[muibubbles]

Disclaimer: I have wanted to make/buy a cnc machine for over two years, meanwhile I have done very little research. I am going to classify this as a NOOBIE BUILD, so If I can do it, Anyone can.

Well I'll start out with, I just saw the quote "bad decisions make great stories" in someone's sig on the forum, and I thought how fitting it is, hence the title haha.

About me: My name is Ben and I am 23 years old, Graduated NJIT with an Industrial Design degree, and started my own company in June called Bubble Tech. Currently we only sell T-shirts and Decals (that i design and silkscreen/cut). But my true passion is fabrication, we are transitioning into making parts (specifically for 1993+ Mazda RX-7's) and will Primarily focusing on Metal and Composite Fabrication.

I also have a huge passion/obsession/addiction to tools. I will buy and hoard any tool i can to make the job easier (if i have the money) hense the cnc build lol. (I have collected a lot of various tools since high school (presents/garage sales/estate sales), so i have a variety at my disposal, so hopefully it wont be too hard to make.

I have a few products in prototyping stages, and a cnc for final production would be ideal. accurate and efficient. And quiet frankly, Im sick of milling them by hand lol.

Here is a picture of the part I am making. This is the reason why I am building this machine. This is 1" aluminum.
old 3/4" Version and New 1" version



Just to get an Idea how/where it is mounted to. Test fitted the 3/4" version. Fits like a glove!


SO, last night I ordered solsylva plans and i'm looking to do a 25x25, frankly because I can not fit anymore tools in my garage, and I'd LIKE to keep this in the basement (is this a bad idea?) I briefly looked over the plans and omg. is it overwhelming how much info is in the package David sends you... i really dnt know what the capabilities are but i already feel like i need to upgrade out the box. (i dont feel like doing it twice)

Soooo, here is my basic plan.

Weld up 1x2 steel frame for the bed.
acme screws
C-channel aluminum gantry
everything else 1/2" aluminum.
bosch full size router

Electronics looking to get this from
CNC Stepper Motor Kit: , Stepper Motor, Servo Motor, Power Supply Automation Technology Inc

G540 3-Axis kit (115V /230VAC):
1 pcs G540 4 axis driver with 4pcs DB9 connectors
3 pcs NEMA23 KL23H2100-35-4B(1/4” Dual shaft with a flat) 381 oz-in
1 pcs KL-350-48 48V/7.3A 115V /230VAC power supply

OR
piece together? from gecko directly

1- Geckodriver g203v
3- stepper motor g723-400-4
Current Rating: 5A
Inductance Rating: 2.6mH
NEMA Size: 23
Holding Torque: 400 oz.-in.

opinions?

Should be fun, I intend to make tons of mistakes and waste a bunch of money, but honestly, I think that's how you learn the best. I am to research more into the motors/drivers and hopefully order them before friday. I will be going to my metal supplier Monday so I will start building Tuesday.(group)

id like to have this done in 2 weeks so i can start my production. hoping to spend no more then 1000-1200 but REALLY hoping its less. i have a bunch of hardware at home, router and pc so hopefully it wont be too much more

[ger21]

opinions?

The Solsylva is an entry level machine designed for light wood routing.
If you want to cut aluminum, at the very minimum, I'd upgrade the linear motion bearings to these. Glacern Machine Tools - Linear Rails and Bearings
I'd also try substantially increase the rigidity of the gantry.

Also note that a Gecko G203V will only run ONE stepper, not 3.

[muibubbles]

hmm, thanks for the heads up.. when I called Gecko the lady on the phone immediately told me that one when i told her what i was doing.

i've seen a few videos (such as cranes) milling aluminum with this kit, and if it upgrade rigidity by replacing the wood parts with Aluminum, wouldnt that be enough since people are milling aluminum with a cnc made out of wood? I'm going to try and upgrade as much as I can without breaking the bank... after i get it rolling ill upgrade accordingly!

[ger21]

i've seen a few videos (such as cranes) milling aluminum with this kit, and if it upgrade rigidity by replacing the wood parts with Aluminum, wouldnt that be enough since people are milling aluminum with a cnc made out of wood?

Depends on how long you want to wait for your parts. The more rigid the machine is, the faster you can cut.

I see on their website they recommend .02" per pass at 25ipm.

[louieatienza]

hmm, thanks for the heads up.. when I called Gecko the lady on the phone immediately told me that one when i told her what i was doing.

i've seen a few videos (such as cranes) milling aluminum with this kit, and if it upgrade rigidity by replacing the wood parts with Aluminum, wouldnt that be enough since people are milling aluminum with a cnc made out of wood? I'm going to try and upgrade as much as I can without breaking the bank... after i get it rolling ill upgrade accordingly!

I built a solsylva also and had cut auminum with it, but it is far from "milling" in that sense. Router spindles don't have the ridgidity of mill spindles, and milling machines are usually made with cast iron and steel for added mass and ridgidity. You won't be taking the same depths of cuts as a mill, nor speeds... just something to keep in mind.

As Gerry states, you'll need three G203Vs, and a power supply for your steppers (probably 60-72v.) You might want to consider a breakout board from PMDX to make wiring easier. I don't think even aluminum C-channel would be enough for a sturdy gantry beam. I would use at least aluminum extrusion.

I use a Bosch router as well, and cut aluminum and G10/FR4 with it, but there are special bits and cutters used for the high rpm's of wood routers. I'd agree with Gerry that you'd want at least round supported linear rails and block; the gas pipe/skate bearing configuration is barely capable of wood cutting let alone aluminum; at least for what you need.

David Steele's instructions are very concise, and everything makes sense and is easy to follow if you go step by step. But you'll end up modifying every aspect to where you end up going astray completely from the plans. I built mine per the book, and used it for a couple years before completely redesigning it.

If you already have a mill, you might want to consider making parts to CNC it...

[Walky]

I'd absolutely NOT use a wood machine for production aluminum cutting because:

1) Finish won't be as good, hand finishing is time consuming.
2) Tolerances will not be very good.
3) The machine will suffer.
4) The router will suffer.
5) The bits will suffer.
6) It will be slow as hell.

You might be able to get away with 1 and 2 depending on what you expect from your parts, though.

[louieatienza]

Depends on how long you want to wait for your parts. The more rigid the machine is, the faster you can cut.

I see on their website they recommend .02" per pass at 25ipm.

That's onliy 50 passes for 1" stock...

Without totally beating down your ideas, Ben, What I have done to overcome flex in my machine when cutting aluminum is to largely ignore it as long as the tool is cutting, since I always do a finish pass to bring a part to tolerance. Again, the finish can be good but nowhere near what a proper mill can achieve, even a cheap Chinese one properly tuned. You then use the machine to make parts for itself, which is waht I've done as well as others. But if you want to get up on production you may find that this eats up a ton of time. And the more your machine cuts parts for itself rather than clients the costlier it gets. So try to use the best parts you can at first, and get spome plans and drawings together. Dead ends are very costly, as well as piles of parts that you can't use.

[muibubbles]

hmmm whataaaa buzz kill lol. so you dont think This machine will be capable of cutting out the piece i showed in the picture? I guess i dont properly understand the term "milling" then... essentially I want to cut that spacer out of 1" thick aluminum plate. I have no idea what the recommend .02" per pass at 25ipm. translates to in real time lol, how long would you estimate that this would take? right now it takes me 4-6 hours to do it by hand....... and its a PITA.

Whats the main difference between wood and aluminum milling machines? I thought the aluminum ones were just more rigid..

also when i say "production" i mean parts i will be making and selling, not clients who want xxx parts by xyz. it will be mainly for my company use if that changes anything lol.

[Devastator]

I built a 24X48 Solsylva. It is highly modified and I doubt that it could make parts like that at least at any reasonable speed. You would probably have to use steel construction and lose the pipe/skate bearings for some linear rails. Mine uses wood reinforced with steel angle and all of the wood has been upsized by a large margin. The biggest weakness I found was in the design of the hand bent aluminum bar that holds the bottom pipe bearings for the gantry. I could rip the gantry right off of the pipes with my bare hands. That will give you problems unless addressed.

I started building it to the plans, then had to modify only to find it still inadequate, then refine it a third time until I was happy. So, almost every part I made three times. I ended up with something that sort of resembles a Solsylva.

You will see once you start building, a need for improvements. You may end up with a machine that will do what you want, but a stock Solsylve ain't it. It is a good guide that can help you set started in the right direction though.

Best of luck!

[muibubbles]

Ok, so a few of you got me worried but I think it will be okay. let me clarify, By production, I dont mean I do cnc work. I am going to be designing and making parts for a niche car market. if it takes a few hours to cut one of my aluminum spacers, I am perfectly okay with that. right now it takes me 4-6 hours each, by hand. With that said, the build will go on!

I've dug deeper and looked over more designs and builds. Mechanically, I think I have the concept down 90%. I think I will be ditching the Solsylva plans and be designing my which will be a hybrid of designs i've seen.

I saw this video on youtube and REALLY like the design. I will be adapting this guys method of "bearings and rails" really neat, cost efficient, and looks like it will be very rigid and no play.

Now this may be a stupid question, but would it be beneficial to make the gantry out of steel? I know it will weight more but it will be very rigid. im thinking 1/2" square bar and weld sheet metal in between (and dimple die them to shed weight/strengthen it)

Besides the rails, what are my other weak points? i guess thats the benefit of building a cheap cnc the first time, you know what needs the most attention and $$. :/

[muibubbles]

devastator thanks for the reply! once again, speed isnt an issue, but if i can upgrade something now to speed it up, I would love to instead of buying and then replacing things wasting money.

I think my linear axis is good as per the design in the youtube video. It utilizes a bearing on the top, bottom, and side of a steel plate. IMO, rigid, cheap, and will be hard to move. perfect in my book. at least for now. i am going to look into pricing. if the price is close to the Glacern Machine Tools - Linear Rails and Bearings i will use those for the x axis and use the other design for the y and z.

If i do decide to make the gantry out of steel, will I need large stepper motors then the 381oz in motors that i previously posted? also since im running 3 motors, should i get one larger one since the x axis motor is spinning two threads?

[louieatienza]

My heavily modified "solsylva" in its second iteration featured surplus linear rails on eBay, phenolic construction instead of wood. And it still could have been more ridgid. As to most home made linear bearing systems, unless you have the bearing surfaces ground to a high finish and hardened, you'll never have the smoothness of movement, tight tolerances, and precise motion of even the least expensive linear rails like Glacern's. And this will show up in the finish of your surfaces as chatter.

Your stepper choice is going to largely be determined not just by the weight of your gantry but also the lead of the drive screw you eventually will use. With a finer lead, you could use a smaller stepper, while a coarser lead would normally require a larger stepper.

The reason for the solsylva bearing design is that there is an amount of flex allowed. This is to take up any discrepancies in pipe thickness or installation error. This is not a problem with commercial systems because they are made to so much tighter tolerances to begin with.

As to steel - sure it would be beneficial to make the parts out of it, and you seem to have the machinery to true all all the mating surfaces. In my "phase 2" machine, I had one 425 in-oz stepper running two leadscrews via timing belt, as in the original Solsylva. with 1/2"-8, 8 start leadscrews (1tpi) with my gantry weighing in over 80lbs, and was able to rapid at over 300ipm with a xylotex box and 24v psu, and over 600ipm with a G540 drive and 48v psu.

My threads on the Solsylva build and the upgrades are here

http://www.cnczone.com/forums/diy-cn...lva_plans.html
http://www.cnczone.com/forums/diy-cn...d_another.html

And my YouTube channel here. You can see how my machine "evolved" from a basically by-the-book Solsylva build to the 8020 aluminum extrusion frame/phenolic plate of my current machine.

AtienzaLouie's channel - YouTube

I noticed you mentioned graduating NJIT... I live in the suburbs of NYC, and currently fabricate some automotive parts for a fellow in Connecticut. I cut quite a bit of Garolite G10 and G10/FR4 in 3/8" and 1/2" thicknesses, as well as 7000 series aluminum in those thicknesses.

[kinghong1970]

Louie,

you've cut alum on your modified machines... what are some of the DoC and feed rates you've used?

[Devastator]

devastator thanks for the reply! once again, speed isnt an issue,

You say that now, but that will change as soon as you start making chips. We all strive to build a machine that will do what we need on the first try, but that rarely happens. Just make it as rigid as you can and plan for the machine to make its own upgrades when time and money permit. Don't get me wrong, I love the solsylva plans. I wouldn't have gotten started on my build if I didn't have them, but many people plan to machine things on them that they just are not designed for. Like buying a Volkswagen and expecting it to work like a Ferrari. Best of luck to you!

[muibubbles]

Louie, thanks for the links, i've glanced at some videos and looked over your first build.. When I have more time I want to go through your whole upgrade/second build thread =)

Devistator- aint that the truth. im trying go bang out everything in one shot. looks like my build budget is raising 30% =(

A lot of you stressed on the rails and flex. so I've looked into it more and went to my handy friend mr ebay. I found a company in china that makes a bunch of parts that might just make my life easier.

What do you guys think?
linear rails
6 Linear Bearing Slids SBR12 200 400 660 12 SBR12UU | eBay

anti-lash ball screws (i believe the leadnuts are aluminum so i dont have to worry about the lead nut from cncdump flexing)
3 Anti Backlash ballscrew RM1605 200 650 1000mm C7 CNC | eBay

Thoughts? Also I'm considering to do 4 motor machine instead... this way I can mount the motors directly on the shaft for the easiest of construction..

The prices on ebay dont seem bad, i can swallow the cost since everyone says they are the weak point and well, it actually will decrease my fabrication time as well so I can get up and running quicker. I really like where my "ideal plan" is at right now:

Steel construction (1/4 & 1/2" square tubing)
linear rails (ebay)
ball screws (ebay)
Gecko g540
4? 381in oz stepper motors

Thoughts? In my head it's going to be a bad ass machine, but I'd like your input! I appreciate it so much, you guys have been SUPER helpful. I feel very confident in building the cnc machine now, once I get some confirmation and double check, Id like to start placing some orders by this weekend. I see my bank account slowly diminishing lol. But you gotta pay to play right? hopefully this will help me turn some profit and be a great investment!

[harryn]

There are different approaches. The most expensive parts to have made are the first 5 or so, as these are often protos and changes happen fast. For this, it makes a lot of sense to have "in house" proto parts capability.

It is handy to have the proto parts of the same material as the final version, but it is not always necessary. For instance, Delrin and PVC can be very useful for proto work / mechanical sizing.

The next 100 parts are cheap, as any decent, local machine shop can make them 10x cheaper than you can do at home. There is a reason that people invest $ 100K plus into professional cnc mills, they are just crazy fast and efficient. Pro machinist buy metal in bulk, so their cost is lower, and they know tricks that save all kinds of time.

If you already have a manual mill (not the baby table tops, but a full size one), consider to just add cnc control to it, sufficient to make your protos and sub 10 unit runs.

Save your energy and innovative spirit for doing product development, and job out the production runs to people who specialize in that area.

Imagine that instead of getting orders for 10 parts / month you get an order for 1 000 parts. If you are used to making them at home, you are really in trouble. If you are used to jobbing them out to a machine shop with a big cnc mill, they just run it a few extra hours while you enjoy being with your significant other.

[harryn]

BTW, while 2 weeks and $ 1K might be possible ( I don't know), I think 2 months and $ 3 K is more realistic for a build.

[muibubbles]

Harryn, maybe milling was not the correct term. i do not have a mill, and just to clarify by PRODUCTION, I do not mean your typical machine shop stuff. I am making parts for a NICHE market, i am perfectly fine if i can only make 1 or 2 a day. I really just dont like outsourcing things. I enjoy making things myself, even if it is the hard way, theres more of a satisfaction when i can say i designed AND built something. Maybe i will look into a cnc bridge port then... the only thing is im VERY tight on space in the garage and i was going to run this in my basement. i think for a used bridgeport and cnc will run the same amount so ill check it out. by all means i do not need anything with .000001 tolerances, I am not trying to run a machine shop (that is completely out of my league) and i dont make anything that will require it.

also I am self employed, I have plenty of time lol. without waiting around for parts, i think i can have it up and running in a short time (thanks to this forum)

[Walky]

What do you guys think?
linear rails
6 Linear Bearing Slids SBR12 200 400 660 12 SBR12UU | eBay

I would go with SBR16 or SBR20, they will be more precise and can take higher loads. SBR12 doesn't have such a good fit on the rail and has less ball circuits.

anti-lash ball screws (i believe the leadnuts are aluminum so i dont have to worry about the lead nut from cncdump flexing)
3 Anti Backlash ballscrew RM1605 200 650 1000mm C7 CNC | eBay

The ballnuts are all steel with plastic seals and ball returns.

The dumpsterCNC nuts are great stuff and are made of a very strong plastic, there's not much to worry about them really.

It would be a very good idea to design the full machine before spending any money so you can get feedback on it, and it allows you to buy exactly what you need (most chinese Ebay sellers will sell custom rail sizes, and some are also able to custom machine the ballscrews by request).

Planning and building a good machine can takes months, and that's usually assuming there's also been some months of previous research on the subject. It's better to make a lot of designs and discard most of them than spend money on a handful of parts of which half will go unused at the end, while trying to fix them to work correctly. Buying beforehand is one of the most common mistakes and one of reasons some machines end up not being as good as they can. Every single detail on a CNC is important, the design, the parts, how it's assembled, what will it be used for... miss one detail and you could be negating any possible benefit from the rest.

The machining time as compared with a metalworking machine is not to be taken lightly. I'll just guess numbers here, but if we compare, say, using a $1500 - $2000 homemade machine to make 1 part in 4 hours (plus hand finishing), and using a $2500 purchased machine to make 2 of the same parts in one hour, with less hand finishing, better tolerances and most likely better tool life (less vibration), the extra machine cost seems well worth it for most situations. Even the power bill will be lower than having a wood router cutting aluminum all day long (assuming it doesn't burns out and you need to get a new one).

[louieatienza]

Louie, thanks for the links, i've glanced at some videos and looked over your first build.. When I have more time I want to go through your whole upgrade/second build thread =)

Devistator- aint that the truth. im trying go bang out everything in one shot. looks like my build budget is raising 30% =(

A lot of you stressed on the rails and flex. so I've looked into it more and went to my handy friend mr ebay. I found a company in china that makes a bunch of parts that might just make my life easier.

What do you guys think?
linear rails
6 Linear Bearing Slids SBR12 200 400 660 12 SBR12UU | eBay

anti-lash ball screws (i believe the leadnuts are aluminum so i dont have to worry about the lead nut from cncdump flexing)
3 Anti Backlash ballscrew RM1605 200 650 1000mm C7 CNC | eBay

Thoughts? Also I'm considering to do 4 motor machine instead... this way I can mount the motors directly on the shaft for the easiest of construction..

The prices on ebay dont seem bad, i can swallow the cost since everyone says they are the weak point and well, it actually will decrease my fabrication time as well so I can get up and running quicker. I really like where my "ideal plan" is at right now:

Steel construction (1/4 & 1/2" square tubing)
linear rails (ebay)
ball screws (ebay)
Gecko g540
4? 381in oz stepper motors

Thoughts? In my head it's going to be a bad ass machine, but I'd like your input! I appreciate it so much, you guys have been SUPER helpful. I feel very confident in building the cnc machine now, once I get some confirmation and double check, Id like to start placing some orders by this weekend. I see my bank account slowly diminishing lol. But you gotta pay to play right? hopefully this will help me turn some profit and be a great investment!

I think 2 motors on the gantry axis is probably better than either driving both screws with one belt via pulleys or a center driven leadscrew. It also frees up the center of the machine for potentially larger parts or even adding a plasma torch.

I would hoewver come up with a comprehensive set of drawings or CAD so that you'll know exactly what sizes of screws and rails are needed. The eBay guys will cut the rails and machine the screw ends to your specs.

I also don't think it's the tolerances that may cause you fits but the overall finish on the parts. Like I mentioned earlier there are some ways around it but it will take a bit of time and experimentation and some broken bits. These routers do not have bearings that can handle heavy thrust loads, and the bearing housings are usually plastic or metal with a plastic sleeve. So you may be fine for a while but you'll probably wear a router out and have to replace it regularly; unless you decide to rebuild them.

Believe it or not, speed IS an issue because routers run at higher RPMs. This means that to cut aluminum and keep the chipload correct for your machine, you will need some speed. Since most routers are optimized to run best at higher RPMs a careful blend of bit selection, feed and speed is important. You just can't say that you'll mill .02" at a time and you don't care how long it takes. Your machine will care, your bits will dull faster on the bottom than the flanks which are supposed to be doing most of the cutting, and you'll burn your router up faster. Being you already have a mill, you could rough out the parts manually and fdo the finish cuts on the router. Having the clearance for chip evacuation is very helpful. I'd even design in flood or mist coolant and air blast if you are doing aluminum mostly.

kinghong, I put some parameters up in Brian's (Grumblau's) build thread, at the sake of not repeating the information again...

Again, consider adding CNC control to the mill you already have; it will perform much better than anything you can make with steel tube and round linear rails.