first of all, I’d like to thank Ryan and his ‘crew’ for sharing all this. I have been reading through all the documentation and many posts. At first, I thought to add some modifications and build some hybrid of mpcnc and lowrider. But the longer I have been reading and thinking about all this it got more an more clearer how well designed everything is - so changing things will most probably make it worse
Here it is, after plotting the crown:
[attachment file=94983]
sorry Ryan, cable ties not yet shortened on some axes (I wanted to wait until I’m sure tension is correct)
But first things first:
I spend yesterday building the table. I recognized that I could use 2 concrete walls, which should add a little stability to the table.
[attachment file=89439]
[attachment file=89440]
[attachment file=89441]
This is 15mm OSB, and on top 18mm MDF (which will be cutted out in the mid for easy replacement later on.
So, now I’m at hitting ‘order’ for the stainless steel tubes and ‘print’ on the printer; but I’m still a bit unsure… I plan to have 120x70cm working area, and this could be OK for the foam parts but too large for the MDF moulds I plan to mill. On the other side I think: try it out, if it is not ok I could then decide to resize the machine and build an additional lowrider
Just loosen the corners and end of axis and move them all in closer when milling the smaller designs. Only go out to the ends when you need to cut the big foam.
It took some time, as I had to replace all bearings, as the fist set was real crap. I ordered new ones, now everything slides smoothly.
I already added the large spindle. A friend of mine strongly influenced me taking the larger one because of the collet size. I’m not so sure; I think I should have taken the 800W version. I could finally align the spindle vertical adding som extra small spacers in the mount. Should work.
It is looking very nice so far. With those stainless steel tubes I don’t think the larger spindle will be a problem, unless you try to move everything really fast.
I was a bit disappointed at first because the roundness is not so good:
put a micrometer gauge bottom up and rotate the tubes: the difference is about 8/10 mm. Quite large.
This non-roundness however is constant along the tube -> sliding along results in variations of about 1/10 only.
I was afraid that the spindle will hang through in the mid of the working area, but the effect is not so large (I decided to build the smaller machine).
I plan to add mid-span supports when the electronics works.
[attachment file=96399]
To recap: my working area is ~ 910mm x 570mm, I have endstops
Repositioning even along the long axis ist mostly exact even when travelling far away and back again; someties 1/100 mm ‘inaccurate’. This is great!
z-axis has backlash of 4/100mm (which is usual for the lead screw setup) - question: I once had a rigid semi-sized machine with ballscrews on all axes which had reversal backlash of some 1/100, which was compensated in software. This worked well. Something similar possible in marlin? (I dont think so).
I the used a scale to measure the deviation from position when applying a pressure:
400g -> 1/10mm
1200g -> 1.5/10mm
1500g <2/10mm
Ryan stated somewhere milling of aluminum requires about 1.5kg -> I think (and this is astonishing to me) that leaving 2/10mm material for finishing, and then using 1 or 2 finishing passes could get better than 1/10mm for aluminum - I will believe this when I saw it
