Hey guys, been a while since I posted anything here but I’m starting to do the leg work for my next project. Me and a few fellow students from my college Robotics class want to make a cnc router for cutting wood. As you can see from my profile I have already built a cnc mill but it has a small travel and is to slow for cutting wood. We want to build it as a kind of learning project and the main goal is to make a good working router on the cheap. Mechanics are not really an issue it’s the cost of electronics so if anyone knows of some good online sources for diy driver circuits and other goodies that we can interface with an Adruino I would appreciate it. Thanks folks.
I am playing with just using a cheap chinese driver board (Under $40 on ebay.) and hooking it up to an arduino through a parallel breakout board ($6). The single 3A stepper boards are $7-$10. I really can’t see how you could get cheaper even making your own.
Check out http://openbuilds.com
Thanks David, I have been looking around on ebay and ya you are right some of those boards look pretty cheap. Any recommended boards that would work for a small 2x2 foot router? And thanks nick I will have to go to openbuilds again been a while since I visited that site, time is getting away from me lol
What is your overall budget for the whole thing?
Good question, I know as far as mechanical parts I’m going to source them from free places like scrap projects so really electronics is the main cost, so going to say as close to $200 as possible, when I built my cnc I had a working man’s budget now I have a college student’s budget lol.
@Will_Dent Depends on how you are going to drive it. Leadscrew doesn’t take as much power as direct drive but you said you were looking for speed. Belt, chain, or gear rack would be fastest. Keep it light and you don’t need huge motors.
@Will_Dent
I’ve made my own electronics for under $200. I’d say one of the biggest challenges I faced was coming up with a power supply the right voltage, that can deliver the current the project needs. Doing that right on the cheap is not always easy. I know it wasn’t for me. I lucked out and found just the right step down transformer for the voltage I needed.
I’d be very suspicious of any cheap surplus switching supplies. It might work, or it might just be a great source of signal noise too. Your PSU is the foundation that the rest of your project runs off of. If it has problems then everything else it powers is going to have problems too.
Making my own boards would not be a problem nor would building our own power supplies because these are the type of things our coarse is teaching us but we don’t always get the lab hours to backup the theory, so brewing our own electronics would be kind of fun. That said If it’s way cheaper to just buy it then that might be the way to go, but if we can build something a little better quality for the same price then that would be the way to go. Know of any designs we could use for drivers?
I would like to be able to work at 2500mm/min (100 inch/min) or higher in wood no aluminum thats what my CNC is for. Not that fast I know but way faster than my CNC and probably realistic for cheap electronics, hopefully. Probably use gt2 belts with either an mdf design or the ox type aluminum structure.
@Will_Dent
Well, there’s my design
The only thing I didn’t include is a current reduction hold circuit. Other than that I think it works great.
Here’s my buffered BOB
It works great too. But there is some kind of a strange back feed issue with it that I never tracked down. If it was a problem I’d have spent some time figuring out what is going on there. But it seems harmless.
I still think today you’re better off just buying one of the imported boards and fixing the thing. They don’t do the current sensing right really.
@Paul_Frederick
Thanks alot that is exactly what I was looking for, didn’t come up in any of my google searches. If I do go the route of home brew drivers I will definitely be making an Honorable mention and putting up your link in the wrightup and online documentation that I’m going to make about this project. Again Thanks and great work on those drivers and b.o.b.
Unless you are doing it for meeting special application requirements or as a learning experience (or just for the fun of it), you cannot come anywhere near the price you can buy them pre-built. If you are going to go to all the trouble of building from scratch, might as well do something interesting like a closed loop system.
well the only real problem I have is buying from chinese sources because of me living in Canada I have to really watch shipping charges, could add 20 bucks to something thats worth 6 and noone seems to sell those boards here, that I know of anyway.
Thanks for the comment Nicholas, I’m probably using linux cnc or grbl with an arduino, if I was looking for a more robust setup I would probably just order another kit from automation technologies but I’m an educated poor college guy now so trying to make my tuition pay for itself 
@David_Kirtley
Before you run around saying, “you cannot come anywhere near the price you can buy them pre-built”, realize that I’ve done exactly that, and better too. It was certainly quite a bit of trouble though. So you are right there.
Today I tell people to just buy the imported boards, and fix them. Because as shipped those boards do have some shortcoming. I can tell that just by looking at them.
The big problem is those boards are so cheap few really invest the kind of time it actually takes to understand them. When I made my drives I was much more motivated, because none of those cheap drives even existed back then.
Now if you’re going to go closed loop then it is a waste to even use steppers. You might as well run real servos then. Here’s an amazing BLDC drive a fellow I know made
http://www.delorie.com/electronics/bldc/
Good luck with a project of that magnitude though. I’d say that is well beyond the capabilities of most folks, me included.
@Paul_Frederick Just out of curiosity, what kind of sourcing do you have? I find Dual H-Bridge chips that can handle 3A like the TB6560 and the STK682 on Digikey for $8-$10 a pop. The built boards on ebay from US supplier (yes, Chinese made) are $10 with free shipping. You are behind even before you start buying connectors, boards, switches and heatsinks.
I can walk down the hall and use the LPKF board router and scavenge some parts from the parts room at work and still can’t build cheaper than I can buy.
Thanks guys appreciate it, definitely got some better ideas on the way to take this project. Will be posting pics when I get started.
@David_Kirtley
Back when I bought my TB6560AHQs they were $2.83 from Digikey. That was before folks started blowing so many of them up in those imported drive boards though. I bought 6, and fried one (whoops, board short!). So I still have 2 extras kicking around now. I figured I’d kill more designing a drive. So I bought some to burn. TBs had already garnered a reputation for exploding even back then. There’s a better drive IC out now. A THB6064AH. It fixes some of the grosser shortcomings of the TB6560, like the low input voltage handling. It has a bit more current handling capacity too. So it is sort of like a TB on mild steroids. It is almost up into the cheaper Gecko drive range in specs.
If I had it to do all over again I’d be seriously considering those. I know with a TB6560 I’m always wondering just what I’m missing out on. Practically not much. Because when I try to run my leads up to 1,000 RPM it is ah, a bit spooky. Yeah, that’s how I’ll say it. There’s theory, then there’s reality. In my case my cheap acme lead screws are not the straightest things going either.
I’ve made drives using the Allegro SLA7026M too. They are pretty pricey. Total PIA to use too, because they don’t have built in sequencers. Least they have the nice adjustable current feature. I really didn’t know what I was doing when I got into those. But they work today nevertheless. I’ll say this for those, they sure run cool. I have the barest of L shaped heat sinks on mine, and that is fine.
@Paul_Frederick I got as far as building my own H-bridges just to learn about it but I haven’t gotten into the low level electronics that much. I hate to get even more partial subsystems and components to get bogged down with instead of actually building stuff. I have 3 (potentially 4) simultaneous builds going on now and nothing completed.