Hello, As I am fighting with my Workbee since lie 1.5 year to finally be able to do something precise enough, after a lot of tests, deceptions, retry, ... I had finally succeed to have a precision about 0.25mm for the squares and it was ok for me! But now I am fighting against the circles :-/ So, with Fusion 360, I have just create a circle with a diameter of 20mm and a depth of 15mm. To avoid issues, I have use the "Multiple Depths" command set to 2mm. I have also tried to tried a stock to leave of 0.5mm and the a finish pass for it. But even like this I have 19.6mm in place of 20mm of diameter. By continuing searching again and again, I have finally found this post: Workbee accuracy - adjustment - calibration and he proposed a great idea that I didn't had tested yet: the "Sideways compensation" so by default it is set to Left (climb milling) so I have tested with the right (conventional milling). So I tested it full of hope and there... 20,4mm and the I cried a lot before posting it. how could it be possible that it is always 0.4mm of imprecision??? And also why too small when climbing and tooo big when conventional. And the most important how could I fix it before the suicided? Regards, Benjamin
I am sure there are better options than that man, we need you around (sell the machine rather!) 0.4mm is very tiny to detect (you could go over everything again, and see if theres anything that is not tight) and checkout How to calculate V-Slot® deflection - it could be that the cutting recipe at 2mm DOC is still just too much for a machine of that design (and thus deflecting) That is kind of the game of scaling up. A minimill with its 125x125mm workbed is much more rigid so can do micron precision. But a large machine is aimed more at large workpieces were a few decimals of a mm in terms of percentage of the larger job size, is acceptable. In the time since your original issues, we've released the LEAD1515 and you'll notice we used double X beams to strengthen it up to combat that deflection, so could be that you may have to either dial back the cutting forces more to cope, or consider salvaging the parts (electronics etc the more costly bits) and upgrade to a LEAD1515? OpenBuilds LEAD CNC Machine 1515 (60" x 60") Other suggestions - Get a BlackBox, I think you had a Duet? We only use the BlackBox, and its got beefy drivers that can help with that side of things - Is you machine Belt drive? If not, the Leadscrews are a good upgrade option too - Press and pull on your frame, look for weak spots, and either consider the 1515 style of machine, or engineer some additional bracing of your own onto your machine - a Once more check for things that are loose - Where is your endmills from? Calipered them out yet? Are they what they say they are (Cheap endmills tend to be under or oversized)
Thank you for your responses! First, some precisions: - I have the Leadscrews version - The mill I have a good quality 6mm and a low quality 5mm but I have measure it and entered the exact dimension. I will re-execute a test with both tomorrow to be sure that they are no differences. About the precision of the measure I am taking my measure multiple times at different places with a digital caliper an also I am comparing with my 20mm drill bit so I can compare the size. The thing that I have difficulties to understand about my precision issue is the reputability of my results. When I execute the same program twice I have the exact same result with like 0.1mm of precision that is the good news but also the reason why I don't understand it. Tell me if I am wrong but if something is moving or having loose it shouldn't be so regular between two run right? Also the chances that the loose is exactly the same on X and Y is very low doesn't it? About the Lead model, tell me if I am wrong but on the X axe I see that they are huge difference in term of structure but on the Y level it looks exactly the same. The only difference I see on this axe is the system to fix the LeadScrews is a bit different and this part is probably helpful to have a better tension on the leadscrew if it can help. So I am correct I didn't expect that it will change a lot on my precision for my Y axe. (X ad Z axes will well more rigid) Regards,
Hi again, I have re-executed (again) some tests. I have done 3 circles : 10mm, 20mm, 30mm First test, I have just use the 2d pocket tool. Results: 9.6, 19.4, 29.3 but the measure where with an imprecision of 0.2mm depending of the diagonal I was using. So it gave me an idea. Those error could be due to the shape generate during the machining who is a snail shape. So, I have added a 2d contour who is a perfect circle and use the conventional climbing for it. and now the results are more regular in term of diagonals. Only my measures are 9.6, 19.6, 29.6 so always the same imprecision of 0.4mm witch looks a very good news for me. Now, I would like to know if they are a parameter to fix this radius issue maybe with the compensation type? But I am not sure that the duet support it and if it is how do I need to configure it in fusion? REgards,
Measured spindle runout with a dial indicator yet? Inconsistencies that stay the same, and does not change based on dimensions of the part, can only be at the tool end - every other calibration error would scale (0.4mm on a 30mm circle, would be 0.8mm on a 60mm for example) So as mentioned earlier, endmill being the wrong size, but maybe theres enough runout that the endmill is cutting larger than it should (because its spinning offcenter) - suspected you'd know if it was that bad, but perhaps, not noticed because of not having a reference of what normal sounds like? For example router knocked a bit in shipping and the shaft bent a little, bad collet nut, etc
Are you using calipers to measure the internal diameters? They will usually read a smaller diameter - especially with small pockets because the edges of the blades are ground flat. Better to use a known diameter piece of rod that's been checked with a micrometer and inserting that in the hole. If the machine has been calibrated precisely, then do as Peter suggested and check the mill diameter. All you need to do, is change the tool bit diameter in the CAM software (before toolpathing) - in small increments, until you get the desired size. For example, if you are using a (supposed) 3mm endmill, but the holes are coming out too small, then change the tool size before creating the toolpath - say 2.9mm, and see how it goes till you get it right.
Oh last minute thought, did you check tramming? A angled endmill will also look like an endmill of a different size
Sorry I am still a newbie in the cnc world and my english isn't so good so I didn't understood your latest message. Here is 2 pictures of the endmill I am using. Sorry for the quality but you can see the shape of it. What is the tramming and how to check it? About the center or the axe that could me not centered tell me if I am wrong but if the mill is not centered knowing that I am doing a pocket and not a contour, it should result in a bigger shape and not a smaller one doesn't it? My dad have the same router as I have so I will ask him to switch with me to test it also but I will need to wait to see him with those corona stuffs.
Sorry about the video being for a different machine, but Winston is a very good tutor and gets the points across so well:
Thanks a lot for the video I understand the principle! but I will take a bit of time to do the same now I will come back here with my results asap !
Well, I never succeeded to make it perfectly no I have mainly learn to live with it. But indeed the video helped to make my router as squared as possible. And I have rechecked all the wells adjustment that have helped a bit too.
