So I wasted a couple days cutting out some aluminium plates really really slowly, and in the end I noticed they were quite skewed and what I had cut out was completely useless to me. The Y Plates were not running parallel on my CNC machine. One side was about 4mm behind the other so I ending up with trapezoidal instead of square plates.
I think what had happen was, at some point the y plates had reached their limit, which is only on one side and as I don’t have physical hard limits, while one side crashed into the proximity limit switch and opposite side kept going. Stupidly, it didn’t occur to me that this would mean I would have to square the gantry again. But I had to and I didn’t.

Anyway lets look at the positives – I just cut loads of aluminium. That’s an achievement in it’s own right. My feed rate is 230mm/min with a step down of 0.25mm and a plunge rate of 80mm/min with the spindle running at 15000RPM… well that’s what I thought I was doing anyway but my acceleration is quite slow and this actually effects the finals speed I cut at, depending on the particular shape and the tool-path I selected. I’m going to come back to that later in this video.

Anyway I decided that I should add some kind of physical stop to the y axis which is going from the front to the rear of the machine, so that if I accidentally crashed the gantry into the end plates again, at least it would stop at the same position.

Ok I made a massive cock up! I think that had happen on the CNC machine was the two Y Plates were not square, so they were off a little bit. On a positive I managed to cut some plates out which is a miracle and in fairness the cut doesn’t look too bad but I have wasted the material. When the bottom is lined up and the top is flipped I’ve literally got a gap on this side or about 1mm and a bit. And because of that it means what should be a circle here because of how I have orientated the pivot holes, is now elliptical. And one ellipse is going this way and plate underneath is going that way and essentially these plates will not pivot. So with the design I have made it is really really important that the CNC machine is properly squared. I actually trammed the spindle so I compensated for the nod, and the tilt in this direction along the x axis but what I didn’t do, which I forgot to do and it was very stupid was to move this back into position so they were parallel. What I could do is take a bit of 12mm threaded rod and screw it in on this section here so that if the gantry was to but into proximity sensor on this side it would hit onto the opposite side at the same amount. I’m not really sure I should be using the proximity sensor at a physical stop although I have managed to ram the gantry into it a few time.

I want to add a machine screw like this one here somewhere on this place so the Y plate, when moving in this direction has a physical stop. I think it makes sense to place this so it makes contact with the bottom of the plate there, down here. But obviously I don’t have the material on the opposite plate so now I’m thinking I should stick with this plate type verity and get rid of the back entirely so there isn’t that option.

So I’m just in the sketch for the motor plates. What I’ve done is created a circle which is the diameter of the coupler (19mm) and I’m going to make another that is 25mm as you can get fixed couplers which are a bit larger. So I’ve got a visual reference of what I need to avoid. Now I’m going to make a random hole which should be the same size as the M5 tap hole which is 4.2mm, and what I’ve done is I’ve already worked out the spacing based on the actual model. So I created a sketch somewhere else on this fusion document position the holes and superimposed it on the model so I could see where I needed to put this. So I know if I click on the centre of that to there, hard stop centre x and that’s moved in. Now I’m going to do the same thing and this is going to be the y. one two three. So this hole is a quarter of the way down of the face plate that will be coming towards it on the y plate. So that’s the new design change. And I think what I’m going to have to do as well, is get rid of these plates as well as there is not material there to place a machine screw like this so it just makes sense to mirror all those parts. So two are the front and two at the back.

So the next thing I’m going to do is square the gantry.

This is not to be confused with tramming which is aligning the spindle along the x and y axis so it doesn’t travel while moves along the z – squaring the gantry is a little bit simpler.

I drilled four holes ever so slightly larger than 6mm and used these broken carbide bits – which I pushed into the openings to measure between each corner. If the gantry is square the space between opposite corners through the diagonal should be the same. That’s 151.70mm and 151.67mm. So it’s ok it’s square now. They were pretty much bang on. I had actually already checked the distance between the front plates and the Y plates with a caliper, turning off the controller temporarily and manually adjusting one side by rotating the coupler until the distances were the same. 75.67mm and that’s 75.69mm. Anyway it’s square now.

I mentioned earlier that the acceleration effects the overall speed. While cutting on bCNC there is a live reading of the federate, and depending on your acceleration and the geometry of your vectors, I noticed that sometimes I would not reach that set speed. So what I’ve done is in the tools tab, and in controller I changed the acceleration of the x and y axis from 10 to 35mm/sec. I then screwed down one of the wasted skewed plates so I could re-cut out the bottom z plate and try use up some of the scrap material.

So what I’m going to do, is try and follow some of the tool paths to remake this plate now that the gantry is square. I should be able to right across there so I don’t waste this plate. But what’m going to do, is use this button here. When I built this machine I added this toggle switch which allows me to go between the speed set in the gcode and to set it manually using this potentiometer. So mid cut I can press that button and change the spindle speed by hand – twizzling the potentiometer. You can also do this in bCNC with a toggle bar, but only once you have upgraded to GRBL1.1. Along side this you can also increase the feed rate plus or minus 100%. This is interesting because it means while you are cutting you can push and tweak your toolpath until the cut sounds and looks right. I guess what would make this even better is if the spindle speed automatically adjusted depending on feed rate – which changed during cuts but that’s a little beyond my ability to programme.

Drilling the 5.1mm holes sounds a lot better through the ear defenders. In the end, while the holes came out in the correct positions I could tell they weren’t perfect. The larger holes however, cut like a dream and the speed was ridiculous. While I had set the feed for the smaller holes to 450mm/min, the realtime feed rate indicated on bcnc actually oscillated between 100-150mm/min – I think it’s worth going half that speed. The feed for the larger bearing holes was another story. I manged up to 750mm/min there. While cutting the outer vectors I reached 855mm/min which is ridiculously fast for aluminium. I was using isopropanol as a coolant but when I began to run out I moved over to mentholated spirit. That evaporates very quickly and helps keep the aluminium cool. I would also periodically check the temperature of the aluminium with my laser temp measure.

So I think in future I’m going to definitely spot mark the smaller holes, and drill those on the pillar drill, and only use the single flute cutter for larger or more unusual shapes which I can’t do by hand. I also think I’m going to simplify the design of the pivoting plate and remove the fine adjustment mechanism. The fact of the matter is most hobby machines don’t have an adjustment for the tilt along this axis. My machine with its pivot holes already means if I were to require tramming the CNC machine I wouldn’t have to take it apart. And that’s already a massive design improvement.

So it definitely looks a lot more square than the first plates. I’m going to go clean this up now. Ok so I just put the Bottom Z plate on, and it feels really good. I can tell just by running my finger along the size and the extrusion and it feels quite square. The holes were a little bit larger at the top. I think I needed to tone down the toolpath but considering everything, what I’ve just had to go through this is a great success.

Anyway that was a bit of a mental video. I’ve learnt a lot about working with aluminium on a CNC machine and I really can’t wait to cut out the new plates and rebuild this machine with it’s new features and so that it’s a lot more rigid. Anyway don’t forget to sacrifice a thumb and let me know what you think in the comments section below and you’ll see me in the next one. Thanks again.