[G02]

I currently build custom speaker cabinets and guitars and am looking to expand into CNC to cut down on work time and also be able to do intricate designs on my guitars, such as these: ESP Angel Sword. I only need 3 axes and a total working area of 2'x4'. Z axis really doesn't need to be more than a few inches. I have a max budget of $2000 (could extend to $2500 if really necessary), cheaper is better, though, and need a starting point. I've seen the various Joes 2006 builds and the Fine Line Automation machines, but I'm wondering which is the best bang for the buck. Accuracy is key here, as well as a metal construction so I can expand to milling soft(ish) metals like aluminum and copper in the future and have a coolant hose and collection tank below (or have the ability to be able to switch between plasma cutter and CNC on the same machine). I really have no idea where to start. Speed isn't really of importance so I'm not sure whether to use Nema 23 or 34 steppers.

I should mention, my budget of $2000 is including everything to get the machine running including the electronics and router (my current one is dying and needs an upgrade). I already have a spare computer and SolidWorks and Maya, so those don't need to be bought.

Any ideas on where to start? Thanks!

EDIT: I just realized I don't have a dust collector either, so I'll have to get one as well. Likely the 2HP one from Harbor Freight for $200. So max budget now is $1800, extendable to $2300 if really necessary. Thanks!

[ger21]

Any ideas on where to start? Thanks!

Double your budget. Seriously. You're not going to get a 2'x 4' metal machine for $2000 complete. Electronics and leadscrews will use up half of that.
It's actually tough to build an MDF machine for much under $2000. You also left out software, which will generally be at least $300-$500, if not more.
I would not recommend trying to build a combination router/mill/plasma cutter. Coolant and sawdust don't mix well, and neither do sparks and sawdust.
Best bang for the buck is probably fineline or CNC Router Parts (both use the same parts), but you'll need to increase your budget.

[G02]

From my lurking, it seems that many people are able to finish their builds for close to $2500, and those were 4x4 or 4x8. For instance, an FLA-200 can be had for $2550, but I've heard that it's much cheaper if I buy the 8020 myself and get thrifty with electronics. I don't have a plasma cutter yet, so it doesn't need to be ready for that, just upradeable. Mach3 is only $140, why do you suggest software gets up to $300 to $500? Doubling the budget isn't going to happen so I'd like to find a way to work within my means. There is a poll thread here somewhere that asked how much it cost people to build a Joes 2006, and the majority of the votes were at or under $1000, so I know it can be done for cheap, even though that machine is MDF.

If I build it myself, and perhaps only buy the hardware package from CNCRP (Or even design them myself, I guess), do you think I could do it for around $2000 or $2500 max? I can always wait an extra week or two to finish the machine and get a couple hundred more.

[ger21]

I didn't realize the complete kit was that cheap. But you're still at $2700 with Mach3, and you need a PC, and CAD/CAM software, plus a router to mount on it. And how much is shipping?
I'm not sure how complete "complete" really means, but it's pretty easy to spend $200-$300 on quick trips to Home Depot for small misc. parts.

There's a big difference between the Joes 2006 and an FLA machine. And keep in mind that since that poll was posted, prices have tripled on many items. For example, I paid $26 each for my 1/2-8 2 start acme screws. Today you can't get them for less than $75

And getting thrifty with electonics will usually result in buying them twice, which is actually more expensive. The Gecko G540 will have double (or better) the performance of anything cheaper. Imo, it's not worth it to go with cheaper electronics.

I don't know how much you can save by shopping around. The best thing to do is make a list, and start shopping for prices.

[jckstrthmghty]

The Joes2006 kit is around $500 but contains only the MDF and HDPE parts. The total mechanical build would cost around $1000. You can get cheaper electronics like the HobbyCNC kit but you sacrifice performance. That in mind plasma routing or milling coolant on a Joes2006 isn't going to work.

Sure the FLA-200 can be had for $2500 but isn't that over your budget? Still need a router, end mills, computer, software, dust collection as you mentioned and many odds and ends.

I'm sure whatever your start budget is you'll go over that by $500 by the time you make your first cut.

Common problem. Performance expectations vs budget.

[G02]

Ok. I've decided to build the whole thing myself since I have access to massive mills, routers, lathes, etc. at my school. I'm going to use the CRP plans and likely just make a 4x8 to future proof it. I can get metal quite cheap through school, including t-slot extrusions, so price to build the actual setup minus electronics should be quite cheap (around $700 if not less). I'm going to start ordering metals to fab the rack and pinion drives and other parts in the kit. Is 6061 aluminum ok or should I go with steel? I'm thinking aluminum would be better since it is much lighter and 6061 offers the same yield strength as the affordable A36 steel, each around 36ksi. Lighter weight means less weight for the motor to move which is better, right? I'll still make the gantry risers out of steel, unless aluminum would be ok for that too.

[PaulRowntree]

I'm thinking aluminum would be better since it is much lighter and 6061 offers the same yield strength as the affordable A36 steel, each around 36ksi. Lighter weight means less weight for the motor to move which is better, right? I'll still make the gantry risers out of steel, unless aluminum would be ok for that too.

I don't think that yield strength is the issue since there is no way you are going to break your machine frame. Think about deformation, which depends on the modulus. Steel's is 3x greater than Al's, and heavier as you say (but you need less of it for the same deformation characteristics). Al is much easier to work with, much more expensive. Can you weld or get it done for you?
If I remember correctly, the strength:weight ratio of steel and Al are very similar, so in my mind it comes down to what is available, at what cost, and the ease of working with it.
Cheers!

[G02]

Sounds good. Buying from online metals, steel and al are pretty close for the correct sizes, so I don't mind. And being that al is much easier to machine, I'll go with that. It's harder to weld, but shouldn't be too bad.

[louieatienza]

Ok. I've decided to build the whole thing myself since I have access to massive mills, routers, lathes, etc. at my school. I'm going to use the CRP plans and likely just make a 4x8 to future proof it. I can get metal quite cheap through school, including t-slot extrusions, so price to build the actual setup minus electronics should be quite cheap (around $700 if not less). I'm going to start ordering metals to fab the rack and pinion drives and other parts in the kit. Is 6061 aluminum ok or should I go with steel? I'm thinking aluminum would be better since it is much lighter and 6061 offers the same yield strength as the affordable A36 steel, each around 36ksi. Lighter weight means less weight for the motor to move which is better, right? I'll still make the gantry risers out of steel, unless aluminum would be ok for that too.

Lighter weight also makes the machine more susceptible to vibration, which will show up in your cuts. Usually lightweight is more important in machines that need high acceleration, like laser, plasma, pick-and-place machinery, etc. Although there is no wasy answer, I'd say use the heaviest material that you can that still allows your machine to travel at acceptible speeds...

[G02]

Lighter weight also makes the machine more susceptible to vibration, which will show up in your cuts. Usually lightweight is more important in machines that need high acceleration, like laser, plasma, pick-and-place machinery, etc. Although there is no wasy answer, I'd say use the heaviest material that you can that still allows your machine to travel at acceptible speeds...

If need be, I'll throw some lead weights on the thing, but I think it'll be quite heavy by itself.

Does anyone have any issues with vibration from their CRP4848/4896?

[louieatienza]

If need be, I'll throw some lead weights on the thing, but I think it'll be quite heavy by itself.

Does anyone have any issues with vibration from their CRP4848/4896?

Yes the extrusions do have some mass to them, but I'm sure the CRS 1/4" plates used for rails do contribute a great deal to the overall mass. I just mentioned this because you said previously about using aluminum because it is lighter.

I used 100200 profile on the gantry of my new machine build, and that 32" length alone weighs 40lbs!

[G02]

Yes the extrusions do have some mass to them, but I'm sure the CRS 1/4" plates used for rails do contribute a great deal to the overall mass. I just mentioned this because you said previously about using aluminum because it is lighter.

I used 100200 profile on the gantry of my new machine build, and that 32" length alone weighs 40lbs!

True that. However the Joes machine only uses MDF, which has a density of .7g/cm^3. 6061 Al has a density of 2.7g/cm^3. Comparatively, mild steel has a density of 7.85g/cm^3. If it comes down to it, I can always weld a big hunk of steel to the gantry risers to give them some extra weight.

On another note, in the spirit of saving money, would someone that has more experience than me check over this power supply and see if it would be a suitable replacement for the MeanWell that CNCRP sells? It outputs 1000W continuous at 48V with a peak output of 1200W, and short-time peak of 1600W. I buy power supplies from this guy when I build audio equipment and his service and the products is phenomenal despite him being in China. I've emailed him to see what he thinks, but at a price of about $150 incl. shipping, it sure as hell beats the MeanWell at $270 + shipping.

Connexelectronic A1000SMPS
PDF Spec Sheet

[ger21]

if it would be a suitable replacement for the MeanWell that CNCRP sells?

That power supply is for the Nema 34 motors. Going with Nema 34's is going to add about $700-$900 to your electronics cost over the Nema 23's. The electronics kit alone is $1500??
What happened to your budget?
If you are going that route, you might want to look at this one.
http://www.automationtechnologiesinc...c12a-duplicate

[G02]

That power supply is for the Nema 34 motors. Going with Nema 34's is going to add about $700-$900 to your electronics cost over the Nema 23's. The electronics kit alone is $1500??
What happened to your budget?
If you are going that route, you might want to look at this one.
High-Torque Stepper Motor, Stepper Motor, Driver, Stepper Motor kit, DC Servo Motor, DC Servo Motor kit, Stepper Motor Power Supply, CNC Router, Spindle, and other Components. Automation Technology Inc

I'm going to make the final decision on Nema 34 or 23 when I have a pricelist made, but I should be able to do a 34 system for not too much more than $2000 if I fab everything myself instead of buying parts from CNCRP. I should clarify: My budget can definitely go higher than $2000. However, I have a number of other things I need to buy as well and spending $4k on a router would make me have to wait a fair while to buy those. I can wait on some things, but building router doesn't help me if I can't upgrade the rest of my shop to take advantage of it. I'm also very limited on space so building a router that can replace other machines is hugely helpful. That's why I'm upping the build size from 2x4 to 4x8. I could then do an entire 4x8 sheet of MDF and have no need for any other saws, which I currently have to use (they're also not owned by me).

Regarding power supplies, how does the voltage that is output from the PSU get transferred to the motors. The Gecko drivers, 201x or 203v, for the Nema 34's allow betweeen 18V and 80V, yet the steppers only use 1.55V. Does the voltage of the PSU matter, or does it just need to be between 18V and 80V? Also, if the motors only draw 7A, why does one need a 1000W PSU?

[ger21]

Read this.
Support

And personally, I have a feeling that you'll spend at least $1000 more than you think you will.

[lgalla]

hay Gerry you had a typo error post #2
I would recommend trying to build a combination router/mill/plasma cutter. Cooland

[ger21]

Thanks. That should have been NOT recommend, and coolant.

[G02]

Read this.
Support

And personally, I have a feeling that you'll spend at least $1000 more than you think you will.

Ah, thank you. I feel much less stupid about motors now.

As for budget, you're probably right. However, I can at least try and if I end up spending more, so be it. I'd just like to try as hard as possible to keep it around that budget. I have other tools I need to buy as well, unless I decide to outsource, which I'd like to avoid doing. I'll just start fabricating parts on for as cheap as I can and see how far I can get. Then I'll make a choice about motors and such (yes I realize nema 23 and 34 mounts are different, so mounts will be the last part I make, but I'll likely just shell out for 34's to prevent any regret and future proof it).

[adprinter]

Well, after reading all these posts, I just thought that I'd throw in my 2 cents worth. The prices of things have SHOCKED me, the amount they have gone up in just the last 2 years. I did some research, posing the question: "What would it cost, at today's prices to duplicate my machine?" Which I originally built complete, for less than $2000. At today's prices for the same list of JUST THE MATERIALS it would cost over $3500 to build! Add to that price for things like: wiring, energy chain, limit switches, motor mounts, couplings, nuts, bolts, and CAD/CAM software. I am thankful that I built my machine WHEN I did, for if I had waited until today- it probably would never happen! However, it would STILL be cheaper to build, than to just buy one. Although the dimensions are not included, here is what I initially built- (See the attached file MyCNCV1.PDF).

[G02]

Well, after reading all these posts, I just thought that I'd throw in my 2 cents worth. The prices of things have SHOCKED me, the amount they have gone up in just the last 2 years. I did some research, posing the question: "What would it cost, at today's prices to duplicate my machine?" Which I originally built complete, for less than $2000. At today's prices for the same list of JUST THE MATERIALS it would cost over $3500 to build! Add to that price for things like: wiring, energy chain, limit switches, motor mounts, couplings, nuts, bolts, and CAD/CAM software. I am thankful that I built my machine WHEN I did, for if I had waited until today- it probably would never happen! However, it would STILL be cheaper to build, than to just buy one. Although the dimensions are not included, here is what I initially built- (See the attached file MyCNCV1.PDF).

Dang.... That's slightly annoying. I'm working on my BOM so hopefully I'll have a better idea of cost once that's done. Is there any good substitute for 8020, or is it just that much better than anything else? I'll be assembling this over the summer with my uncle, who has a very (and I mean VERY) powerful welder, among other useful tools like a mill and lathe. Looking at onlinemetals, A500 (mild steel) rectangular tube in 1.5"x4" is only $38 for 5ft (I'm scaling back to 4'x4' since I don't really need 4'x8'). It's also roughly twice as strong as 6061 Al and 3 times the weight, both good for CNC frames. This is also less than half the cost of t-slots and also saves me money on nuts, bolts, and time spent tapping holes. I'm not planning on ever taking this apart so I don't need the ability to break it down like with t-slots. Thoughts?

I'm also wondering, what is the strongest frame design? I thought it was x braces to prevent racking, but 8020 obviously can't do that. If I were to weld it myself instead of using 8020, should I plan on making x braces instead of perpendicular slats? X braces should theoretically be stronger and use less material.