So by ‘pluck’, do you mean audibly hear a pluck?
Yeah. It isn’t too critical. The forces are small. Just so it isn’t loose, it is probably fine.
Thanks… my belts are nowhere near “audible” so I guess they are “loose” I was worried about overtightening them. It would be nice to have some quasi-objective measure and process of how to determine if you are in the proper range of tension or not. Maybe something like: “With gantry 4 feet from end stop, with one finger in the middle of the belt (2-feet from gantry and end stop), you should be able to offset the belt ~ x inches from resting point” (though I realize this is flawed as well, just using it as an example.)
edit: sorry if my propensity to use scare quotes bothers people… I don’t mean it negatively…
1 Like
I’m curious about this possibility.
I assembled my Lowrider such that the Z tubes are very close to the table. Less than 1mm of play, I think. Wheels slipping is not reallly possible. If there is a recommendation regarding this dimension in the assembly instructions I haven’t found it (I have looked…).
I did sand the edge of the table smooth, and I have had no problems as a result of this arrangement… When I shut the machine down and the gantry slowly “falls” one side may hang up a little as a result, but a little poke will get it to fall all the way.
How close SHOULD the Z tubes be to the table?
They should not actually touch, and it shouldn’t matter. From an engineering stand point the closer the better to keep the rails as short as possible.
You really just need to make sure the front of your Y plates and back are equidistant so they drive straight. If they do not drive straight add a guide to the wheels of some sort.
1 Like
Things are much better now, not perfect, but not bad either. I’ve tensioned all the belts (quite a bit), checked the grub screws (all seemed tight but I broke them loose and retightened), installed a set of rails along the side of the wheels and adjusted the position of one of the back tube to get the back wheel of the front Y-plate better in line with the front wheel… It seemed off by a little… it was easier to see when the rails were added.
I ran an outside profile of 7-inch by 7-inch square (8 mm/s feed rate, 2.5 mm/s z axis, 3.2 mm doc) and did just two passes with 1/8 inch single flute bit. Both passes seemed to overlap each other well, but the size is still off (not by much though) and there’s still something going on with the initial plunge…
If you look, you’ll see the initial plunge is lower ( lower in the sense of the Y axis ) than that of the path of the router bit as it returns to that spot after doing cutting the square shape. The 7-inch square came out about 6-15/16 inch tall and 6- 31/32 inch wide. I suspect that if the bit had returned precisely at the plunge point, the square would have been a little taller making it closer to 7-inches. I checked the gcode and there’s no lead-in or anything in it. It starts at X188.433 Y6.033 and ends there.
Also, I used the default settings I found in V1CNC branch… maybe its not appropriate for the mini rambo?
I can deal with this (perfect is the enemy of good), I’ve dealt with much worse with other CNCs, but curious if anyone has any ideas as better is always better.
Are you sure your machine is parked square before you start the cut? You can check this by measuring the diagonals.
Another thing you should be doing in the beginning is driving over the entire cut before making it. The easiest way to do this is outline your part in CAD or CAM and do a pass in the air as the first tool path.
The dims seems okay, did you do a finishing pass?, but the plunge anomaly is not okay.
If you are slotting, the rotation of the tool will push the tool slightly left relative to the direction it is moving. This can cause dimensional errors depending on how fast and deep you are going and how hard the material is.
If the inside dimension of the hole is also small, and if a pen drawing is the correct dimension, then I would suspect it is deflecting. (If the inside dimension of the hole is not small, then it is more likely tool runout).
If you have deflection then the traditional remedy is a finishing pass. The first cut leaves enough material that the part is guaranteed not undersize, and the finishing pass has very light loads so it maintains precision.
1 Like
I squared the machine when I it had loose belts and misaligned wheels, no rails, etc. I have not checked yet to make sure its still square… have to get the pen reinstalled.
I went back out and redid the measurements, this time using lining up the 1-inch mark as the origin rather than using the tab… From what it looks like both the X and Y are 6-31/32 inch… not sure why I didn’t get better measurements first time around, sorry, but it’s very close. The little drop of the plunge is 1/32 inch off from the bottom of the router path. Both diagonals measure the same so it’s nice and square.
I’m not sure how flying the path will help with this… Are you suggesting doing it in general or specifically to address this issue?
No finishing path. All I did was two passes, 1/8 inch each pass just to see if the they lined up. I use easel for the time being for my CAM and its not very sophisticated. I’m looking at estlcam, but I do CAM on three different machines depending upon where I am and three licenses are not attractive.
Yeah, its an oddity.
1 Like
In the beginning this makes sure you are not going to hit anything including the sides of the table, clamps.
This also means your machine is square.
If you are measuring things to the 32nd you need to use a full depth finishing pass to offset cutting loads. This is the only way to get accuracy. Others call it a spring pass.
It could also be a wonky plunge, ramp, too fast, a lot can explain this if your job came out reasonably accurate and square. Much worse if it cut into your project. I can’t say for sure what any other program does on a plunge other that estlcam.
1 Like
Yeah, I like estlcam because it can do all that… maybe I can set up a server somewhere and just use remote desktop to run it from wherever I am.
Easel doesn’t do anything fancy. It just plunges at whatever z-axis plunge rate. Once I redesign my piece (third time is the charm), I’m going to try to use estlcam to generate the gcode and include a finishing pass.
That could explain it… 1/64th on an inch perhaps? each side yields 1/32th inch total and there you go.
Oh, that would never happen to me! … looks around nervously …
1 Like
Me hides the cut in half hold downs…
Cnc.js will show you the overall dimensions and the preview shows you the bit relative to the pattern, so that is pretty useful for checking against hold downs. With my Low Rider, I’ve used exclusively screws, or screws and some long 1/4"x2"x12"ish strips with a screw down the middle.
I also discovered that my lowrider was a true lowrider… the back end of the 611 plate sits about 1/8 inch lower than frontend. Turns out that the back tube has a bit of a dip in it. I turned the tube to get the bend upward (or so it seemed) and the tube seems to sit level with the sled in the middle… Progress.
I did some test holes (6.35 and 5 mm) today and found adding -0.2 mm finishing allowance, with no tool to a 6.35 mm hole got me very close and I could slide a 1/4 inch bolt through. A 5.5 mm hole also needed -0.2 mm to come out close…
Re: The plunge… when I do paths on fusion there is a ‘lead in’ setting so it plunges slightly outside (or inside) the cut and moves to the actually cut line.
I tried that with Estlcam but it looks like you can’t do a lead-in and tabs at the same time. I found a post on some forum that I think was from the author that said the reason for doing a lead-in was to avoid plunging into the cut-line and tabs do exactly that, plunge into the cut line… so why would you do both? 
You can adjust the plunge angle so it travels along the path while pluning. You can do that with tabs too.
1 Like
Unless I’m doing things wrong… I click on lead-in and set it then click tabs to set the tabs and my lead-in converts to a tab… I no longer have a lead-in.