[rthorntn]

Hi,

An old-school machinist buddy and I scrounged some metal and built a budget CNC for me, honestly he did the majority of the build on his own. I have the CNC in my workshop now.

I'm about to bring it online and I've read and understand the importance of it being flat and square so I need to check for and remedy any issues.

I'm a beginner and I want to eventually get to milling this in 200x100x5mm aluminium plate:

www.thingiverse.com/thing:3704406

It's only a 750W servo driven spindle so I wont be pushing it hard. I don't need super-dooper accuracy.

So could someone please help me with a simple checklist, with the tools to use to measure the accuracy so I can set to work fixing things, I have a straight edge and I can see that the table isn't perfectly flat (the table is strips of steel bolted on to two ally plates, the table is high in the centre where the two plates meet). I totally understand this will be time consuming and it would have been easier during the build but it is what it is?

Also I messed up and we made the table a bit too low, any ideas on how to get the workpiece up closer to the spindle, say 100mm, keeping it rigid, I don't want to cut down the uprights or anything drastic?

Oh, last thing, bearing in mind the low power spindle would you bother filling the frame tubes with sand?

Thanks!

[joeavaerage]

Hi,

Oh, last thing, bearing in mind the low power spindle would you bother filling the frame tubes with sand?

No, waste of time. I presume that you think that filling with sand will improve the vibration damping?
It is really a forlorn hope. In order for the sand to damp some vibration it must move and then disippate energy. The steel frame is very stiff and therefore will flex very little, and the sand must therefore
move an even smaller amount, so how can it dissipate energy?.

The truth is that if you use a thicker wall steel section, that would be stiffer, and the stiffer a structure is......the lesser problem you have with vibration. If you make it stiff enough then it won't vibrate
worth a damn. If its too flexible (compliant) then it will vibrate no matter what you do, and sticking sand (or anything else) in the tube is trying to put lipstick on a pig. The bottom line is that rigidity trumps
vibration damping hands down every time.

The simplest tests often give you the best feel for your machine.

For instance do a circular interpolation path, maybe 6 inches in diameter. Then measure as accurately as you can the diameter of the cut part at 20 degree increments to see how 'far out of round'
it is. Another simple test is cut a square shape, say 4 inch square and then measure the diagonals. Any 'out of square' will result in diagonals of different lengths.

Many static tests require things like a granite square, which, even cheap ones, are still quite expensive. Even worse, you'll use them once or twice and then hardly ever again.

I've got a Precision Granite (CA USA) AA grade (50 millionths of an inch over six inches) granite square, paid a bloody fortune for plus shipping to New Zealand and I've used it about four times. OK, its great and superbly accurate when you do want to use it
but its a lot of money to spend when you are not likely to use it a lot.

Craig

[rthorntn]

Thanks, I should add that I have a granite square and a dial indicator.

I'm pretty sure the frame is 75x50x3mm RHS.

The width of the table is quite daunting and I've no idea how accurate I should try to make everything, like maybe 0.2mm over 500mm or something like that?

[joeavaerage]

Hi,

I should try to make everything, like maybe 0.2mm over 500mm or something like that?

That depends on what you are trying to make. My guess is that 0.2mm might be OK in a wood working situation but would be be hopeless for an engineering part.

I would suggest either measure the machine or make parts which reflect the machines capability. Then ask yourself 'is this good enough'. Often times the basic accuracy of a machine is built-in....
its not something that you can correct later, its either there or it is not.

My own machine is a mill, primarily intended for metal small parts, say 10kg or less. As such I have tried to make it accurate to 1 um, but inevitably missed that by quite a margin.
What I have achieved is about 8um over any 150mm cube, which for a home made machine I think is pretty fair. I use my machine daily for my business and the level of accuracy it can and regularly
delivers is entirely satisfactory for the work I do. If I were having to make high tolerance automotive or aerospace parts then my machine would be stretched, and is not really suitable.
Would I like to have a machine capable of that level of accuracy.......well yes I would, but can I afford such a machine, then no I can not. I am happy with what I've done.

In recent months I bought a new spindle 3.5kW(S1), 3.4Nm(S1), 10krpm(rated), 40krpm(max), 400V, HSK32E tool interface. I had to buy a new VFD, make a spindle monitoring board. Since
Christmas I've made a refrigerated chiller for the spindle cooling water. I've just built a new coolant tank, 216l, with a three phase pump. I completed my trunnion and fifth axis last year. So all in all
my machine is progressing.

I think the bottom line is don't waste too much time trying to measure this and that, but get on and start making parts. That is what really informs you about your machine. It may well prove that the accuracy of
the machine is good enough, but the spindle is shonky......or you need flood coolant......or something else entirely. You won't know until you start making parts.

Craig

[rthorntn]

Thanks!

Wow your mill looks great!

I want the CNC for DIY computer enclosures and so I think that 0.1mm over 500mm would be plenty.

[joeavaerage]

Hi,
0.1mm over 500mm sounds very doable. With any sort of luck it will already be within this spec. I imagine any old school machinist would leverage is experience to get
he best result.

Try this idea, it relies on geometry rather than expensive gear.

Get yourself some 2mm drills, half a dozen or so.

Using your machine drill four holes in some stable material like a sheet of aluminum say on a 300mm square.
The Gcode would look something like:

G0 X0 Y0
G82 (or whatever drill cycle you use)
G0 X300 Y0
G82
G0 X300 Y300
G82
G0 X0 Y300
G82

Install a 2mm drill in each hole, they are being used, in effect, as precision ground pins, but drills are just so much cheaper and are every bit well ground enough for our purposes.

Now measure the diagonals of the square.....they should be identical. The diagonal measurement is quite sensitive to the axes being out of square. Even a small 'out of square' will
result in a moderate difference in the diagonal measurement.

Another simple trick is to repeat the process with a fresh piece of material. Now you have two identically drilled pieces of material. If you rotate one piece by 90 degrees if everything is
perfect the holes will line up exactly. If either the axes are slightly out of square then they will not, nor either will they if one side is 299.9mm whereas the other side is 300.1mm.
Rotating the topmost piece another 90 degrees allows you to estimate whether you suffer from 'out of square' or 'different length sides'.

Using geometry like this can tell you quite a bit about your machine without having to spend a lot of money.

Craig

[rthorntn]

Thanks, yeah my friend would have got it pretty square.

My straight edge tells me my table isn't flat, I'm a bit lost on how to approach correcting that, it's all bolted together so I could start shimming, I don't have access to a surface grinder and using the CNC to flatten its own table ain't gonna happen for a number of reasons.

I feel like I cant measure for accuracy unless I have a reference point, would that be a flat table?

[joeavaerage]

Hi,

My straight edge tells me my table isn't flat,

So, who cares? Did you not say that you need some sort of block or riser to elevate the workpiece so that the spindle can reach it? Then just place you block/riser or whatever on the table
and then use the machine to flatten that.

I don't have access to a surface grinder

I call BS. You live in Australia and Australia has a large engineering industry and there will be plenty of surface grinders if you look. What you are saying is 'my local engineering shop
does not have one, therefore it cant be done'. Rubbish. You'll have to search a bit wider than that.

The biggest in Christchurch, to my knowledge is 1.2m x 400mm. In Auckland there are two Proth surface grinders (2.2m x 1.25m x 0.7m and 1.6m x 1.3m x 0.7m), and that's in New Zealand,
are you suggesting that similar is not available in Aussie?

https://www.egwhiter.co.nz/index.php...formation_id=8

https://www.egwhiter.co.nz/index.php...ormation_id=10

Craig

[rthorntn]

Many thanks, makes perfect sense.