[native34]

I have kind of a dilema. I am tying to understand why my parts are not coming out right, but something doesn't make sense. When i measure the wall thickness it is coming out to proper size, but all my otherdimensions are coming out not right. I have attached a picture of what i'm machining and the results of what i machined are as follows: 2.384, 2.131, .941, .690, wall thickness .125 all dimensions are in inches. What could make these results?

Attachment 313154

[native34]

Any help please?

[MichaelHenry]

It might help to mention which brand and model of mill you have.

[native34]

I'm sorry I thought it would have been understood that it was a charter oaks mill, as I placed it in the charter oaks section of the forum. Anyway it is a new charter oaks 12Z, also used to be called industrial hobbies mill. I have a vfd with 3600rpm 3hp motor, the old style IH cnc kit with kelinginc servo motors, amt102-v selectable encoders set to 250cpr, cncdrive.com drivers model dg4s-8020, cnc4pc c32 breakout board with Ethernet smoothstepper. When testing each axis with an 1" travel indicator they measure what they should. I.e. If I command a 1" travel it measures 1" on the indicator, but when I am cutting parts the dimensions are not coming out as they should.

[SignTorch]

the ratio of measurements/dimensions is about 1.003 - which suggests the XY steps per unit may need to be multiplied by 1.003 - then make sure you take up the backlash when checking travel with the dial indicator - just guessing...

[native34]

So are you telling me that after I calculate the needed steps per value and I test of hat value, to multiply the difference by the current steps per to get the new steps per value? I'm going to include an example below. Let me know if I am correct.

Ex. Calculated steps per value 5301.0000 for x, and y. I then machine a 1" square and measure it. X measures.995 and Y measures 1.006. So the. I take my steps per for x and multiply it by .005 to give me my new steps per for x, and take my steps per for y and multiply that by .006 and that will get me my new y steps per value. If this is true how does this factor in when using a 1" dti I'm getting the move I'm commanding? The dimensions are only off when actually machining. I've already taken into account for any mechanical backlash.

[SignTorch]

I was just suggesting something you could easily try - if the part comes out more accurate then I'd suspect calibration issues (incorrect calculations, steps or measurements) - if it makes it worse then I'd suspect mechanical issues (flex, backlash, etc) - I don't have enough information to elaborate

if you move 1 unit and measure it then you multiply the steps by the measurement (for .995 measurement you'd multiply steps by .995 not .005 - for 1.006 multiply steps by 1.006 not .006) to get the new steps per unit value

I've always calculated steps based on the drive components and the only way that can be off is by measurement error and/or mechanical issues - both come into play when measuring moves vs. cuts then using that to set steps

you just have to try it and see what happens and go from there