I've been unsuccessfully attempting to cut aluminum on my new MiniMill. My setup is MiniMill with Makita RT0701C and 269oz in Nema 23s. I am using a 3/32 in carbide end mill from Kodiak tools. The part cuts perfectly in foam and 1/8" plywood. I am hoping to cut the final part out of 1/8" aluminum . So far what I have on hand is 1/8" 2024 T3 (I think - it's leftover from a plane project and I can't find any markings), but I have some 6061 T651 on the way. I am using Fusion 360 with the generic Grbl post processor. Here is the part: I am using pocket cuts for the holes, and a contour cut for the outside. I've tried both climbing cuts, and conventional (had better luck with climbing). One of the attempts (climb) forced the bit down from the collet some. The other attempt (conventional) just stopped. Both attempts ended with the first hole (the large one). I'm using the minimum speed on the Makita (~9000 rpm) and I think I used a feed rate of 800mm/min on the first attemp, but lowered it to 254mm/min on the second attempt (after consulting a feed/speed calculator). I suspect I was trying to mill too deep with each pass (0.5mm). But I have a couple of questions: 1. My bit is ~2.68mm diameter and I had used a stepover of ~2.26. Does that matter on spiral pocket cuts? (I may try again this morning with 0.6mm stepover and 0.2mm depth/pass. However, that increases the machining time on this simple part to almost 1.5 hours - that doesn't seem right!) 2. Am I correct that a climbing cut is better? (My endmill is 2-flute.) 3. Am I correct that a "lowest" spindle speed is better? 4. When cutting a "slot" in material (which is the case for the contour cut), does climb vs conventional really matter? I haven't made it to the contour in aluminum, but climbing worked fine in the plywood. Any answers or other suggestions will be much appreciated. Thanks, Scotty
might be too large, depends on your depth of cut per tooth. remember the chip has to have somewhere to go in the cutter flute, too large a cut and it gets jammed. always always make sure the chips get out of the way! recutting chips is very bad. also, aluminum tends to stick to the tool, which is why it needs a cutting fluid. kerosene is good, WD40 works too. yes, but it does depend on the machine and material too. no! spindle speed is calculated from the SFM for the material, and the kind of cutter. The manufacturer of your cutter will have numbers for you. Here are the formulas https://www.kennametal.com/content/...owledge/TechTip_030_CuttingToolFormulas-1.pdf depends how stiff the machine is, but it is usually better to cut the slot with a bit that is smaller than the final slot. then you can use techniques that are easy on the machine.... '2d adaptive' and so on. also note that you should avoid plunging straight down, endmills are not good at that. Rather ramp or spiral into the cut.
I can not get slotting to work reliably with aluminum either. I have done 1/2" 6061 with trochoidal milling using Estlcam in 3 passes (1/8" bit), IMO it's much easier to do than with Fusion. Estlcam Trochoidal Milling Video
Thanks. I'm still having some trouble that I don't understand. When attempting to cut aluminum, it appears to work OK for a little bit, then the motion just halts. The controller (Gradus M1 Pro) will not respond to any further commands, but the board doesn't appear to show an error (with the leds). could be that it did for an instant, then wouldn't continue. I put an extra fan on the drivers just in case it was a thermal issue. After that, I thought: "well if it's an overload thing, I should be able to duplicate it by trying to jog beyond the end of an axis". I tried that and it happily just banged and skipped away, but the controller didn't care and didn't "halt". It seems to me that if I were overloading while milling, it would just break the bit. Granted, jogging is just a single command. So. I'm curious if there is something in grbl, where if there is a long string of commands and the machine "bogs down", it just quits sending and puts the board in an error state. BTW, after all of this, I tried another cut in plywood, and it worked fine. Any thoughts?
what version of GRBL is that board running? what USB chip has the board got? The CH340 has known problems. There is a protocol for talking to GRBL that is supposed to prevent the buffer overflowing. What GUI are you using, and does it support the latest protocol if the board is running GRBL 1.1? I have been using the UGS nightly build and have had similar freezups. But I thought it is my Nextbooks USB port playing up rather than an actual GRBL problem.
I'm running GRBL 1.1f and Grbl-Panel 1.0.9.15. The USB chip is the ADUM4160 (isolated USB - whatever that means). The Gradus M1 Pro came with GRBL 0.9j, but I couldn't get that to communicate with my computer, so I flashed 1.1 (but now that makes me wonder about my computer...).
ok, try different USB cables, they are not all created equal (-: in particular look for ones with a big lump at one end or even both ends. those lumps are filters that prevent electrical noise in the cable. also, shorter cables are better. cables supplied with cellphones are generally bad for this job. Another thing you can try is to slow down the communications. GRBL defaults to 115200 baud but if your cable/controller/computer has a problem then running slower might work better. you need to edit the config.h in the libraries/grbl folder , go right down to 9600 and test. Then increase, trying every step. Not all USB ports will support every speed so if some don't work do not be surprised. in fact 250000 baud might work better because it is a multiple of the clock frequency while other standard rates are not always good multiples of the clock frequency (which is why cables make a difference) Here are the normal rates to try 9600, 14400, 19200, 28800, 38400, 57600, 115200, 250000 if only a slow rate will work reliably, then your problem is most likely the cable or computer, but of course there may be noise injected on the Gradus board itself from poor design. high frequency electricity does some weird stuff that has to be circumvented by good design. have you searched the Gradus web site or forums for similar problems?
Good thought on the USB cable. I had to get one with the ferrite chokes for my 3D printer. I will give that a shot first before I try different baud rates. There is woefully little info on the Gradus M1 Pro out there (I've looked). I've started documenting stuff I've discovered or found on another thread here. I will add this to the list (once I have determined the culprit). Thanks for the suggestions!
David - thought I'd give you an update. Turns out, it was probably my aging laptop. I was just getting ready to test with the new USB cable, and my laptop completely died. I'm not really upset, because I'm not really a PC person anyway. I was just using it because it had been laying around not getting any use and it was all I had that would run grbl-panel. I prefer Mac, and wasn't aware of UGS when I started this adventure. Looks like UGS will work fine on OSX, so I will install that on my Mac laptop and see how it goes. Fortunately, the PC had the good grace to die before I ruined another piece of aluminum. Thanks again for your suggestions. Anything I should look out for when running UGS?
Another update: UGS "Platform" nightly update running on a Mac with Java 8. So different computer and OS, new USB cable with chokes at both ends, new bit, new aluminum, new flat spoil board, all fans running. Well, you can guess.....it froze again. At this point I would suspect the Gradus - but since you had something similar happen, I'm not so sure. I'm gonna do some research, and try the lower baud rate, but I also might try downgrading to an older grbl. David, what board did you have your freeze-ups on?
whats the max character length set to in UGS? are the lines in your code reaching 50+ characters including spaces
Max command length is set to 70 - and no, no lines are even nearly that long. I think I may have narrowed down my problem though - I think it's EMI interference.
David, I think "electrical noise" is getting to the heart of the matter. I've been doing a good bit of research/reading on the general problem of random stopping/freezing and I heavily suspect I have an EMI interference problem. The Makita spindle is not grounded, plus I had all of my cables bundled together and on the same power strip. From what I've read, this is just asking for trouble. I am going to try a few things: Add an earth ground to the Makita, and to the mill chassis itself. Plug the powersupply and computer into a different strip and/or replace the strip with a filtered battery backup. Anyway this will take me a couple of days, but I will report back with results.
Hiya google 'star grounding' and do it . in short, everything must be grounded to one place, (separate power strips does not do that!) anything needing grounding must have its own ground wire to the star point. also, move signal wires away from power wires. signals are limit switches and step/direction type stuff, power is AC power, DC from the PSU to the drivers, and from the drivers to the steppers. the usual noise source is the router, esp one with a speed control. run a long job with it off and see what happens?
yes that's a big one for sure, it's very easy to replace your router cable with a shielded pair cable, run the drain to the star ground, for the Makita, it only needs to be 18awg. And that fits in the cable chain easier too.
Well, finally - a successful cut in aluminum! I can report that the grounding (especially the Makita router) seems to have solved the problem (knock on aluminum...). Not sure if what I did qualifies as "star grounding", but from a couple of diagrams I looked at, it's close. Everything is grounded to a single surge strip, which is grounded to earth through my house wiring. I didn't use shielded cable on the router yet (I will, but I wanted to try a quicker method to test the idea first.) I used 18 gauge wire and grounded the the metal case (from the outside of the case). The other end I attached to the ground pin of a DIY 3 prong plug - then plugged it into the surge strip. Ditto for a ground to the minimill frame. I replaced my original plug strip with a surge strip that has some EMI filtering built in. I moved the USB cable away from all other wires (as much as possible). I also decided to try the cut without WD-40 this time as it seemed to me last time it tended to gunk up the works. My dust shoe and vacuum did a good job of keeping the chips clear. In the photos below, you may notice some "rough" spots - especially the center hole. This was a scrap from a previous attempt that stopped before the first pass finished the center hole and I didn't get it set back exactly right. Anyway, I am very pleased! I appreciate all of the suggestions everyone had for me. Thanks! BTW, the part is a stepper mount for my 3D printer, which has been acting up lately. Hopefully this mod will help solve its problems! (Then I can print version 2 of my dust shoe.....)
