Hey I am in the market for an XY table that can be controlled by DC motors. The XY table will have some sort of advanced position sensor (cm accurate GPS or laser positioning, etc), so it is always aware of its position. Therefore I don't need a modified servo or stepper motor, DC should do just fine.
The problem, is I have an idea of what building needs to be done, I just don't have the resources, i.e. a machine shop where I currently am. The other issue is the table has to be pretty big. 4x4 feet, perhaps bigger
So does anybody know a resource where I can buy a pre assembled 2D XY table along with the DC motors for it?
If the table is 4ft square how much movement do you need? If it is 4ft each side of the centre that means you are having to have a 12ft square area for this device. That is enormous. I have been looking at x-y tables for the last year and there is nothing remotely at this size available. I am not sure what you expect to pay but something like this is probably going to cost at least $5000 to have made for you.
Stepping motors are common because they are cheap and easy. With DC motors you have the problem of integrating the position sensing with the motors. This gets very computationally heavy and is known as "servo systems". It is all about defining step response time and calculating critical damping, or under damping feedback functions depending on how you want it to move. It's the sort of thing people used to do with analogue computers in the 60s. You can do it with digital computers but it involves integration and differentiation.
I don't think GPS would update fast enough to allow you to use it. Are you sure you have to move the table and not say a pen or extruder over the surface?
I'm worried about the slippage of the stepper motor. If the motor skips a tooth, my position is out of whack. I was looking into laser triangulation or something so that the device always knows its position.
Are you sure you have to move the table and not say a pen or extruder over the surface?
I'm not entirely sure of what you mean here. I thought an XY table was the entire table. I want to sense over a 16 foot surface area. This is what I am imagining.
It would be something like this. The red circle would have my microcomputer, a sensor to sense the area I am in, my position sensor, and some sort of battery to power the unit, and my motors. My motors could be moved to the rails and made stationary, but I would rather not have to do this, cause of the wiring issues.
From what I reckon, the setup is going to be pretty heavy if it has to move all these components. So i wanted to know if this setup was even possible.
If its not, I'll look into just moving the 2 sensors and battery, mounting the motors at the rails, and having sensors send all the information wirelessly to an offline computer.
If you look closely at the picture you will see that it is using optical encoders for the position feedback sensing.
Stepper motors are probably the best solution for this kind of X/Y table. It's not a big deal if the motors skip or slip steps, as long as it doesn't miss encoder steps used to keep track of the table's actual position.
This is a good application for an Arduino board. First get the encoders working well then move on to commanding the stepper motors.
Those aren't encoders with stepper motors, those are servo motors with external potentiometers.
Using optical encoders gives me two problems - missed steps (which I've always had a problem with), and no memory on reset. I know I can interface with a RAM, but that still doesn't eliminate the missed steps issue. Plus if I have to get a robust stepper motor and driver, that adds cost anyway, and might justify the triangulation method. The idea of using a triangulation method, was that the device will knows its exact location all the time.
Also I was hoping for a more plug and play solution for the table and motors, cause I don't have the tools or the time to do much building on my own, and just have to worry about mounting the sensors, and the closed loop - read data from position sensor, move table based on data, repeat.
You need a router table.
If it is high speed it will probably have dc motors already fitted. Then all you will have to do is figure out the absolute position sensing.
The easiest way if outright accuracy is not important is to use multiturn potentiometers.
Anything with stepper motors will be desperately slow.
Anything with stepper motors will be desperately slow.
No they can move quite fast if you give them enough voltage and a chopping current limiting supply and then accelerate them to top speed. In other words do it properly.
Well that is not what I call a table, jabber was right you need a router, although I am not sure what you want to use it for because what you want to route will determine the materiel it's made of and the design.
There is also the problem of backlash, some solutions like a ball race drive are very expensive. What you have to decide is what sort of accuracy you want / need.
There are router kits / assembled mechanics from about $400 upwards, best bet is to go on ebay and see what you find. It depends on where about in the world you are as to what is available.
For home made look at the reprap machines (gogle it) , full open source plans and variants. Then there is the cup cake and the maker bot that have a following as well.
There is nothing to stop you using a position sensor or optical encoder with a stepping motor but you will not skip pulses if you get the setup correct. The problem with liner sensors is that they are expensive. You can always solve the memory problem with limit micro switches to allow you to calibrate on switch on.
Great! So I need a router table not an XY table. No wonder I wasn't finding what i was looking for when I was searching.
I actually dont have to cut anything. The moving center, wont actually come into contact with the surface, its just sensing it. Kind of like a scanner. Does this drastically reduce my costs?
Now with backlash, do you mean the spring effect when the center reaches it's hard stops?
I am currently in Sweden. I am actually a Canadian but I'm her for the next little bit. Thus I am looking to buy either a kit/assembled machine, cause I don't have the ability to make one myself.
I looked at ebay, and googled a bit, but mostly got router saws. I essentially just want the table and the motors. Most of them come with controllers and software and I'm seeing them start at 20,000 and look much heavier than anything I had in mind. Any suggestions? Anybody come across something?
I have also been looking for routers on Ebay and found that it is important to include the 'cnc' bit in the search terms, if it isn't there then hundreds of normal router tables show up in the listed items.
The cheapest I have seen (buy it now for a large-ish machine) is 400 uk pounds for a table 2.5 mtrs by 1.5 mtrs, described as old but serviceable.
They seem to be a desireable item and anything with good linear bearings and ballscrews reaches a high price.
I'd seriously suggest you look outside Sweden. The market here is small so limited choice and high prices (Not helped by VAT 10% higher than, eg UK). Same is true of second hand stuff, low volumes, high prices.
Think you also need to be clear about what performance you need to achieve, how fast does your 'payload' need to move?, is 'point to point' positioning good enough, or do you need to follow a path?, what forces are involved - are you just moving the carriage or are you doing work with a cutter of some kind?
Something along the lines of the system shown above but using extruded aluminium profiles and ball bearing mounted carriage instead of wood and draw sliders would be pretty nice, structurally sound and eminantly scalable. There are commercial systems of this kind (eg Bosch Rexroth www.boschrexroth.co.uk - look under Linear Motion) but a DIY solution using whatever aluminium profile you can find cheap, and a bit of inginuity for the carriages is quite doable.
As far as control and positioning, there are some good answers above. Google "motion control" should find you more pointers. Also "DIY CNC" (Computer Numerical Control)
I actually dont have to cut anything. The moving center, wont actually come into contact with the surface, its just sensing it. Kind of like a scanner. Does this drastically reduce my costs?
Yes, theoretically. A lot of effort goes into commercial router tables so that they're able to hold and push a heavy tool against a workpiece. If you're willing to build it yourself, something like the picture you posted would probably work pretty well for your application, but be pretty questionable if you wanted to mount a machine tool in the middle.
Unfortunately, the commercial implementations are aimed at tooling (there are "lots" of commercial designs designed to cut pieces out of standard 4x8ft pieces of plywood or plastic. Web search "shopbot", for example.) So in practice you probably won't find an "off-the-shelf" cheaper table. (4x4 shopbot seems to run about $16k)
You might search for "CNC foam cutter"; they're designed for the smaller loads, but they also seem to mostly be aimed at 3d applications...
Yeah I think I'm giving up on finding a cheap router table, and am exploring options of building one. The reason I wanted to find some sort of off the shelf solution is cause I didn't want to worry about material type, dc motor specifications, max width/lengths, etc; and only concentrate on the sensors and programming as thats a substantial task as well.
Anyway, I put together a few drawings yesterday and I have a meeting with a local machine shop early next week. Hopefully the language barrier (I know no swedish, and the guy at the machine shop knows no english) won't be too much of an issue!