PDA

View Full Version : Getting everything square



daveshorts
27-01-2010, 10:40 AM
Hi
I am in the process of buying things to build a gantry router with a cutting area of about 600x900mm with two steppers on the x-axis so I can mill the top of large things if I want to.

I was going to make most of the frame out of Aluminium profile, but I have noticed that there are no surfaces in my house that are flat to within a couple of mm let alone anything like the tolerances I would want in a router, and if I am not careful it is all going to end up on the squiff.

A spirit level seems a bit lacking in accuracy, and that is where my imagination has stopped so far. Are there any neat machinist tricks to get around a lack of planar tables? or am I going to have to see if I can borrow a flat surface somewhere on which to do up all the bolts?
thanks

BillTodd
27-01-2010, 04:30 PM
A spirit level seems a bit lacking in accuracyA good builders bubble level will be OK to a couple of mm over its length. A mechanist's level is capable of measuring to fractions of a mm. (e.g. <0.5mm/m) but at a price

You could probably make a precision level out of a good straight edge and a length of glass tube bent over a spacer in the middle of the tube (see attached). The tube has to be partially with water and a little glycol (antifreeze) to leave a small bubble. It doesn't have the internal barrel shape of a proper level bubble, but it'll still work.

Calibrate it by placing a a reasonable level surface, mark the bubble position, turn it around 180 degrees, mark the bubble position. Half way between the two marks will be level.

The flatter the bend in the tube the greater the accuracy and the smaller the bubble will have to be.

irving2008
27-01-2010, 04:46 PM
go to your local kitchen supply place, see if they can do you an offcut of granite worktop where they've cut out for the sink. i dont mean a b&Q etc as they don't do it themselves but template it and have it done elsewhere...you need to find somewhere that actually supplies the granite and does the cutting... I know of one locally to me, Planen, at Staples Corner... they have this huge CNC cutting and polishing machine, impressive to watch. I got a 600 x 400 piece of granite 40mm thick. It really needs to be mounted in a frame to stop it twisting for ultimate accuracy, (a real granite surface plate is 70+mm thick) but its good to 0.1mm as it is just sat on the bench...

I've seen the same done with 20mm plate glass but not so large obviously...

daveshorts
27-01-2010, 05:16 PM
I don't really want to use granite as a base, as it would be even more of a pain to move (I am in rented accomodation so am going to have to move the garage in 18months or so), and storage is a bit of an issue.

I do like the idea of a diy level though, I guess the important thing is to make the curvature constant and very small near the point the bubble sits. So the mounting wants to be as symmetrical as possible. I will see what I can knock up. I guess perspex tube would work too, though it wouldn't be quite as stiff so the curvature might be higher over the lump. I will have a go and see how it goes.

Was the glycol, to increase the viscosity of the water? or do something to the surface tension?

pavlo
27-01-2010, 06:21 PM
I have been thinking about this too. The solution I came up with is to make sure that the diagonals of the frame cross in the same plane.

Get a straight edge and lay it from corner to corner. Now measure down (using a dial gague or caliper) from the centre of this staight edge to a fixed point and record the measurement. Lay the straight edge across the other two corners and measure down to the same fixed point. If the measurement is the same, you know your frame has no twist in it.

Dan Brown
27-01-2010, 08:05 PM
a longish length of clear plastic tube filled with water works well for a makeshift level, just tape the first end to an upright with the opening faceing up, hold the other end and top up with water untill you reach the desired hight.

now you can walk around fairly freely with your end to check around your entire build.

i watched a couple of polish builders in portugal lay the main slab of a new house this way

BillTodd
27-01-2010, 08:27 PM
Was the glycol, to increase the viscosity of the water? or do something to the surface tension?

It wets the glass so reducing the meniscus.

daveshorts
28-01-2010, 10:38 AM
It wets the glass so reducing the meniscus.
That makes sense, I was worrying about that.

Robin Hewitt
28-01-2010, 11:34 AM
Rather than trying to make it square to planet Earth, surely you want the gantry square to whatever the material rests on.

Can't you simply put on a piece to be cut on the machine then adjust until it touches wherever it is supposed to touch?

For flat, a cheap laser pen glued to a flat bottomed lump and resting on the bed should hit the wall at the same height when moved around.

Robin

daveshorts
28-01-2010, 11:44 AM
The main thing I am worried about is that it won't be planar. You can take a rectangular frame, and twist it, just by rotating the joints. So quite a tiny error in the joints could produce a few mm error in the frame. This will mean that what you end up milling will not be flat either. I have a particular problem as the table I have just built for it is very definitely rough woodwork, so I definitely can't rely on that for a plane. Horizontal is just a convenient plane to use.

daveshorts
29-01-2010, 02:30 PM
I had a go last night with a 12mm ID perspex tube, there was definitely a problem with the bubble sticking with pure water. I added some glycol, which helped, but still wasn't perfect. I then added some mouthwash (don't ask) which seemed to help a lot.

I am not sure if it is the surfactant in the mouthwash wetting the tube better, or the alcohol pushing the glycol out of solution so that it preciptiated out on the surface of the perspex.

The bubble seems to move smoothly over the best part of a cm when I move the end up and down less than a mm, which isn't bad.

HS93
31-01-2010, 09:03 AM
I am new so may be talking rubbish , but can you not test the setup when finnished by placing a cheap DTI on the motor head touching the bed and then send the head around the table to see if it is the same all over ?

Peter

irving2008
31-01-2010, 09:11 AM
I am new so may be talking rubbish , but can you not test the setup when finnished by placing a cheap DTI on the motor head touching the bed and then send the head around the table to see if it is the same all over ?

PeterYes you could, but its still a good idea, and preferable, to get the frame square when building it. Waiting till everythings assembled is a tad late, by which time you are chasing shadows... is it the frame or the gantry or the bed thats not aligned?

Robin Hewitt
31-01-2010, 11:21 AM
I am new so may be talking rubbish , but can you not test the setup when finnished by placing a cheap DTI on the motor head touching the bed and then send the head around the table to see if it is the same all over ?

Hi Peter

Unfortunately the bed and gantry can be far from flat, but, having the same distortion, line up perfectly with regard to tool height.

This may be fine and dandy, only becoming a problem when the thing you want to cut doesn't want to conform as well. Or worse, starts out by comforming then changes it's mind when you start to cut it smaller.

I still like my idea of gluing a laser pen to a lump then moving it around watching where it hits on the wall. It's simple, cheap, amazingly accurate and would save all this faffing about :rolleyes:

Robin