July 25, 2017 at 3:35 am #39277
Trochoidal gives you a larger slot for the chips to evacuate.July 25, 2017 at 11:12 am #39306
So you got it working perfectly, impressed the heck out of all of us, then tried to double the depth? All at similar speeds and step over? Killing me smalls.
You are allowed to take multiple passes.
I think the previous video is seriously pushing the maximum for a machine the size of yours. You had amazing results, stick with them and see how it goes. Don’t forget the deeper it goes the less rigid it is on the MPCNC, the lowrider is opposite.
A few things to consider. I would try the single flute again. The more flutes you have in a deep cut the less time you have for a chip to get out of the way of the next blade. A deeper cut also should be run a little slower for the same reason. Air could help but slowing down the feedrate or the spindle rpm will help as well.
Results at 2mm will not be the same as results at 8mm, CAM is a tricky tricky thing.July 25, 2017 at 1:12 pm #39313
I pretty much gave up plunging. I tried pecking .5mm at a time, and on the last depth, it went really nasty. Made this low rumble but the bit didn’t break. I will just half predrill the holes then finish them on my hand drill. No biggie there.
The test was done on a single flute. Only now I only have 1 left… A 3mm slot 9mm deep doesn’t allow for any chip evacuation at all. I will remedy this by doing the wider deeper cuts. Technically this wastes more material but at this point I just want to finish the part without something going really bad.
True about the rigidity, but my z is so short as it is I don’t think it will make a difference. There is a diminishing return line like I said. At one point all the flex is not the z axis turning, it is the linear flex of the rails, zip ties, etc.
I got teased into my success lol. I think my cam was great for thin parts but I got to change it up now that I am trying to go all the way through 3/8″. The further the bit goes into aluminum the more scared I get cause it’s hell down there D:
Still trying to fix the awkward chatter points. Picture a circular cut, the bit chatters at 12 o clock and 6 o clock. At 3 and 9, there is no chatter at all and the harmonics are great. I tried to test if it was the fact that my x axis conduit was twice as long by test cutting at the very corner of my machine. No luck. 🙁 This is by far the problem I am concerned with the most. It’s weird because even on light finish passes, at 12 and 6, this low rumble occurs. It get’s so bad you can actually see the bit vibrating back and forth. Yikes. I tried taking off my shims and it had no effect. I am able to run really aggressive settings with no chatter except when it hits 12 and 6. I don’t get this at all…
I am going to keep testing and trying until finally I get a finished part. 🙂July 25, 2017 at 1:20 pm #39315
Easy, there you go blaming zipties again.
Truly it is a CAM programming thing. This is people full time jobs that they go to a lot of schooling for and get paid insane amounts of money. You cut wonderfully and then more than doubled the load on the entire machine and are wondering why it isn’t working. I can barely cut 9mm of soft pine at once, I made a video to prove it.
Have a look through my previous post again, I gave you a full breakdown.
As for plunging, always peck if you can’t helical cut, never ever straight plunge unless you are using a drill bit. You can’t just jam a bit made to cut on its side faces into material and hope it works.July 25, 2017 at 1:34 pm #39316
Wait, by double load do you mean the thickness of the workpiece? Maybe I wasen’t clear, but I didn’t ever try 9mm doc lol. That would be an instant disaster. I see load as the rate at which it is removing material combined with HOW you are doing it. I actually reduced my DOC and speeds a lot. In fact, in one test of my now butchered 3/8 inch plate, I reduced my doc to 2mm and my feed to something like 400mm/m with a .5mm radial cut. I am talking about adaptive clearing, not slots. It sounded better for sure, yeah. But that chatter at 12 and 6 was still there only lighter. So in short, I did go more conservative.
So let me do another test right now. It takes something like 10 minutes. I will be really conservative this time.July 25, 2017 at 1:40 pm #39317
If you go too light you can work harden aluminum.
There is a sweet spot, why not use the exact settings you had before?July 25, 2017 at 1:53 pm #39324
My very first 3/8 part test were my exact settings. The dimensions came out good, as expected. Finish was great. I was feeling really good like, hell yeah here we go! 😀 But this was before I realized I cant just plunge into the material. I did peck it but too deep, 3mm each peck. So hard to watch. I should of stopped it but I didn’t so on the 4th and last hole it just snapped off 🙁 I need to order more bits I keep ruining them. I. am on my last single flute. 4 dollars adds up fast for me…
I know I have said this 3 times already, but those weird chatter spots are really throwing me off. Like I said on the first test with my what I thought were perfected settings, ended up chattering super bad on 12 ad 6 areas. On my second test I went more conservative on doc thinking it would help and it actually ended up breaking on the finish pass. ConfusingJuly 25, 2017 at 2:17 pm #39335
Check for perpendicularity, sounds like you have an angle, in or out from 12&6. You almost never notice them until you get to metal.July 25, 2017 at 2:54 pm #39338
I did try shimming with tape and I got really close along the Y. That’s why it didn’t make any sense to me. I just tried taking them off and it made little to no difference. But if we shim the dewalt mount, aren’t we just putting a bandage over the issue? Cause if I shim it and get it accurate, great. But won’t the z axis still not be perpendicular? Maybe this is coming out wrong.
So when we are shimming the mount, are we trying to fix the error in the mount or the entire z axis? Cause then if you shim it, your bit will be perpendicular, but your z will still feed at an angle.July 25, 2017 at 2:57 pm #39339
If your z axis is not perpendicular you have to fix it, first.
Tape is just for the little tiny bits that are from the shape of the dewalt case.July 25, 2017 at 3:08 pm #39342
Ok so I should adjust the tension bolts in the gantry. I have done it before but if it comes to taking it out and reassembling I will. My parts already have the break in thing, my xy squareness is spot on. I only can get at most 4cm of z pipe coming down, so the square is hard to see.July 25, 2017 at 3:17 pm #39343
If you take the dewalt mount off you should be able to get a square in there pretty easily.
Or you can just figure out your cam. I feel like you want it to be something else. You can get things progressively more accurate but with your previous video I am telling you your machine is spectacular. Just work on your CAM program.
You are doing this the hard way and going hardcore into 3/8″ aluminum, most of us have spent 99% of our time in wood, then some plastic and a bit of aluminum or brass, you are insisting on pushing the limits in the beginning. There are so many subtleties to learn. You seem to be frustrating yourself. Maybe you can step back and make some things out of wood, or even test your gcode in wood, then acrylic, then metal.
The MPCNC is going to take more precise CAM than a machine that has a few more zeros in the price tag, but honestly, not by much. Add a few more zeros still and then yes aluminum gets a bit easier. If we could do steel you would see it get exponentially harder still. Any time I have been around large machines, the program gets checked by another programmer before they try and run it if it is steel.
I have not run the same program twice yet. I run it, watch, take notes (seriously), tweak the CAM, Run the next job. So many sets of lowrider flat parts and I am still changing it every single time. And that is .25″ wood.July 25, 2017 at 3:34 pm #39345
I guess the only reason I am so concerned about it is that I know it is the little things that are taking away from really really good results. I have done really aggressive stuff with no issue at all until very specific points. I feel like if I can little by little improve things, that I can get the results I want. Honestly although it seems I am pushing it to the limits, I have done a lot of much lighter/slower cuts with the same fundamental issues.
Upon inspection my z is indeed off, in the Y direction. I am dissassembling right now to tension the bearings. One of them wasen’t gripping the conduit, bad sign.
I am switching to estl cam.. Seems much more efficient for deeper cuts. I really like f360 but it doesn’t have the trochoidal milling I like. Adaptive clear only works for pockets. I am really close to getting a good test part out, I will post pictures when I get there.
I made this specifically for aluminum in mind. I knew going in I would have issues. I don’t want this to become a life story but, I am only in high school. Can’t get a job quite yet so my budget was low. The MPCNC was an attractive option. Even though I sound frustrated, I am having a lot of fun and would much rather mess with CAM then paying the extra 0 to make it more forgiving.July 25, 2017 at 3:39 pm #39346
Okay, we are on the same page then. I just wanted to make sure the tweaking part was the fun part for you.
I just bought a big laser and I have spent 10 minutes cutting and etching and about 4 full days making changes. More on this in a different thread. So I like to get the most out of my machines as well.July 25, 2017 at 3:41 pm #39347
And…For a high schooler you are killing it, you have a bright future for sure. Messing with a cnc at that age…If you were local I would offer you a job.July 25, 2017 at 3:58 pm #39348
Thanks 🙂 Yeah if I was in norcal I could bag the hardware in the kits for minimum wage! That actually sounds quite fun to me.
I just want to fix the little things. That video was really just a showcase, it wasen’t even properly edited or filmed. I don’t really care how long the part takes, I just want it accurate and most importantly, reliable. Included a pic of the part incase you’re curious. It is a motor mount system for an electric longboard conversion.
I wanted to do a small production run of 20 and sell them. If I could get one made per hour I would be happy, considering each one sells for about $40. Watched the interview you did which got me getting a little interested in making some quick cha-ching.
Attachments:July 25, 2017 at 4:05 pm #39351
My college senior project was an electric skateboard, wirelessly control by muscle impulses. flexing your forearm made it go, like a small kick but smoother, regen braking all of the bells and whistles.
https://endless-sphere.com/forums/viewtopic.php?f=35&t=62689July 25, 2017 at 4:28 pm #39356
Looks a lot more fancy than mine. If I ever made a lowrider I would want to carve out a new board.
I am visiting sdsu tomorrow, what are the odds. If only the mpcnc could make me perfect grades. Where are the toolpaths for the SAT?July 25, 2017 at 4:34 pm #39360
I spent a lot of time in the machine shop downstairs. I think they just got a new engineering building though…July 26, 2017 at 7:49 pm #39490
Okay, got everything mechanical fixed today. Spent hours with the wrenches messing with tension bolt C. Getting the z axis perpendicular was sort of a lost cause, so I just shimmed the hell out of the mount and it seemed to do the trick. Not ideal but as far as I can tell everything works great.
Besides making the bit actually perpendicular, I added another 1’x2′ particle board sheet to my bed, making it around half an inch taller. My z is now as short as it can be, with z pipes flush against the gantry.
All the chatter is gone, now it sounds like cutting butter again, yay!
How do I turn off estlcam tool changes? Everything just pauses forever in the middle of the toolpath. I put 2 different tools in my tool list, but they’re the same “tool” just different settings for different operations. One is the contour passes where I use trochoidal, the other is a pocket setting where I use parallel pocketing.July 26, 2017 at 8:22 pm #39495
Just delete everything in the tool change tab. I was looking at it as well, I prefer to do my helical cuts real slow when they are small and the tool change thing would have been irritating. I ran mine for 2 hours straight today, lots of lowrider parts.
Can’t wait to see your next video!July 27, 2017 at 5:01 pm #39573
Aaaah dang. It keeps cutting out in the middle of a job. It’s going really well but it’s not finishing! grr.
Hope the ramps didn’t die, everything looks good. Thought my drivers are overheating but I just added a fan and nothing. Repetier starts freezing and everything just stops. It doesn’t respond to commands anymore and I have to force reset. Could be the laptop I am using since I never had any issues prior to this when I had it next to my desktop.July 27, 2017 at 6:34 pm #39576
Well it seemed to stop, for whatever reason. Had some success today though. Finally made through 1 tool path without any troubles. I really need some compressed air though. The total toolpath was about 30min, and vacuuming chips for half an hour gets old. Still though, 3/8 aluminum plate is tough and not easy to go through especially with a $400 machine.
I ran out of good single flutes, so this was with a sharp dual flute. I like single flute more but it is far better than using a dull single. Tolerances check out, except for that estlcam doesn’t seem to have an oversize finish pass option, so the part comes out the exact dimensions in the CAD file. This is good and bad, just means I have to oversize on my dxf drawing.
Later tapped the holes M4, and did the M6 set screws that go along the sides.
Attachments:July 27, 2017 at 7:13 pm #39582
Looks really nice! Just FYI most cnc’s don’t have a ‘finish oversize’ option. When designers are making the CAD they will typically design with tolerances in mind. The software doesnt know what dimensions you are OK with being over or undersized. Sometimes its even left up to the machinist at the machine to determine what offsets he wants to set on OD and ID cuts based on how his machine is cutting that day and the tools being used. Just keep it simple for now. Remember you can always take more off, but its a bitch to put it back on….if you can.July 27, 2017 at 7:39 pm #39585
Great result, bravo!
30 minutes seems a bit long for such a small part, what parameters did you use?
About your problems of machine stopping, instead of using repetier, I suggest you to use the SD card directly. Using the SD card will rule out any computer related issues.July 27, 2017 at 7:41 pm #39586
Thanks, and I will keep that in mind. Normally on fusion 360’s CAM, I do a contour finish pass where under “Stock to Leave”, I put a negative value which actually does oversize on purpose. It’s a cnc router so tolerances are pretty loose. I am getting within .1mm which is great for my purposes. Very true about oversizing, it’s so easy to file fit vs trying to…idk shim the fit with something else.July 27, 2017 at 7:51 pm #39588
SD card sounds like a good idea. 30 minutes is mostly due to the fact that my toolpaths are trochoidal. The advantage to this is that I can do really deep cuts without the chips getting burried in a 3mm wide slot. The disadvantage is that I actually have to remove more material than “necessary” to create the cutout geometry.
7% trochoidal step
70% trochoidal width
900mm/m (This can be faster but I if I go faster, I get weird interruptions, probably the arduino usb bottleneck) Son of a diddly!
Really need an idea for a cheap air blast. I literally wouldn’t care how long the toolpath is as long as I don’t have to watch it with a vacuum. To think most of my previous 3d prints took WAY longer than 30 minutes, this seems fine. I care more about repeatability since I want to do a production run of at least 20 of these. Knowing I can hit the green button then do other things would be awesome.July 27, 2017 at 8:02 pm #39590
Well, that is still quite an aggressive setting, I’m impressed.
70% width is quite a lot, but if it works fine, then go for it. During my tests I’ve set it up to 50% or less.
Do you have a workshop air compressor? You could easily hook it up to the cnc, then just add a simple valve to set up the air flow, and a flexible tube to get it where you need. You could even add a solenoid valve to start/stop it whenever you want.
If you don’t have an air compressor, well… buy one! It is one of the most useful tools around the shop, a must have! Plus it doesn’t cost too much, a small 30 liters one should be enough.July 27, 2017 at 8:04 pm #39591
About the speed, what you can do is to lower the microstep setting. I suppose you are using the 1/32 step mode, so try 1/16 or even 1/8. That’s what I use on my 3040, it gives much more room for speed and doesn’t seem to have an impact on accuracy. Could worth a try.July 27, 2017 at 8:14 pm #39593
That and a drill press are clearly missing from my makeshift dumpy garage. I don’t really have any income…so this may get kind of ghetto rigged. If I could still print something I would try to get like 4 12v high static pressure(and noise) fans and then combine their air into a single stream.
Doubt it will have enough pressure though. Just wondering if buying all these “upgrades” is worth it.
I also owe some money in ebay fees…don’t tell them I was here! 🙂
Good idea on 1/16th stepping, guess I have to change my jumpers. If there really is no change in accuracy I don’t see why not.
I am running outa power in the garage here. Not that many outlets, and I have an electric car charger that’s 1400W on the same breaker.
You must be logged in to reply to this topic.