to be reliable and fast for semi-production runs, exactly what would be needed in a labor intensive chinese factory. I guess if you were making 1000's of "carved" jewelry boxes for export to the US, this is exactly the machine you'd want. The ball screws keep it working long term, and the rest of the machine is just good enough to do the job fast and cheap.

For me then, I'll probably go with a lead screw/stepper motor/small mill configuration. Slow and steady wins the race for me. So I'll look closely at Sherline/Taig/Sieg; especially the 2026ER as it is only $2155, so that may be my baby.

[blades]

Originally Posted by cwm9 Regardless, I think "tried and true" is better for me than "buy and hope..."

Got that right buddy!

Now I can clearly see why it only weighs in @ 210. That's a REALLY wimpy looking table casting!

I believe you're right about it being a table-top router. Nevertheless, it looks like it would be very nice for machining anything non-metalic, such as plastics or wood. Haven't seen any plastic injection molds made out of plastic or wood though.

An ACME thread is a typical square thread (not sharp or pointed like standard screw threads) and are used on machine tools (lathes, mills, etc.). Most all manual mills & lathes will utilize acme leadscrews, although I have seen a couple Bridgeport type mills that have ballscrews. No ACME screw company that I'm aware of. Wyle E Coyotee probably knows for sure though.
Ballscrews serve you much better on a CNC for many reasons, not the least of which is negligible backlash. This is an issue with my super-budget X2 CNC conversion. My steppers aren't powerful enough to allow me to ride the brake, so backlash becomes a much bigger issue. I'm just playing though, so I don't need ballscrews or bigger motors.... yet.

[cyclestart]

Originally Posted by blades Here's a turn-key converted X2 mill, but it's priced @ a rather hefty $3500 (not worth it, imho) http://www.syilamerica.com/product_X2.asp

There's 2 prices on that linked page. The $2395 price would be a more fair comparison to other desktop solutions.


cwm9
Here's a bit more reading material.
http://www.tormach.com/FAQs.html
Page down to their reasoning for choosing steppers. The meat is in the pdf document. Some of what's there is sales talk but there's some useful info as well.

Not suggesting a Tormach, you just sharing some info. Although a Tormach would be a lovely bit of overkill