Thread: Mach3 VB programmer needed

This is a great idea, something like this as a wizard maybe. Set size of bed and distances between measurements, then at the end it could spit out a list with measurements from a probe. Not sure the accuracy compared to a dial but i couldn't see it being completely out of whack. Of course this is all taking into account that the frame and linear components are leveled first.

Daniel

Hi Daniel

Thanks for the boost.

I don't see that it really matters if the frame and rails are leveled first.
Both sets of rails MUST be the correct distance apart from one another as otherwise we would not be able to play trucks with them before we hook up the motors, and I can't see what difference it would make if the frame was twisted as the twist would be taken out with the shims or other adjusting methods.

As to accuracy, my copy of Mach3 tells me that it is accurate to 0.0001mm. I have a small (huge) amount of trouble believing this, but even if we drop the last two decimal places, it would still be accurate to 100th of a mm and that is more than close enough for me.

Bob

Bob,

no worries at all, its a great idea and to me there wouldn't be much in the software side of things (i'll wait to be told different by someone with a little more knowledge than me) other than the trying to post the results to a file but then again i spose its the same as a point cloud takes its measurement. Making it a wizard is better and will require a few extra variables but would be widely universal then.

I agree with the accuracy thing, 100th of a mm is great, and worst case is once at this level i guess most people would then skim the work surface anyway.

As for level, the main rails probably aren't too much of an issue but the gantry needs to be level because if one side is lower than the other side then your measurements would be out showing low spots on the higher side and high spots on the lower side.

Here's hoping this great idea gets off the ground, i am building up to my wedding next month but if i can find some time between now and then and no one with better VB scripting come forward i'll have a shot at it.

Daniel

Originally Posted by twistedfuse

As for level, the main rails probably aren't too much of an issue but the gantry needs to be level because if one side is lower than the other side then your measurements would be out showing low spots on the higher side and high spots on the lower side. Daniel

I don't believe that the gantry would be an issue either Daniel because that is what you are using for your datum line. As such, it MUST be correct because that is what you measuring from. So even if one side was 100mm higher than the other, then the base plate would be adjusted up accordingly until it was parallel to the gantry on both sides. Yes it would look a bit odd.

Congratulations on your upcoming marriage. I hope it all goes well for you.

Bob

Originally Posted by Bob Willson

Hi Daniel

I don't see that it really matters if the frame and rails are leveled first.
Both sets of rails MUST be the correct distance apart from one another as otherwise we would not be able to play trucks with them before we hook up the motors, and I can't see what difference it would make if the frame was twisted as the twist would be taken out with the shims or other adjusting methods.

As to accuracy, my copy of Mach3 tells me that it is accurate to 0.0001mm. I have a small (huge) amount of trouble believing this, but even if we drop the last two decimal places, it would still be accurate to 100th of a mm and that is more than close enough for me.

The twist is still in there Bob. You have only adjusted the table with a twist to match your X and Y axis rails. For 90 per cent of jobs it will not matter because when you mount your job and clamp it to the table you will flex (twist) the material to match the table surface.

All good so far because the path of the router bit will follow the same twist.

The result is that when you unclamp the job and it flexes back you may find the cut edges are not at 90 degrees to the face and the distances between points on your drawing will be longer on the cut material. The 10 percent of jobs that will not work are the ones that are too rigid to flex to the twist in the table. While you have got your rails parallel (equidistance) on the horizontal plain the cure is to get your X and Y axis rails on the same vertical plain. Daniel was pointing to this in his post.

The variations will be minute and I am not being smart by saying this but it is important that forumites understand what is happening.

I would highly recommend using a dial gauge to zero in your table. Steppers can hold half a step at best and the rest is theory. For Gecko drivers that have 10 microsteps this means at best the motors can hold 5 of those microsteps. The value of the microsteps for steppers is to more accurately round off small numbers to position the motors.
Add to this that your ballscrews are probably C7 rolled specs then their accuracy is 0.05 in 300mm. Without going into too much detail you are operating software assuming the machine position matches your software settings.

I hope I have got a few of you thinking about how all this works but having spouted this most the the manufacturers recommend surfacing the table with the machine which is in effect what Bob is doing.

At least you know a bit more about your machines if you understand my waffle.

Bob,

Rod got it in one but has a hell of a better way to explain it. It is all minute changes we speak about but in a machining sense are major factors and can mean the difference between close tolerance fitting parts or slightly out, which to 90% of people is ok since most woodworking is measured in mm not 100ths of a mm, except most of us aim for the best especially when doing things like photocarving lithos and the likes which does require fair accuracy.

As Rod stated, a dial indicator will be the best thing to use, especially with the details like Rod outlined with microsteps and the likes since it is possible to lose steps and also loose accuracy with the rounding done by software and steppers etc. In saying that i didn't expect this to replace a dial indicator, more so to help get the table close before running a dial and also for those who are on a tight budget or do not need the accuracy some of the others do. For me it would be a simple way to quickly assess the table to shim and get it close to what i need, then run a surfacing bit over the sacrificial table and follow with a dial to check. This is a cheats way of getting it a fairly close. Probably cost a fortune but i wonder if there is a digital dial guage with output, that would be pretty cool getting the accuracy of the dial and still having it logged in the computer.

Bob, thanks for the well wishes, it can't be all bad, my partner already let me build 2 machines so i'm going ok and i'm in the process of buying a combination planer/thicknesser to help with panels etc and am still looking for a table saw as well. After that i have the go ahead to re-build my cnc frame so i count myself lucky. She would just like to see some more work done for her or for our house....)

Daniel

***EDIT: Shocked but a 0.01mm digital dial is about $60-$100 for an entry level and most have a digital output so there maybe some merit in this also, but am yet to see one with more than 0.01mm accuracy which makes me think twice****