Hello everyone, I was referred to this website from the Flite Test forums when a model aircraft project evolved into researching a scratch build for a CNC mill/dremel. I've got a rough idea of what I'm looking for in the mechanicals but am running into some problems with the electrical and software end. I currently own an arduino uno rev3 and was hoping to use a gshield to cut costs but I noticed that most of the nema 23 steppers I had planned on using had a max 2.8 amps per phase and the gshield was only rated for up to 2.5 amps max with cooling, not mention it seems to have come into short supply recently. I was hoping someone here would be able to provide some recommendations or advice. I have heard of several other alternatives but am unsure of how to rate them or what I should be looking for in each. Up to now my degree has exposed me to CAD/CAM work that has been mostly automated drag, drop, and cut. This is my first real endeavor to get into the "nitty gritty" of the systems so any help would be much appreciated. For anyone who would want some backstory of this up to now, here is the original post I started on flite test. http://forum.flitetest.com/showthread.php?20441-CNC
Hey Capn, Shoot for low inductance motors. A TinyG will handle the 2.8A motors. Give it a 24v supply and little fan to make it happy. Has your degree exposed you to basic machining and manual cnc programming? If not, I recommend a used copy of "Computer Numerical Control" By Robert Quesada. It covers the two really well in an easy to read and logical order. Its also a great reference to have around. Welcome to Openbuilds. Joe
Yeah during my degree I did some project with both manual lathes/mills, manual g-code creation, and utilizing surfcam with solidworks models. I've been using my old notes to look over some generated practice codes. As for the tinyG it seemed a tad on the high end of my budget. Is there much difference between the Gshield and the TinyG besides the extra axis? I came across this the other day, http://www.sainsmart.com/new-4-axis...ver-controller-board-kit-57-two-phase-3a.html , it looks like it has the option for either serial or usb interfacing. I also saw this, http://www.sainsmart.com/arduino/co...controller-card-interface-breakout-board.html , but thought that it would need a separate purchase of driver board(s)
You'll note that not only was the original OX build was done with a gshield, it had 2 nema 23s wired into a single driver. I would suggest checking with @Mark Carew to see how that played out. As it appears you are only wanting to build a moderately low power machine, this approach may be fully workable. Another project you might look at is @Robert Hummel's Stinger design which only uses nema 17s. This design should have no problem pushing a dremel through balsa or foam.
The first looks like it'll work to get you moving. I'm skeptical of some info given on the data sheet though. research tb6600 and tb6560. the tb6560 also does not have a quarter step setting which you may or may not need. The second will need drivers as you said. If you can use a parallel port over usb then do it. I think that the first package is usb for power only, but double check that. Download mach3 free, set it up, and throw code at it. I wouldn't run two motors off of one gshield driver. Some will get away with it because they're not pulling much to begin with. Two 17's maybe. Two 23's nah.
The design I had planned so far was going to be fairly similar to the OX model but there was some concern on the Flite test forum about having 2 nema 23 motors for the single x direction port on the gshield. I was planning on just getting something together that would be functional for some woodwork using a dremel 4000 but had wanted to eventually upgrade the spindle into something that could handle some small aluminum work. I had planned on a max build area of 3'x3' and was against the idea of a moving work bed since it wasted so much space. Would nema 17's be capable of aluminum work, or could the feeds be slow enough that it could be worked around? Another concern that I posted on the other forum was that there may be some stretch with belts under a tool force against aluminum. A bit peeved that my degree showed me how to operate these types of machines but hardly touched on how they actually worked.
I do like to try out different boards from time to time to keep up with the latest in drivers. The gshield works out good for me on the OX although I did learn that it seems to work even better with a good cooling fan on it. Keeping it cool seems to insure steady operation. I also like the xPRO as an all in one plug and play solution. I am partial to the Open Source alternatives out there though There are a lot of options out there to choose from and whatever you decide on we would love to hear the outcome. As for help we all will do our best to help get you up and running with whichever controller you go with. Mark
Woo that mean a lot, thanks guys. Watched all your videos before I started doing some real research into what I wanted to get done. I got my degree with focuses in mechanical and manufacturing so I don't think I'll have too much trouble figuring out feeds and speeds, or the frame work and rails; had more than a year of classwork in just those. This here is the toughest part for me, the electronics. For now I think I'll stick with a nema 23 set up and keep an eye out for all in one boards or complete controller/driver kits for sub $100. Have you encountered any stretch issue with your belts on the OX or having to re-tension them regularly?
I have my OX from the start and have only needed to adjust the belt once and in fact the reason for the need for adjustment was not the belt but a loose pulley that needed some thread locker on the set screws to keep them in place. You will be fine and you will have fun doing it. Once your done you will also know the machine inside and out for any future upgrades or idea changes/mods you may have in mind. Mark
Got it at Arizona State, still looking for degree related work which is a secondary reason why i'm doing this. I figure "designed and built a CNC machine" will look pretty good on my resume and I'll get my own CNC mill out of it. At this point it's really more "Researched, spec'ed, and built a CNC machine based on existing opensource concepts and ideas." That xPro all in one looks pretty slick, would save me some cash with the power supply, I think I have an old computer supply around here somewhere. Any idea on what kind of wattage would be needed?
at this point I'm pretty settled on the xPro V2, After looking at their wiki I am a little confused on weather or not I would need a fuse if I was using a computer power supply. I am still unsure of what kind of wattage this supply would need.
I took mechanical and manufacturing. AC/DC Electricity and digital electronics were prerequisites for me as well as manual and CAM programming. Don't short change yourself. You have the capacity to figure these electronics out. They're not exactly industrial PLCs. You just have to read, read, read, and ask questions. Once you take the dive into the hardware you'll have it figured out in no time. Personally, I'm a fan of the parallel port breakout boards. The grbl controllers are decent ai1 packages that appear to make life easy, but imo are really still in their early days. I had a TinyG and while I really loved that board we had to part ways for the time being. I just don't have the time to deal with the intricacies of grbl and the programs involved. So I went with a parallel port solution as they're pretty tried and true. I recommend (Not on behalf of Openbuilds just me) the TinyG over the Xpro because the drivers can handle .5A more which puts one in the ballpark for the common 2.8A stepper motors out there. TinyG also has pinouts for upgrading to different drivers latter on. I would fan and sink either of these. The cost is so cheap for the added protection. Regardless of what electronics you choose, I think what Mark was getting at is that planning, building, and operating your own piece of cnc machinery is quite the personal journey. The forum is filled with members on that same expedition and with members passionate about helping others along on their voyage. On that note, if money is a concern, figure out what you want your OX to do in the end and plan for that now. A system economically optimized for balsa and foam with a lightweight spindle can be very different mechanically and electronically than one optimized to 2.5d carve hardwoods and aluminum. Building the former and upgrading to the latter can be considerably more expensive (subjective) than building the latter to begin with. Good luck Capn. Joe
Well after reading another thread comparing the Ox and the C-beam I'm leaning towards the c-beam for accuracy. I do plan on eventually getting into some aluminum work. So I'm trying to design a 3 stepper C-beam set up. I thought that the specs for the max driver amperage for both the TinyG and xPro was 2.5 a? I was leaning towards the xPro cause I have an extra computer power supply laying around. As for the computations I assumed that the wattage requirement for the power supply would be a summation of P = I*V, mostly coming from the stepper power draw which I am assuming is the peak amperage times the number of phases multiplied by the input voltage? Then there's the power draw from the board itself, but I was under the impression that would be pretty minute. A full sized CPU draws up to 95 watts, I can't imagine a small control board even getting close if it doesn't require a heat sink. My degree was actually a general engineering degree, not even a split major. It covered a little of everything and about a semester extra in what I wanted to focus in. Really regretting not just going all the way with a dedicated mechanical degree. I did get some education on AC/DC but it didn't go much past the basics; charge, power, parallel/series, etc. I'm willing to do the foot and number work but I'm still pretty vague on what that translates to and back in terms of applications and components.
Hey Capn, About a year back I read a very long detailed post over on cnczone explaining why the combination of hardware and programming on the TinyG was superior to the Xpro, but since I cannot find that post now Ill retract my argument on current. It had something to do with the way the drivers sink to the boards pads and the way the jerk algorithm ran the motors. Either way it's a moot point to begin with. If a stepper isn't getting full current it will still run just not at full torque. A formula for stepper voltage I picked up from Gecko is: V=30*(sqrtL). L=motor inductance. However, it isn't really useful if you know you're going to be running a 12 or 24v system because of electronics limitations. For instance, I use a higher inductance stepper at 6.8L. (sqrt6.8)32=83v!! This tells me these motors would love an 80vdc driver, but that may not fall into everyone's price range. Geckos g201 80v drivers are $120 each. PS get pricey too. Another handy formula is Ips=(sumA).66. Ips=Current power supply. sumA=Sum of all motor current ratings. If I have 4 2.8A motors this formula tells me that my power supply should be rated at least 7.4A to have enough juice. Figure out what voltage your atx supply is, then use the power rating to figure out if it'll provide enough current for your set up. Don't worry about voltage, but worry about voltage! If you use the voltage formula above and end up with a high number too..don't fret. Drivers limit current to the motors. So you can supply whatever voltage to them and not have them burn up. The thing about voltage is that the higher the voltage you have the better the performance you'll see. Many using a 12v supply with an Xpro report laggy and sluggish motors before switching to 24v. So, while the Xpro seems appealing because of the atx capability, you may end up having to purchase a power supply to get max performance out of your setup anyways. Gecko has a ton of useful reading material on this topic.https://www.geckodrive.com/gecko/images/cms_files/Step Motor Basics Guide.pdf I really recommend one determines what they expect out of a machine prior to purchasing electronics for that machine. Just kinda makes sense right? I'd hate to see anyone buy a full nema17 and grbl board setup before building a machine (even virtually) and expecting to mill aluminum....happens I guess Joe
Thanks for the info Joe. I'm fairly new to the "hobby" cnc scene so I'm still getting my bearings on where to get my info and supplies. Looks like I'll have to figure out my power supply situation before i settle on a control board. I'll need to look into the tinyG a bit more aswell. Any other top choices in this category I should know about? I keep seeing mach3 pop up but I don't have a serial port and i haven't heard too much sucess about serial to usb adapters.
The UC100, plus a cheap breakout board, and cheap motor drivers will get you there. The UC100 is basically a usb to parallel port cord. I use an Ethernet to parallel port motion controller card. I wanted to avoid any potential problems with usb interference and pass off higher order processing, but that cost me. I could have just used an old pc (I have a couple around) and a bob, but the controller I chose gives me three parallel ports(more inputs/outputs for things like limits and mpg controllers). A UC100 will do usb to parallel conversion for one port, but the pc will still be processing everything. The UC300 puts out five ports and does higher order processing which is an excessive number of ports for one machine imo. It's important to note that many people use usb to parallel port devices like the UC100 and old pcs with printer ports with hardly any problems. Adding, less than $100 pc's are readily avail and one can download a copy of mach3 to play around with. It's actually free to use up to 500 lines of code with mild limitations (something to consider if you're doing simple 2d work). There are many options out there. It's a good time for diy cnc. Keep us updated with whatever you choose. Don't forget to start a build thread to keep us updated and check out others builds while you're at it. There is a lot of great and inspiring ideas floating around here. Enjoy Joe
Ok from what I understand Mach3 removes the processing element from the control board and has the connected computer do the grunt work and the control board becomes more of a router/power distributor, correct? So a control board that would use Grbl sender or univseral gcode would not be compatible with Mach3? I had heard somewhere that mach3 isn't compatible with OS's past windows xp, what about simulated OS's, like xp mode in windows 7? If i went with this route I would probably get a pci slot parallel port and sticking it in my personal computer. http://www.newegg.com/Product/Produ..._re=parallel_port_card-_-15-124-083-_-Product Sorry, I know this is the control board section of the forum and not the software but up to this point I hadn't really considered Mach3 as an option because of the price tag, but if it's free for up to 500 lines of code per job then I should be able to make that work no problem. You say for mild 2d work, does the demo have restrictions on curved faces or anything?
grbl boards not compatible with mach. trial limitations: http://www.machsupport.com/software/mach3/ system requirements:http://www.cncrouterparts.com/mach3-cnc-control-software-p-165.html It'll run on 32bit Windows back to win2000. 64bit with motion controller. If you have a 32bit system around then your good. 64bit starts to get pricey if you consider a motion controller. probably cheaper to buy a used 32bit pc unless you wanted the extra ports.
Is a parallel port to ethernet adapter a possibility for mach set ups? It seems that mach4 is more 64 bit OS friendly and has a similar demo package as mach3. http://www.pmdx.com/Mach4Hobby I think I also found that discussion comparing the TinyG and the xPro CNC you mentioned http://www.reddit.com/r/CNC/comments/2io0oo/tinyg_vs_cnc_xpro/ Could you explain this a bit more? From what I found on the wiki, there are a set of pin holes for using external drivers but it lists "J17 also brings out 3.3 volts and can be used to draw up to 50 ma." I'm assuming this is for signal purposes only, but would these pin outs enable you to use drivers capable of amperage higher than the default 2.5 amp limit on the board? I'm starting to think i need to bring these questions to the their creator specific forums.
You can use mach3 and 4 on a 64bit system with a cat5 to db25. It requires a motion controller. I'm not sure what that j17 is for. Maybe for external drive enable. Those drive pinouts are also only at cmos logic levels. 3.3v. I think typical 5v logic systems have a lower threshold of 3.5v. So that might be a small hurdle.
I was under the impression that the motion controler requirement for 64 bit opperating systems came from a lack of software drivers for the parallel port.
Came across this bundle, for the price it seems like a pretty solid deal. http://www.pibot.com/set-of-pibot-electronics-kits-for-cnc-2-0-c.html The drivers included seem pretty good, up to 4.5 amps http://www.pibot.com/pibot-stepper-driver-rev2-2.html The driver chip they use is a Toshiba 6600 and I haven't been able to find anything bad about it. I actually haven't been able to find much information about pibot at all really. I've seen it mentioned a few times but nothing in depth. For the price it seems that it would be all the electronics that I would need and then some that and all of the components are already heat sinked. Anyone here know about any potential concerns I should have about pibot?
I'm not really sure what's going on at the back end of 3. I don't know if it's driver compatibility or the programs architecture. Some have run 32bit XP in virtual mode with limited success. https://www.machsupport.com/forum/index.php?topic=13849.0 There are a couple members that have quite a bit to say about the tb6600. I think some of it can be read on page 90 and forward on the main OX build discussion. It's stickied to the top of cnc mills/routers. I don't have any experience with pibot. Joe
As it stands i think I'll go with a usb based controler, which one I'm not sure yet. From here i think it would be best if i finalized my mechanical design, did some soft number crunching, pick out some ballpark components and start a build thread. Or do you think that it's still too early?
Thanks to everyone who responded to this thread, hopefully I'll have a build thread up in the next week or so assuming fusion plays nice with the v slot resource files
Update time, finally got enough cash and research done that in comfortable in making my first order. Narrowed it down to two different controllers the sainsmart 4 axis usb mach3 controller http://www.sainsmart.com/sainsmart-...controller-card-interface-breakout-board.html Or an aluminum cased digital dream controler https://www.amazon.com/gp/aw/d/B00QY9CU2C/ref=ox_sc_act_image_5?ie=UTF8&psc=1&smid=A1U2FGX08PIQ6R Is the extra 100khz on the max pulse frequency worth the price? And i was going to pair the selected controler with 3 of these drivers http://www.sainsmart.com/cnc-2m542-stepper-motor-driver-controller-4-5a-support-nema17-23-34.html
Welp black friday was just too tempting, almost 6 months later I'm taking my first real steps forward in this project. En route is 1 of my y axis c-rail assemblies and a few misc. hardware pieces for me to toy around with. So while I wait for that to get here I'm using amazon to get my electronics. After a recent review on that USB sainsmart 5 channel I was interested in, I've decided to stick with the more tried and true parallel port set-ups, that and I was able to snag an old xp office computer from work. Going to use the openbuild's store nema 23's just for convenience. Using the Gecko formula that Joe provided I got Idea voltage would be ~52V and that I'd need atleast 7.4 amps for 4 of the open builds nema 23 steppers or 5.5 amps if I can get my 3 stepper design to work. With that in mind I selected http://www.amazon.com/gp/product/B0...ue&ref_=ox_sc_act_title_2&smid=A10EAPE4CAYC9P for my control board http://www.amazon.com/gp/product/B0...ue&ref_=ox_sc_act_title_1&smid=A10EAPE4CAYC9P and 1 of these to test out my single en route actuator on http://www.amazon.com/gp/product/B0...e&ref_=ox_sc_sfl_title_16&smid=A33TCMZEAWOM9U or http://www.amazon.com/Switching-Pow...TF8&qid=1449364289&sr=8-43&refinements=p_85:1 or http://www.amazon.com/Switching-Pow...TF8&qid=1449363703&sr=8-41&refinements=p_85:1 with that in mind I'm also ordering one of these as my power supply That last one has almost the ideal voltage but I don't like how close it's cutting it to the amperage minimum. Mechanical design is almost complete only thing left is the gantry, I keep thinking I've found a way to improve on it.
