This one is promising, but I haven't tried to cut Aluminum with it yet: www.amazon.com/Machine-4040-XE-Spindle-Engraving-Engravable/dp/B08M96HHKW/ref=sr_1_1?dchild=1&keywords=cnc+router+4040&qid=1605163511&sr=8-1 I will be making a serious video on that CNC machine in the near future. At that price piont I think your only other option is to DIY something.
@@ClintonCaraway-CNC I went with sainsmart 3018 to test the waters. I'm stuck at resetting machine coordinates to zero. Been trying and researching for the past few days with no avail. getting frustrated for sure.
@@pentachronic I'm in a few of them. I'm having issue with machine zero coordinates. Got everything else. Seems like I'm missing the software side of it. Playing with Candle atm.
It was cutting metal parts that were good enough to work but needed a couple changes to cut better (not more parts). I've been cutting metal on my machine for 10 years (it's a couple tiers up from that machine) *Tips for anyone trying to cut ALU or Brass:* 1. *Suck the bit up* as far in the collet as you can (not on the flutes, but as close as you can). It makes a big difference. 2. *Deep cutter-width slots are a no-no!* Set your profile cuts at least to at least *1.5x cutter width* (also helps with clearing chips). It will take twice as many passes, but your bit will thank you and your surface finish will be much better. 3. profile cutting: if you can, precut/drill a hole in the middle and make sure your path starts from the hole. You will be able to take deeper cuts as it is *side-cutting instead of slotting* (As soon as saw your deep narrow slot, I knew immediately it would break a bits and/or lose steps.) 4. If cutting dry/air only: *Use WD40* Frequent squirts cool the bit and give it a coating that resists chips sticking. 5. *Buy better/midrange cutters* $5-8 cutters make your machine cut so much better than $1-3 cheapies. (don't buy expensive cutters until you learn how not to break them) 6. *Cut chips not dust!* Dust = wearing out bits fast with almost no material removal. (small chips are ok, but grey dust isn't) 7. *Don't cut steel.* Just don't. Your machine will damage it's self even if you can't see the damage immediately. e.g. My motor shaft coupler started slowly fatigue cracking and then later snapped and caused a nasty crash that actually bent my Z ballscrew... sniffle. Purchasing tip: 8. Get a machine with the largest dia *ballscrews (not leadscrews)* and largest rails you can. The water cooled spindles are also much quieter. Upgrade tip: 9. *Flood coolant* with water/cutting fluid: if you plan to cut metal a lot and are willing to set it up, this *will massively increase your machine's cutting capability.* Seriously, I went from struggling to cut metal at all - to wondering how far I should push my cuts. Not only does it blast chips away and cool the bit, but it also reduces vibrations and noise! (straight water will rust everything including your cutter's crisp edge needed to cut well) I use a cheap fountain pump and water soluble cutting fluid. My machine sits on a tilted table with angle alu edges to keep the water in, then it all runs into a rubbermade tote with the pump in it. Linelock makes directing the flow much easier and isn't too expensive for a foot or so of it, but I started with a coat hanger wire ziptied to a piece of hose! PS. I'm not saying you should go back and try that machine again. It is just going to be on the painful side. But it was doing it until the too deep narrow slot started chattering too much. So for anyone else, who already has a machine and can't afford to just go buy another machine, I hope these tips help.
Videos like this - where things Don't work out - are way more truthful and helpful than polished and edited videos where everything works perfect the first time. Thank you, sir!
This video will save countless viewers from failure for aluminium routing projects - thanks for the time, effort and expense you went through to produce this video.
Actual machinist here: It may seem odd, but you can often remove chatter by increasing feed, or decreasing rpm. Keeping a high load on the cutter makes it unable to jump around. Adjusting feed/RPM while making a cut is a good way to find the right numbers. What the read in online calculators is just a starting number, you find the ones you need by adjusting and listening.
That's definitely true, and part of the problem with these little hobby machines is that they generally don't have enough muscle to give a carbide bit what it wants to eat.
That is true for a better setup, but this is likely due to a lack of rigidity and power that will not be overcome without losing your precision. Might work to a degree for roughing but I doubt it in this setup.
@@Skinflaps_Meatslapper I think people glorify carbide as an industry standard whan HSS is generally a better option for low power machines too. It's easier to get it to bite since has a finer edge usually.
@@brahtrumpwonbigly7309 It's surprising how dull most carbide inserts are when new, and that's part of the problem. You need more force to get the duller insert to cut, but since it's such a hard alloy the additional force doesn't speed up the dulling process like it would with HSS, it just ends up cutting better. Maybe the really high end carbide tools have a better edge, but those are usually out of my price range. I hit mine with a fine grit diamond wheel if possible right out of the box and less aggressive feed rates work a lot better, but they're still just a bit more susceptible to chatter than HSS is. Could be due to a more rigid tool and harmonics, I don't really know to be honest. The slop in a cheap machine plays a big part in that, so unless I need carbide, HSS does most of the cutting in my shop. It's also pretty handy because I can forge, shape, and sharpen HSS to whatever kind of cutting tool I want...carbide is way out of my league LOL
I'm used to larger, corded routers ruclips.net/user/postUgkxfQ5_mgwq6PcudJvAH25t-I4D-3cTPz4z so this was a different experience for me. Basically, this is an incredibly sweet machine, fits well in the hand, etc. It has slightly less power than I'm used to, but that's understandable. Maneuvering freehand takes some practice. These days, it's especially nice to be able to avoid power cords. This is above the level of a hobbyist but below a pro level.
Few points of interest here: 1) The terms most often used is Backlash, or play, in 22 years, I've never heard anyone simply abbreviate it to "lash" 2) judging from what's shown at @00:07:00 that anti-backlash nut is installed completely incorrectly. The spring is NOT supposed to be compressed that hard, nor is it supposed to be compressed against the frame, the way you "fixed" it completely negates it's use. The way these devices work, is they put tension between the nut, the follower, this ensures the nut is always engaged with the threads of the lead screw This image displays the proper installation of a spring loaded anti-backlash nut: images-na.ssl-images-amazon.com/images/I/71xI4rTFy4L.jpg 3) At the scale you're talking about, rigidity issues on the "cheap" versions has very little to do with the frame, and more to do with the hardened guide rods, and the fact that on the machine you're reviewing ,they are offset, and the machines you pointed out, they are in line. This will lead to "nod" and tram errors int he cheaper machines as they plunge into work, and chatter as they move through the work. However, again, with the scale of these machines, you shouldn't be putting sufficient cutting forces with any of them to bend rods, you're not working with VMC's, so trying to rooster tail 4041 with this isn't going to happen anyway. Feeds and speeds are where most of your issues are going to come into play. (yes, 12mm hardened rods will be stiffer) 4) Moment of inertia calculations: Wow.. talk about going a long way for literally no reason... Yes, the MoI for 8mm rods is 201, and 1017 for 12mm rods.. However, these numbers actually don't tell us anything practical about the rods, their material, their hardness, bending forces required, or how they relate to the application they are used here. That's like telling someone the tensile strength of nylon when talking about a wire brush.. If you're supplying enough force into the work with your cutter, that you're bending, or displacing the rods (2 hardened rods, and the lead screw) then you're obviously over-loading the machine, and operating it unsafely. again.. you're going out of spec for these machines to try to prove a point, that literally, doesn't need to be proven. yes.. 12mm rods, and a thicker gantry cross member will be more rigid, 100% accurate.. literally all you had to say... once.. 5) False equivalencies abound. You point out that the connections between two parts are not equally rigid.. um.. ok.. of coarse they aren't going to be equal, they are 1) different connection planes, 2) different connection points, 3) different components, 4) different profiles... To assume they would be the same is illogical at best... Anyone, can simply take a look, and see they are different in every aspect except for perhaps, the fasteners used.. .on one machine, you have a piece of extrusion, supported with two right-angle blocks, that are clamped onto the extrusion, and the frame (total of 4 connection points, on 4 faces) and on the other, you have 6 connection points, all on the same face... Of coarse they are different, they are different designs.. Also, comparing 10mm thick solid aluminum plate to extruded aluminum..... really... The plate will be less likely to tilt as you've described than the extruded aluminum, mainly due to the way they are clamped, and the fact that solid aluminum vs a hollow extrusion.. that 10mm of extra wideness on the extrusion probably isn't going to compensate as well as you might think. However, that being said, if you're putting enough lateral forces on either machine, you're not operating your machine safely, or properly. 6) It's not an "axel" it's a spindle. An axel is for wheels, regardless if it spins or not.. a spindle is any rod which spins within a mechanism. yea.. i'm not going to go any further into this review.. sorry my guy, but it appears you're trying to look like you know what doing, but there are a myriad of technical issues here, everything from terminology, to mechanical assembly, to the way you describe things shows this isn't something you're very knowledgeable about.
I read and enjoyed your retort. What you say is true, but unfair, especially as you criticise for being verbose, with probably the most verbose comment on this page. Lol Of course, it’s important to be accurate in what you say and write, even the spelling of “of course”. However, everyone has their own style (including me and you), but I think it’s best to accept that styles are often imperfect in their entertainment and education. The one big takeaway from this video for me was his educational value in failure, which he provided abundantly.
You needed linear rails supported every 50mm instead of round shafts supported say 300mm apart at the guide ends, ball screws to drive the slides, and a stiff spindle shaft supported nearer to the Z slide base. But that, as you concluded, is a whole new more expensive CNC router ! I recently bought an all metal, linear rail, ball screw, Anolex 3030 EVO Pro ($660) built like a tank, emphasizing increased rigidity and low backlash, for these same reasons. It cuts Aluminum well, at moderately high spindle speeds and feed rates. I have just upgraded to the Anolex 1.2kw VFD spindle to replace the stock spindle; man it is hefty !
3 years later this video is fantastic. Thank you for documenting the effort and failure. Tired of seeing videos where in 20mins everything works and the hard work is skipped.
Hi I know this video is a year old or so but I just wanted to drop some suggestions from a CNC programmer. as far as your bits collecting aluminum in the pores of the carbide soak the bits in hydrochloric acid overnight and that should takeoff any aluminum that’s residually on them it works better than lie. also instead of air they sell cheap mist cooling systems that use air and a water soluble coolant to cool the tool and clear the chips away they don’t make much of a mess and give you a tremendously better finish , cut speed and tool life. Also consider some three Flute mills I’ve noticed they cut down on chatter due to the fact that there is always a flute in contact with the work and the pressure never really comes off of the bit in a cut as it does with even number flute end mills. Anyway I hope someone get some thing out of my comments great video and very informative thank you!
Your upper left and liwer right quadrants are swapped from industry standard. IOW :"X" is always called first in a geometric expression, "Y" is always called second, "Z" is therefore third, etc.. And a good reason to set your work station coordinates in the middle of a part is when you are working on a part that is round and you're doing work in the center of the round feature, as in cutting a tapered pipe thread (inside or ourside) into the feature. Another good reason would be that your blueprint establishes the origin at a feature well onto the part. Say, in a program meant to cut a block plate for the bellhousing mounting surface of an engine. Those are almost always given an origin at the center of the engine's crankshaft.
Yes, that’s what I was wondering. It just kind of seemed different from what I’ve been seeing and doing on my lathe and mill. Anyway I loved your video and would love to see you try to cut using a coolant like WD-40!!
I was wondering if anyone else caught that. I ran that section of the video three times to endure I hadn't misunderstood and that the horizontal line was the x axis.
I think the frame of this little machine is probably stiff enough to handle alu. Upgrading the motor is a good step, but now other factors come into play and they are all compouding factors. The mill bit really wants much higher RPMs. The mill bit is probably not of the best quality. The lead screw has too much backlash and the anti-backlash nut solution does not take away the backlash with the forces involved. The linear bearings are probably cheap ones with too much play. And after all those issues are fixed you might even find that the steppers do not have enough torque.
Well, that saved me a bunch of typing 👍😁 Too much work but profiled linear rails, ballscrews and ditch the Mickey Mouse steppers; get closed loop servo motors on there. That cantilever bothered me from the get-go.
i went to 92 oz in steppers, 500 watt spindle shitcanned all the plastic parts and upgraded to aluminum replacements. started using Yonico and Amana bits and feel confident on any wood or plastics jobs but im not going to destroy the little machine on trying to cut aluminum. i also converted the motors to hybrid closed loop steppers and then went through every wire and replaced them with all sheilded jacketed cables because the RF interference was causing lost steps.
38:21 They usually oversize the fan so that it is more effective at cooling the motor with the air flowing over the top and cascading down the sides. When you make the fan the same size as the motor it causes the fan to dead end right into the top of the motor and won't have the same air flow down the sides.
Feeds and speeds also assume a rigid machine with a motor that will not stall at the proper RPM. You've discovered most of this already though. I've seen a guy make successful cuts on a cheap machine, BUT it was a heavily modified machine. Adding extra rigidity to the spindle, using coolant, changing to a high power motor with a VFD, and changing to ballscrews. All of that in a machine a bit larger than what you are using there. Also being able to hold 0.0005" tolerances. But we are talking about a complete redesign of the machine. You might have gotten a tad farther with this machine and got slightly better results but I think it would have taken more money.
I have been considering buying a cnc router and I found your video most informative. Thank you for all the work you put into this. In the world of hobby machining we use WD40 as a lubricant when machining aluminium. It eliminates the galling which blocks the cutter leading to all those broken end mills.
A channel named "Awesome CNC Freak" recently made a vid on how to upgrade this machine. He decreased the backlash, the chatter, and even milled steel. They look like low cost improvements; he used the original motor but added pulleys and a better spindle. I'm super tempted to try.
To avoid aluminium sticking to the bit you need to take larger chips. The heat caused by the cut will be ejected with the chip, limiting heat buildup in the stock. Ideally you need a faster feedrate, but you may have to try a slower spindle RPM if faster moves are out of the question. Also compressed air to blow away the chips will be helpful.
How come no one has suggested using coated end mills. Even the cheap titanium coatings would be helpful here but there are many more advanced coatings which are all designed mostly for this specific issue. Your comment regarding chip loading absolutely applies as well.
This is a fantastic video. I've been playing with a dirt cheap 3018 for over a year which has taught me a lot but I learned loads more from this vid. I will just note that the title should have been "Proving That It Is Impossible To CNC Cut Metal For Under $400".
I'm currently in the same situation, I have a Shapeoko 2 with lots of upgrades and I've finally taken the leap to cutting aluminum, with the hopes of making parts to make it cut aluminum well. First, you're absolutely correct that speeds and feeds matter. I'm currently cutting 0.6-0.7 mil per tooth that seems to be a sweet spot for my machine, using either 2 or 3 flute end mills. Second, I found that performance definitely improves with the right surface speed. That depends on RPM and bit diameter, but carbide really likes it fast: 800+ s.f.m. Third, Fusion 360 is free for hobbyists and can produce the trochoidal mill path you describe. It's much easier to clear the chips when you cut a wider slot this way. Unfortunately, trying to cut 0.7mil per tooth is useless if the backlash is 4mil and spindle is chattering all over the place. You end up with some teeth cutting nothing and others cutting 5 mil. On my machine I've got the backlash pretty small, but I believe the current limitation is overall machine stiffness, the opposite of the problem you have.
I can tell you with absolute certainty that the chatter and drift in these types of machines is entirely due to the backlash nut essentially being loose. You either need to put a much stronger spring in, make a backstop that can be tightened against the nut, or find small ball screws. I've run dial indicators all over these machine types and they are actually remarkably accurate, but not under high load. The cutter is bouncing around as it cuts. These machines can and will do 1000mm/min adaptive cuts with a 1/4" full depth. I hope this is the answer someone was looking for.
When I work on projects that require accuracy (pretty much everything), I work off centers to be sure that my parts are balanced. Using all the Cartesian coordinates enables ease of symmetry for parts or structures based on center. Using all positive x,y coordinates will make symmetrical layout a bit more difficult. For example, if you are cutting brackets that require holes equidistant from center all you have to use are identical integers for distance and assign positive / negative values in either x and y or both. Anything you create that is radially balanced on center would be a bit more difficult using all positive coordinates. In the end what works for you is what works, so I do understand that when things are confusing, using a method that enables you to move on and be productive is most important. Thanks for the video.
When I started out I was naive enough to think I could build a large gantry style mill mostly from scratch. I ran into almost all the same problems as you. It was a very humbling project but I learned a TON.
I think it is possible. I built a large CNC machine out of 4040 and it is just like too much of everything. It has big Nema 34 motors and I am a Nema 17 type of person.
I am glad someone finally said it. They are playing the Frauenhoffer game. Give people something for free, lock them into that software/product and then when people are locked in, squeeze them for money. Rinse and Repeat. Frauenhoffer did it with MP3 20+ years ago and Autodesk uses that technique even now. Also .. there is a blue anodized version of this machine that has linear rails as opposed linear rods. It is not as readily available though.
Matt, I have looked a bit at the issue and for machining soft aluminum, you need at least 800W spindle or bigger with water-cooling for a proper control of the minimum rotation speed and cut speed. Your video has many pertinent points, you have learnt a lot from this experience. Keep the good work.
A few tips: keep your cutter as short as possible. Use cutting fluid. Use the lowest RPM that will cut effectively. Keep your feed rates as high as possible. Climb cut whenever possible for better finishes, except when you have lots of backlash in your axes. Make starting holes with a drill bit before plunging with a mill bit. (two setups, though). Rigidity is the most important thing for clean cutting.
If you get a little tub of "MIG Dip" from a welding store, and dip your cutters into the compound, it helps prevent the aluminum from sticking to the cutter, it helps to heat the cutters to 110° or so to liquefy the dip compound
Single lip cutter or a D-Bit with gasoline for a coolant. This I’ve observed from professional engravers for 35+ years (for Aluminum only) cutting oil otherwise . We make the cutters on a Deckel single lip cutter grinder. We sharpen the cutter in the collet which is belt drive. It insures perfect center. We relieve the metal on the backside of the cutter which you probably know. I’ve seen cast cement frames made encapsulating the frames to take out the harmonic frequencies. The RPMs are fast , 20-30k. The main ingredient is a good fast spindle and a rigid frequency deadening frame. I know this this from watching hand operated Gordon And Deckel pantographs for decades. I’ve never seen them use a mill type cutter or a spiral flute. These machine are like the Flinstones other than the spindle and a rigid frame
You were a huge part my 3d printer start back when you did alot of the delta style printers. That tevo little monster was a dream of mine lol. Glad to see your videos popping back up on my feed. Always enjoy your explanations and indepth approach for all of us who are alittle slow or new to this type of hobby. Makes you feel confident about taking on the next challenge. Awesome video as always
Hiya, not sure if any one has already said this but a few tips, kerosene is the best lube for machining ali, the tool speed for hss is N = 1000 x S / π x D where n is spindle speed in rpm S is surface speed in M/pm (avlb from material data sheet) ali is about 100 from memory, D is diameter in MM, if using coolant 10-15% faster, if using carbide rpm x10. Rule of thumb is 0.1 mm per tooth feed. Unfortunatly the above is great if you have brand new machines so, if it chatters slow the spindle increase the feed, make sure your tool bit is as short as possible and sharp, ali is really abrasive, if using HSS don't let it rub as it will take the edge off the cutter making it a paper weight, don't climb mill unless you have a rigid machine with 0 backlash or your tool will fracture in seconds, if you can climb mill it cuts far better. when selecting a cutter for side and face machining make sure the trailing cutter starts the cut before the previous finishes. the only other thing i can think of is the collet, but the only way to fix that (apart from a clarkson chuck) is mabey using a larger collet and machine a nice long snug slieve out of brass and put a small slit down its axis, it works quite well. OK woffled enough. g luck
You're absolutely right! I'm totally going to rezero my MR-1 with the positive all available, although obviously the z axis will still go negative, but I think I will like it much better also.
When cutting aluminum, in order to prevent clogging of the cutting bit, you don't blow air into the bit! You absolutely need to lubricate often the bit. Nothing replaces lubrication.
I bought a 4180 machine of similar quality and even added the same 500 watt spindle . It's fine for wood or plastic but yeah for aluminum it's a struggle . I saw a fellow who said the run out in the spindle was bad because there is a rubber bushing around the drive shaft bearing to cut motor vibration but of course it causes runout as the shaft can bend inside the rubber bushing! He was a machinist and replaced the rubber bushing with brass bushing he made on his lathe. Thanks for the great explanation of hardware and software for milling , cad and cam too. ☺
Hello, just as an FYI if you are cutting aluminum. Use Varsol as a lubicant and coolant. Thats machining 101. Aluminum is very Gummy and a low melting temp thats why its gumming up the tool. No charge by the way.😁 Do not use Carbide with alluminum. Carbide likes to work at high temps not good for alu. Use HSS (High Speed Steel) 2 fllute.
You can do a little more to make this machine much more stable and even more accurate. I filled all of the lightweight Alu profiles (i.e. these hollow aluminum parts) with a lead / tin mixture - you can also use pure cement but a lead / tin mix is better. That makes the whole machine heavier and it runs smoother and more stable and is a little more accurate when milling. The steper motors are powerful enough to handle the extra weight.
My first comment to you is about the topic of starting in the logically appropriate position on your workspace (and physically on your work piece). This is your piece of wood or something more expensive (like metal or stone or whatever). I used to have a CNC plasma table and I told the builder of my table that I don't want my table to start in the middle of the work piece because of waste and he changed the focus to the lower left position. I don't remember any problem with +,+ vs -,- because my tool could run a dummy or sample of the program so this would visually SHOW where the head of the machine would travel. This would therefore validate if being too low is a problem. This is the right way to start anyway (in my opinion) I agree with you. The major reason I was not satisfied with my machine was because of one feature in how it was built. I know now that it was built using the wrong technology (the most important part) that affected the accuracy of the head. The X and Y and Z movement on my machine was being controlled by straight cut gears, the worst choice. No matter how the adjustments were made these gears would affect the quality of the curve or corner of each movement. It was after this that I was then noticing so many ads for so many machines were now becoming available.
Clickbait title: "How to CNC cut metal for under $400" Then proceeds to say it cannot machine metal worth a damn even after attempted upgrades. An hour long video that could have been condensed into 10 minutes of explaining why it cannot happen with that machine...
Great detailed video. One suggestion. It is not just about rigidity it is about dampening vinration too. There is no enough material in z axis to damp vibration. You will need a aluminium plate to hold spindle clamp. It will improve performance 200%, there won't be chatter and cleaner cut.
I had a cheap 3018 that I upgraded to a 500wat spindle and circuit board to. And found that harder Aluminum cut better as the softer stuff gums up the tooling and causes the tool to wonder. if you cant buy harder Aluminum then uses a lot of coolant. don't use oil as it will cause it to thicken up and go gloopy Also use a single flute bit, If you have a good cam software that supports adaptive cutting then it will definitely help the chatter problem If not then look at your ramps Meaning the angle of let say 2 degrees will help a lot And be aggressive as to the cut that that the machine can handle and then use a nice finishing pass
In the world of machine shops either the work is held still or the tool is held still and sometimes both. If it is rigidity you want switch all motion functions to the bed to which you fix your work, so the bed moves X and Y and Z and the tool is rigid and solid. Just like the big boys.
What i've done is up the feedrate to about 700mm/s, depth of cut 0.1mm. Spraying it with cutting oil, if you don't have cutting oil wd40 will do the trick. Spray some more when you start to hear chatter. Use a endmill with a coating and small diameter, 2mm would work for your requirements. The number of flutes are the least of your problem because the chatter does not come from the tool itself, i would use at least 2 flutes, but i've used 4 flute endmills with the same setup with no issues, just don't cut completly dry with such small endmills
I have the FoxAlien vasto. It’s a great machine for aluminum so far. Ball type Leadscrews all around. Motor couplings for easy upgrading. Heavy duty Dual linear carriages with rails for extra rigidity. 3 different sizes spindle holders.
Aluminum is hard to work with. I think you are using a bad alloy. It smears so much. Also you should be using a coolant. Then again you need the right tools to machine alumium. And at the end of the day, the right speed of rotation and the right feed. There are tables that recommend all these settings. I think it is not the machine that caused the problems, but primarily the material that was being used. Greetings from Germany!.
@@DesignPrototypeTest Flood coolant is impossible for this setup. Use WD-40. It is a perfect cheapo coolant/lubricant to prevent gummy material like aluminum from sticking to your cutter, which eventually cause your tool to fail. Works very well for me. It works for plastics too, preventing heated chips from binding to itself to form a big blob.
@@DesignPrototypeTestwho would have thought that if you have 0 idea of machining it doesn't work out first try 😂 and yeah there is a steep learning curve about 40% of this vid is bs single flutes are superior on small machines dump the rhino stuff and use fusion like everyone else and suprise suprise a 400$ mill doesnt perform like a professional one 1310 mills do work, with knolege and limitations, dont make sawdust, make chips and so much more
Aluminum is hard to work with? No son, you just don't know what you are doing. Aluminum is easy as fuck. Try inconel, titanium, or any high nickel alloys. Lmao. Aluminum is hard, that's the easiest metal besides brass 😂
You have switched the x and y axis I think. Normally, the X axis goes left to right, and the first coordinate is X. This means the the upper left rectangle is actually (-,+), not (+,-)
As a machinist I work from a center all the time! There are many benefits that I didn’t notice starting out, especially if the part has symmetrical features. For me its easier to visualize one side as positive and the other side as negative in the same direction. As opposed to having it just look like random numbers in the G code. But if the part has a feature like a hole or a bore I like to indicated and set the origin to the center of that!
I could see (0,0) being a useful center of my piece when its going to be a round geometry, because then I could make the measures for one cuadrant, and then adding negative signs to the X or the Y or both would turn those into Cuadrant 2 - 4 coordinates. At least that's my believe coming from someone who's not done any CnCing.
Moving (0,0) to center of the x-axis, and keep y-axis at the edge of the bed, makes it easier to design bisymstrical models, by changing the X coordinates , to -X coordinates. Place (0,0) at the center of the bed, and do the above to both X and Y, to make quadsymetrical objects super easy.
Thanks for posted this. I wont buy one lol. I started watching this thinking I could give this $400 setup a shot. Not many people post perceived 'failures' but it's probably saved a lot of people some money and head to desk banging.
I'm into one of these as well in fact I bought the 3018 as a test model to design a better one using SBR linear rails and 12mm ball screws and a 500 watt motor......3/4 finished but still on the back burner due to another project taking more priority........I would have to say that the one in the video is totally adequate as long as you don't push it ....scaling is a factor you have to experience and it won't like forcing the cut..........aluminium must be cut wet or you'll always get a cutter weldon with the chips.
Good to see a review on this machine. I didn't see if there were any drivers needed to install the CNC to your computer. Sorry I'm a total doofus when it comes to IT. I think the stage with the drivers was skipped. It's where a lot of people seem to have trouble with the other small machines. They can't find the drivers for their machines for this or that OS. Mostly the drivers are for Windows machines but people may want to use Mac's or Linux computers. I've seen people using both Mac's and Linux systems to control the cheaper CNC machines so it must be possible etc. But as far as I can see it's not clear what's needed to make them work, so the stage of using UGS can be reached.
The Benbox board on the 1310 uses a plug-in Arduino Nano clone. If it's a faithful clone then it should work out of the box on Windows, linux and macOS using the standard FTDI USBSerial drivers that ship with those operating systems. More likely, though, it uses a CH340x or CH341x USBSerial chip so you'll need the appropriate driver which you can download for Windows, linux and macOS from www.wch.cn/downloads/CH341SER_ZIP.html
@@anthonyrich1592 That's extremely useful. Thanks. I've got a little Windows 10 netbook I'll be using to control the machine myself but I've seen more than a few people posting about not being able to find drivers. It might be that if this board is easily recognised, it may give the product an edge over boards which need proprietary drivers etc. Hopefully your comments will help others who come looking.
@@pentachronic yeah i dont understand why people think more flutes cuts better. the more flutes the faster you have to cut with a deeper cut to get the chips the right size. some people think dust is cutting and all producing dust does is wear out bits. i started using all O flute bits for my wood and plastic work and the cuts have been much faster with a far better result
You're right about "milling bits" not liking to drill, that's why there are two types, end mills which don't have drilling capability to speak of or swarf clearance, and slot drills which do like to drill, do have only two flutes to clear swarf and also can cut slots full width due to there only being two flutes. Mix those two up and you'll break tools you needn't.
welcome to the rabbit hole that is called cnc-ing metal. I fell in it too. Build myself a momus a couple of years ago. You wont stop upgrading once you see what others can do on youtube :-).. I can recommend the momus though if you like diy. You can build it using simple tools only (I know, I did) and its relatively cheap (In CNC, cheap is always relative :-) ) . Your cnc looks nice too although I fear you will one day upgrade the spindle motor. If you ever decide to spring for one of those chinese high speed spindles, do pick one with an ER16 collet though. Much more choices when selecting cutters as the ER11 only fits a maximum of 7mm and the choice for 8mm and up is really big. Don't ask me how I know...
Thank you. You just saved me a lot of time, money and frustration on buying or DIYing such kind of machine. It turns out that for my needs it's cheaper and better quality just to order required parts (tested with 3d print) in some kind of CNC cutting workshop.
For real! Every time I try and give this guy another chance, he does this kind of stuff again.. I'm going to just block him from my search and feed at this point.
And why did he keep saying how sturdier this version was compared to the previous one... if he then ended up printing a holder for the spindle out of PLA? Why are we even speaking of sturdiness, making a bunch of changes, if you then use a holder made of soft plastic? Are we kidding?
@@xhorntail would you prefer he lie to you? You don't want to know the truth, and I hate clickbait as much as the next guy, but you wouldn't have listened if he'd titled it "you cannot cut metal for under 400".
@@brahtrumpwonbigly7309 this and other machines CAN cut but not for pro jobs where precision is a must. He printed the holder, because of the new motor. I have news for you, the other one for 1K is also no good too. Must go up in quality, sturdiness, etc. meening >$
Check Alex Moon's videos, he figured out how to mill aluminium with a CNC1310. He uses a much faster spindle. I have the 500W spindle you have and only goes only to 11000 rpm . I think we need more speed.
The soft aluminum you used is too soft for proper milling without cooling. I tried that once during an internship with a professional milling machine and it was just chewing gum.
I do alot of aluminium fabrication, I drill, countersink, circular saw with a tungsten tooth blade, jigsaw and router alloy, if I don't use a penetrating fluid or a stick wax it will only get hot and pickup on the bit as you have shown. It's a friction problem. Try slowing down the rpm to disperse the heat, not to slow to overload the bit. WD40 or a low vis cutting fluid. Hope it helps. Like your vid, thanks.
I think you should have tried to mill your intermediate spindle mount out of delrin. Even if you still have some play, you should get more rigidity, equivalent or better tolerances, and a proof of concept for your aluminum design (and you could have shifted that spindle hole a couple of millimeters to accommodate the fan)
Its hard to tell what the issue is without a video of the actual cut, but generally when you see aluminum galling up like that, its due to a surface speed that is too fast, or full slotting with too high of a feed rate. That being said your spindle only does 12k rpm which isn't terribly fast. I would try doing a shallow depth of cut start at .01mm with multiple passes, it looks like you were trying to make these cuts in a single pass, which just isn't possible at 12000 rpm. full slotting metal is difficult for an endmill, and it typically needs to have a machine that's very rigid, has coolant, or can evacuate the chips, and also done at low rpm, maybe around 2000 for a 1/8in endmill. Also I would highly suggest getting into fusion for this type of work, the cam in fusion is awesome, and allows you to control a lot of variables to the tool paths.
I had that same spindle. It was terrible, the shaft is really thin and flexible. The chatter was unavoidable, no matter what gcode you use and even when cutting wood. When i upgraded to a 1hp watercooled spindle it made a huge difference, now i can cut aluminum in 1-2 mm passes
I believe most machinists would not opt to take full width , full depth cuts with this setup. Rather, they might use smaller diameter bits to "peck" at a rough cut of the initial pass till the desired depth is reached. Once most of the material is removed to the final depth, a single "conventional climb" full depth cut would be used for a final, accurate cut.
wait do not throw in the towel.... you are doing great... I can make the adaptive gcode for you, and you can try it again. 3mm single flute looks good but better to enter the material from outside rather start inside, chips will go out faster.
I remember LibreCAD when I played with pen-plotting! Neato EDIT: I split this watch-through into two segments and it's very informative to see how BAD these things chatter! Wowza
If your bits came from China I can tell you that's gonna be a problem. I couldn't cut crap with the carbide china bits, got a brand name carbide bit and solved most of my cutting problems.
There are a few machining strategies that should be used for weak machines/setups. HSS tools, even 4 fluted tools, are preferable to carbide. You want to always conventional mill, never climb. Climb milling kills machining passes when the rigidity and backlash get involved. I feel like a very short 2 flute HSS low helix tool conventional milling with small step downs and high feed would work best here. Also, I doubt know why everyone thinks soft aluminum cannot be cut without coolant, that isn't true. It just isn't, and there isn't much more to say about it. You encompass that factor into your feeds and speeds. Aluminum only gums up if you heat up the material during the cut. If you don't, it will not be a noticeable issue.
It’s normally called a tool not a bit. CNC software has backlash compensation and lead screw compensation. Look at your feeds and speeds. Usually the electronics are in an enclosed box. You need way covers to protect your lead screws. You need a motion control algorithm with acceleration and deceleration.
Hi Matt (is that your name), Bam, I subscribed. Great video. It shows the value of failure for education, which you make entertaining too. If you agree (that failure is important for success), then I’d ask you to carry on with this project and not give up altogether. Instead, learn from the failures, take on some of the points made in these comments above, about topics such as metallurgy, coolants, etc and continue to explore what’s needed to get this machine to do the job. You’re far from finished with it - I hope. Looking forward to an update later this year 😄
Hey man, what a great video packed full with quality information. Maybe you said it and I missed it, but did you use conventional or climb milling? Or both? The difference in outer dimensions vs. inner dimensions at around 41:40 could be explained by the mixed use of both milling types. I'm sure you know this already but climb tends to push the bit away from the material whereas conventional tends to pull the bit into the material. Also spring passes (100% overlap with the previous toolpath) could be used to get better dimensionality. At 43:14 you mention the problem you're having with chatter. In my experience with a similarly built cnc, it helps to experiment with different (scary low in my case but that of course depends on the machine) spindle speeds until you find a speed where the frequency of chip creation doesn't positively interfere with any of the machine's resonance frequencies. But that quickly becomes a shitshow ;). As for the optimal (literature) feeds and speeds: I tried them and they didn't work. My gut feeling is that those work on tormachs that actually have the required rigidity for those settings but they are incompatible with hobbyist-machines. To the cooling: I tried compressed air, but the airflow required to cool my parts was so high that it would bend the bit/shaft/motor-mount which would create even more heat and chatter. I had good luck with isopropanol as a cooling agent because it evaporates really fast and doesn't leave a mess. And about the adaptive clearing toolpaths: With a small workaround you can get Fusion360 to do those perfectly fine. Even with the free version afaik, but I understand that Fusion isn't as cool as it used to be due to their new pricing model. The last thing you shouldn't forget about is that aluminium is terrible at absorbing vibrations. It's light and stiff which is great because it will make the machine light and stiff ;) But at what cost? Professional machines are made mainly out of cast iron in part because of its dampening characteristics. In this case I think dampening the frame by different means might be an option. The easiest idea would be to print 0.5 to 1mm spacers out of petg which also has great dampening characteristics and put them in between each of the aluminium-aluminium joints on the machine. For that same reason, I think that those cnc machines like the foxalien or the shapeokos with openbuilds-styled delrin or nylon rollers are a great compromise because of the dampening characteristics of those rollers.
some of the larger hobby machines are definitely capable of cutting aluminum but you really want at bare minimum 1.5 kw spindle, 4-600 oz in steppers and a machine using cast parts and not extruded parts. cast iron is better than cast aluminum. but then youre talking a minimum of 2-3000 dollars for a machine like that. im building a gecko g540 machine using a completed chinese machine with cast iron parts and gutting it of all crap electronics and shitty wiring. but people hear CHEAP and they think wow i can do all that for $500 and its just a lie. I have over a thousand into my what started out as a 3018 that cost $300 and its capable of cutting wood great but wouldnt attempt it on any metal and the size is just too small for practicality.
I also prefer a positive only coordinate system. With a positive/negative central location system, if you get distracted for a moment, and you forget to sign your command, well... I would really only recommend conventional cutting with these machines. The rigidity is nonexistent. The motor itself isn’t meant for this. The new motor is much better. But the machine itself isn’t capable of taking usefully deep cuts for thicker metal. The new motor is far from the rails, giving it a greater ability to bounce up and down. So some of the chatter is due to that. The fan is wider than the motor body so that the flow moves around the sides of the motor. When the fan is the same diameter, you get a very turbulent flow because all of the air is now hitting the flat rear of the motor. Assuming this is designed properly. You know that aluminum sticks to cutters. You need bits coated for aluminum. Since you know you need coolant, I’m trying to understand why you’re not using it. I often use water diluted sawing coolant for much of my work. For this tiny machine, a few drops here and there is all you need. Why not use it? A slight mess is part of the game. Blowing these fine chips away is not the solution. I see at the very end you discuss that motor overhang. There is no way to eliminate that. My machine weighs around a ton, and it’s considered to be a light machine. Chatter still happens, though with much heavier cuts. I’ve played around with a CNC Sherline, and with anything more that a 0.002” cut, it vibrates like crazy. I don’t know how anyone gets anything done with these little devices.
3 flute is way too big for a machine this size. youll destroy bits like noones business. the more flutes you have the faster the spindle has to spin and the faster the machine needs to travel to get optimal chipload. a 1 flute or O flute works best on these smaller machines. if you produce dust you are burning bits if you produce chips the temp and friction on the bit is far less and the bits last 10x longer. also never buy crappy chinese bits. if your bits are less than $15 ea your running crap. i run Yonico on the cheaper side and Amana on the higher end side depending on the job determines the bit i will use.
Design Prototype Test, I don't know if anyone committed on this before, but I want to let you know that you made a mistake when explaining the Cartesian coordinates. The top right is (+,+), the top left is (-,+), the bottom left is (-,-), and the bottom right is (+,-). This is not criticism, just clarification. On the other hand, thank you for your very informative video.
The Path workbench on Freecad has the adaptive clearing feature you were missing in this video. I'm not sure if they had it already when this was filmed, but at least they have it now :) Ever tried it? Great Video by the way. I learned a lot listening to your comments throughout the video!
Adaptive tool paths were available in FreeCAD 18 so I think they should have been available when this video was filmed. The only slight gotcha would be that adaptive profiling and tabs don't seem to play together nicely in FreeCAD (applying tabs to an adaptive profile seems to send FreeCAD into an infinite loop when I try it). The work-around would be to use adaptive to get most of the way down then do the last bit, with the tabs, using a normal profiling operation. I do agree with you that it's a great video and I'm sure many people will have found it useful.
I agree with you on the lower left of the machine should be 0,0 the only reason I can think that they would make the center of the cut area 0,0 is their programer doesn't have any real world experience. Most drafters, and other trades people were trained to use 0,0 as the lower left. in the end it doesn't really mater, it still math but it would bug me too!
I suggest wiring the 500W motor as if it was direct mains connected - mains rated wiring, wiring protection and earthing of the motor and metalwork. I bought one of these sets a few months back and the power supply doesn't provide mains isolation.
Was a really good video going through so many different aspects of your decision making and the basis of a good cnc, to software etc. felt the pain at the end. I think we all want a working machine ideally cheap to buy. How would you make your cnc or what do you think needs changing in your final machine . Your spindle holder was not great being short in height. maybe worth revisiting when the pain has gone.( not to make us want to buy the machine but build it.) See Aurora Tech for 3018 based/robo cnc metal cutting though limited. Please consider using her spindle holder and/or drilling holes to mount your spindle to see if it helped. Did love your video - was just really sad for your hard work ending with disappointment and you seem to have abandoned cnc.
@43:50 i can explain you difficulties. The bouncing back and forth is because of your motor holder it's made of PLA it not only gives in the long run but it has e very elastic deformation E-module all the vibration and forces bend the motor holder elastically.
What machine do you recommend to cut aluminum if the budget is $1000?
This one is promising, but I haven't tried to cut Aluminum with it yet: www.amazon.com/Machine-4040-XE-Spindle-Engraving-Engravable/dp/B08M96HHKW/ref=sr_1_1?dchild=1&keywords=cnc+router+4040&qid=1605163511&sr=8-1
I will be making a serious video on that CNC machine in the near future. At that price piont I think your only other option is to DIY something.
@@DesignPrototypeTest Thanks for replying. I'll have to look at that machine.
@@ClintonCaraway-CNC I went with sainsmart 3018 to test the waters. I'm stuck at resetting machine coordinates to zero. Been trying and researching for the past few days with no avail. getting frustrated for sure.
MPCNC
@@pentachronic I'm in a few of them. I'm having issue with machine zero coordinates. Got everything else. Seems like I'm missing the software side of it. Playing with Candle atm.
It was cutting metal parts that were good enough to work but needed a couple changes to cut better (not more parts). I've been cutting metal on my machine for 10 years (it's a couple tiers up from that machine)
*Tips for anyone trying to cut ALU or Brass:*
1. *Suck the bit up* as far in the collet as you can (not on the flutes, but as close as you can). It makes a big difference.
2. *Deep cutter-width slots are a no-no!* Set your profile cuts at least to at least *1.5x cutter width* (also helps with clearing chips). It will take twice as many passes, but your bit will thank you and your surface finish will be much better.
3. profile cutting: if you can, precut/drill a hole in the middle and make sure your path starts from the hole. You will be able to take deeper cuts as it is *side-cutting instead of slotting*
(As soon as saw your deep narrow slot, I knew immediately it would break a bits and/or lose steps.)
4. If cutting dry/air only: *Use WD40* Frequent squirts cool the bit and give it a coating that resists chips sticking.
5. *Buy better/midrange cutters* $5-8 cutters make your machine cut so much better than $1-3 cheapies. (don't buy expensive cutters until you learn how not to break them)
6. *Cut chips not dust!* Dust = wearing out bits fast with almost no material removal. (small chips are ok, but grey dust isn't)
7. *Don't cut steel.* Just don't. Your machine will damage it's self even if you can't see the damage immediately.
e.g. My motor shaft coupler started slowly fatigue cracking and then later snapped and caused a nasty crash that actually bent my Z ballscrew... sniffle.
Purchasing tip:
8. Get a machine with the largest dia *ballscrews (not leadscrews)* and largest rails you can. The water cooled spindles are also much quieter.
Upgrade tip:
9. *Flood coolant* with water/cutting fluid: if you plan to cut metal a lot and are willing to set it up, this *will massively increase your machine's cutting capability.*
Seriously, I went from struggling to cut metal at all - to wondering how far I should push my cuts.
Not only does it blast chips away and cool the bit, but it also reduces vibrations and noise! (straight water will rust everything including your cutter's crisp edge needed to cut well)
I use a cheap fountain pump and water soluble cutting fluid. My machine sits on a tilted table with angle alu edges to keep the water in, then it all runs into a rubbermade tote with the pump in it.
Linelock makes directing the flow much easier and isn't too expensive for a foot or so of it, but I started with a coat hanger wire ziptied to a piece of hose!
PS. I'm not saying you should go back and try that machine again. It is just going to be on the painful side. But it was doing it until the too deep narrow slot started chattering too much. So for anyone else, who already has a machine and can't afford to just go buy another machine, I hope these tips help.
Videos like this - where things Don't work out - are way more truthful and helpful than polished and edited videos where everything works perfect the first time. Thank you, sir!
This is what I thought too.
This video will save countless viewers from failure for aluminium routing projects - thanks for the time, effort and expense you went through to produce this video.
Agree
Actual machinist here: It may seem odd, but you can often remove chatter by increasing feed, or decreasing rpm.
Keeping a high load on the cutter makes it unable to jump around.
Adjusting feed/RPM while making a cut is a good way to find the right numbers. What the read in online calculators is just a starting number, you find the ones you need by adjusting and listening.
That's definitely true, and part of the problem with these little hobby machines is that they generally don't have enough muscle to give a carbide bit what it wants to eat.
That is true for a better setup, but this is likely due to a lack of rigidity and power that will not be overcome without losing your precision. Might work to a degree for roughing but I doubt it in this setup.
@@Skinflaps_Meatslapper I think people glorify carbide as an industry standard whan HSS is generally a better option for low power machines too. It's easier to get it to bite since has a finer edge usually.
@@brahtrumpwonbigly7309 It's surprising how dull most carbide inserts are when new, and that's part of the problem. You need more force to get the duller insert to cut, but since it's such a hard alloy the additional force doesn't speed up the dulling process like it would with HSS, it just ends up cutting better. Maybe the really high end carbide tools have a better edge, but those are usually out of my price range. I hit mine with a fine grit diamond wheel if possible right out of the box and less aggressive feed rates work a lot better, but they're still just a bit more susceptible to chatter than HSS is. Could be due to a more rigid tool and harmonics, I don't really know to be honest. The slop in a cheap machine plays a big part in that, so unless I need carbide, HSS does most of the cutting in my shop. It's also pretty handy because I can forge, shape, and sharpen HSS to whatever kind of cutting tool I want...carbide is way out of my league LOL
@@Skinflaps_Meatslapper HSS will be better than cheap carbide over time. Good carbide however is unbeatable.
Thank you comments. You saved me 50 minutes. I now know that this doesn't actually show me HOW to cut metal on CNC.
I'm used to larger, corded routers ruclips.net/user/postUgkxfQ5_mgwq6PcudJvAH25t-I4D-3cTPz4z so this was a different experience for me. Basically, this is an incredibly sweet machine, fits well in the hand, etc. It has slightly less power than I'm used to, but that's understandable. Maneuvering freehand takes some practice. These days, it's especially nice to be able to avoid power cords. This is above the level of a hobbyist but below a pro level.
Few points of interest here:
1) The terms most often used is Backlash, or play, in 22 years, I've never heard anyone simply abbreviate it to "lash"
2) judging from what's shown at @00:07:00 that anti-backlash nut is installed completely incorrectly. The spring is NOT supposed to be compressed that hard, nor is it supposed to be compressed against the frame, the way you "fixed" it completely negates it's use.
The way these devices work, is they put tension between the nut, the follower, this ensures the nut is always engaged with the threads of the lead screw
This image displays the proper installation of a spring loaded anti-backlash nut:
images-na.ssl-images-amazon.com/images/I/71xI4rTFy4L.jpg
3) At the scale you're talking about, rigidity issues on the "cheap" versions has very little to do with the frame, and more to do with the hardened guide rods, and the fact that on the machine you're reviewing ,they are offset, and the machines you pointed out, they are in line. This will lead to "nod" and tram errors int he cheaper machines as they plunge into work, and chatter as they move through the work. However, again, with the scale of these machines, you shouldn't be putting sufficient cutting forces with any of them to bend rods, you're not working with VMC's, so trying to rooster tail 4041 with this isn't going to happen anyway. Feeds and speeds are where most of your issues are going to come into play.
(yes, 12mm hardened rods will be stiffer)
4) Moment of inertia calculations: Wow.. talk about going a long way for literally no reason... Yes, the MoI for 8mm rods is 201, and 1017 for 12mm rods.. However, these numbers actually don't tell us anything practical about the rods, their material, their hardness, bending forces required, or how they relate to the application they are used here. That's like telling someone the tensile strength of nylon when talking about a wire brush..
If you're supplying enough force into the work with your cutter, that you're bending, or displacing the rods (2 hardened rods, and the lead screw) then you're obviously over-loading the machine, and operating it unsafely.
again.. you're going out of spec for these machines to try to prove a point, that literally, doesn't need to be proven. yes.. 12mm rods, and a thicker gantry cross member will be more rigid, 100% accurate.. literally all you had to say... once..
5) False equivalencies abound. You point out that the connections between two parts are not equally rigid.. um.. ok.. of coarse they aren't going to be equal, they are 1) different connection planes, 2) different connection points, 3) different components, 4) different profiles... To assume they would be the same is illogical at best... Anyone, can simply take a look, and see they are different in every aspect except for perhaps, the fasteners used.. .on one machine, you have a piece of extrusion, supported with two right-angle blocks, that are clamped onto the extrusion, and the frame (total of 4 connection points, on 4 faces) and on the other, you have 6 connection points, all on the same face... Of coarse they are different, they are different designs..
Also, comparing 10mm thick solid aluminum plate to extruded aluminum..... really... The plate will be less likely to tilt as you've described than the extruded aluminum, mainly due to the way they are clamped, and the fact that solid aluminum vs a hollow extrusion.. that 10mm of extra wideness on the extrusion probably isn't going to compensate as well as you might think. However, that being said, if you're putting enough lateral forces on either machine, you're not operating your machine safely, or properly.
6) It's not an "axel" it's a spindle. An axel is for wheels, regardless if it spins or not.. a spindle is any rod which spins within a mechanism.
yea.. i'm not going to go any further into this review.. sorry my guy, but it appears you're trying to look like you know what doing, but there are a myriad of technical issues here, everything from terminology, to mechanical assembly, to the way you describe things shows this isn't something you're very knowledgeable about.
I think the video is aimed at the home hobbyist, not the anal retentivest.
I read and enjoyed your retort. What you say is true, but unfair, especially as you criticise for being verbose, with probably the most verbose comment on this page. Lol
Of course, it’s important to be accurate in what you say and write, even the spelling of “of course”.
However, everyone has their own style (including me and you), but I think it’s best to accept that styles are often imperfect in their entertainment and education. The one big takeaway from this video for me was his educational value in failure, which he provided abundantly.
You needed linear rails supported every 50mm instead of round shafts supported say 300mm apart at the guide ends, ball screws to drive the slides, and a stiff spindle shaft supported nearer to the Z slide base. But that, as you concluded, is a whole new more expensive CNC router !
I recently bought an all metal, linear rail, ball screw, Anolex 3030 EVO Pro ($660) built like a tank, emphasizing increased rigidity and low backlash, for these same reasons. It cuts Aluminum well, at moderately high spindle speeds and feed rates.
I have just upgraded to the Anolex 1.2kw VFD spindle to replace the stock spindle; man it is hefty !
3 years later this video is fantastic. Thank you for documenting the effort and failure. Tired of seeing videos where in 20mins everything works and the hard work is skipped.
Hi I know this video is a year old or so but I just wanted to drop some suggestions from a CNC programmer. as far as your bits collecting aluminum in the pores of the carbide soak the bits in hydrochloric acid overnight and that should takeoff any aluminum that’s residually on them it works better than lie. also instead of air they sell cheap mist cooling systems that use air and a water soluble coolant to cool the tool and clear the chips away they don’t make much of a mess and give you a tremendously better finish , cut speed and tool life. Also consider some three Flute mills I’ve noticed they cut down on chatter due to the fact that there is always a flute in contact with the work and the pressure never really comes off of the bit in a cut as it does with even number flute end mills.
Anyway I hope someone get some thing out of my comments great video and very informative thank you!
Your upper left and liwer right quadrants are swapped from industry standard. IOW :"X" is always called first in a geometric expression, "Y" is always called second, "Z" is therefore third, etc..
And a good reason to set your work station coordinates in the middle of a part is when you are working on a part that is round and you're doing work in the center of the round feature, as in cutting a tapered pipe thread (inside or ourside) into the feature. Another good reason would be that your blueprint establishes the origin at a feature well onto the part. Say, in a program meant to cut a block plate for the bellhousing mounting surface of an engine. Those are almost always given an origin at the center of the engine's crankshaft.
My dyslexia kicked in and I reversed those two quadrants. Woops! My bad. Hopefully a million people don't see this video too. :)
Yes, that’s what I was wondering. It just kind of seemed different from what I’ve been seeing and doing on my lathe and mill. Anyway I loved your video and would love to see you try to cut using a coolant like WD-40!!
No wonder you like working in the +,+ quadrant :)
I was wondering if anyone else caught that. I ran that section of the video three times to endure I hadn't misunderstood and that the horizontal line was the x axis.
Thought the video was mirrored or something
I think the frame of this little machine is probably stiff enough to handle alu. Upgrading the motor is a good step, but now other factors come into play and they are all compouding factors. The mill bit really wants much higher RPMs. The mill bit is probably not of the best quality. The lead screw has too much backlash and the anti-backlash nut solution does not take away the backlash with the forces involved. The linear bearings are probably cheap ones with too much play. And after all those issues are fixed you might even find that the steppers do not have enough torque.
Well, that saved me a bunch of typing 👍😁
Too much work but profiled linear rails, ballscrews and ditch the Mickey Mouse steppers; get closed loop servo motors on there. That cantilever bothered me from the get-go.
i went to 92 oz in steppers, 500 watt spindle shitcanned all the plastic parts and upgraded to aluminum replacements. started using Yonico and Amana bits and feel confident on any wood or plastics jobs but im not going to destroy the little machine on trying to cut aluminum. i also converted the motors to hybrid closed loop steppers and then went through every wire and replaced them with all sheilded jacketed cables because the RF interference was causing lost steps.
38:21 They usually oversize the fan so that it is more effective at cooling the motor with the air flowing over the top and cascading down the sides. When you make the fan the same size as the motor it causes the fan to dead end right into the top of the motor and won't have the same air flow down the sides.
Feeds and speeds also assume a rigid machine with a motor that will not stall at the proper RPM. You've discovered most of this already though. I've seen a guy make successful cuts on a cheap machine, BUT it was a heavily modified machine. Adding extra rigidity to the spindle, using coolant, changing to a high power motor with a VFD, and changing to ballscrews. All of that in a machine a bit larger than what you are using there. Also being able to hold 0.0005" tolerances. But we are talking about a complete redesign of the machine. You might have gotten a tad farther with this machine and got slightly better results but I think it would have taken more money.
I have been considering buying a cnc router and I found your video most informative. Thank you for all the work you put into this.
In the world of hobby machining we use WD40 as a lubricant when machining aluminium. It eliminates the galling which blocks the cutter leading to all those broken end mills.
A channel named "Awesome CNC Freak" recently made a vid on how to upgrade this machine. He decreased the backlash, the chatter, and even milled steel.
They look like low cost improvements; he used the original motor but added pulleys and a better spindle. I'm super tempted to try.
To avoid aluminium sticking to the bit you need to take larger chips. The heat caused by the cut will be ejected with the chip, limiting heat buildup in the stock. Ideally you need a faster feedrate, but you may have to try a slower spindle RPM if faster moves are out of the question. Also compressed air to blow away the chips will be helpful.
How come no one has suggested using coated end mills. Even the cheap titanium coatings would be helpful here but there are many more advanced coatings which are all designed mostly for this specific issue. Your comment regarding chip loading absolutely applies as well.
This is a fantastic video. I've been playing with a dirt cheap 3018 for over a year which has taught me a lot but I learned loads more from this vid. I will just note that the title should have been "Proving That It Is Impossible To CNC Cut Metal For Under $400".
I'm currently in the same situation, I have a Shapeoko 2 with lots of upgrades and I've finally taken the leap to cutting aluminum, with the hopes of making parts to make it cut aluminum well.
First, you're absolutely correct that speeds and feeds matter. I'm currently cutting 0.6-0.7 mil per tooth that seems to be a sweet spot for my machine, using either 2 or 3 flute end mills.
Second, I found that performance definitely improves with the right surface speed. That depends on RPM and bit diameter, but carbide really likes it fast: 800+ s.f.m.
Third, Fusion 360 is free for hobbyists and can produce the trochoidal mill path you describe. It's much easier to clear the chips when you cut a wider slot this way.
Unfortunately, trying to cut 0.7mil per tooth is useless if the backlash is 4mil and spindle is chattering all over the place. You end up with some teeth cutting nothing and others cutting 5 mil.
On my machine I've got the backlash pretty small, but I believe the current limitation is overall machine stiffness, the opposite of the problem you have.
I can tell you with absolute certainty that the chatter and drift in these types of machines is entirely due to the backlash nut essentially being loose. You either need to put a much stronger spring in, make a backstop that can be tightened against the nut, or find small ball screws. I've run dial indicators all over these machine types and they are actually remarkably accurate, but not under high load. The cutter is bouncing around as it cuts. These machines can and will do 1000mm/min adaptive cuts with a 1/4" full depth. I hope this is the answer someone was looking for.
Your videos are so underrated! Love your videos and honest opinion!
When I work on projects that require accuracy (pretty much everything), I work off centers to be sure that my parts are balanced. Using all the Cartesian coordinates enables ease of symmetry for parts or structures based on center. Using all positive x,y coordinates will make symmetrical layout a bit more difficult. For example, if you are cutting brackets that require holes equidistant from center all you have to use are identical integers for distance and assign positive / negative values in either x and y or both. Anything you create that is radially balanced on center would be a bit more difficult using all positive coordinates. In the end what works for you is what works, so I do understand that when things are confusing, using a method that enables you to move on and be productive is most important. Thanks for the video.
When I started out I was naive enough to think I could build a large gantry style mill mostly from scratch. I ran into almost all the same problems as you. It was a very humbling project but I learned a TON.
I think it is possible. I built a large CNC machine out of 4040 and it is just like too much of everything. It has big Nema 34 motors and I am a Nema 17 type of person.
@@jeffbeck6501 I agree that it's possible, I know I greatly under estimated the cutting forces the frame and motors have to resist.
You had me at "parts for your pew pew"
LOL I know right, you know pew pew nothing suspicious going on here.
Probably obfuscation due to RUclips-algorithm paternalism
Same lol
I am glad someone finally said it. They are playing the Frauenhoffer game. Give people something for free, lock them into that software/product and then when people are locked in, squeeze them for money. Rinse and Repeat. Frauenhoffer did it with MP3 20+ years ago and Autodesk uses that technique even now.
Also .. there is a blue anodized version of this machine that has linear rails as opposed linear rods. It is not as readily available though.
Matt, I have looked a bit at the issue and for machining soft aluminum, you need at least 800W spindle or bigger with water-cooling for a proper control of the minimum rotation speed and cut speed.
Your video has many pertinent points, you have learnt a lot from this experience. Keep the good work.
A few tips: keep your cutter as short as possible. Use cutting fluid. Use the lowest RPM that will cut effectively. Keep your feed rates as high as possible. Climb cut whenever possible for better finishes, except when you have lots of backlash in your axes. Make starting holes with a drill bit before plunging with a mill bit. (two setups, though). Rigidity is the most important thing for clean cutting.
If you get a little tub of "MIG Dip" from a welding store, and dip your cutters into the compound, it helps prevent the aluminum from sticking to the cutter, it helps to heat the cutters to 110° or so to liquefy the dip compound
Single lip cutter or a D-Bit with gasoline for a coolant. This I’ve observed from professional engravers for 35+ years (for Aluminum only) cutting oil otherwise . We make the cutters on a Deckel single lip cutter grinder. We sharpen the cutter in the collet which is belt drive. It insures perfect center. We relieve the metal on the backside of the cutter which you probably know.
I’ve seen cast cement frames made encapsulating the frames to take out the harmonic frequencies. The RPMs are fast , 20-30k. The main ingredient is a good fast spindle and a rigid frequency deadening frame. I know this this from watching hand operated Gordon And Deckel pantographs for decades. I’ve never seen them use a mill type cutter or a spiral flute. These machine are like the Flinstones other than the spindle and a rigid frame
You were a huge part my 3d printer start back when you did alot of the delta style printers. That tevo little monster was a dream of mine lol. Glad to see your videos popping back up on my feed. Always enjoy your explanations and indepth approach for all of us who are alittle slow or new to this type of hobby. Makes you feel confident about taking on the next challenge. Awesome video as always
Hiya, not sure if any one has already said this but a few tips, kerosene is the best lube for machining ali, the tool speed for hss is N = 1000 x S / π x D where n is spindle speed in rpm S is surface speed in M/pm (avlb from material data sheet) ali is about 100 from memory, D is diameter in MM, if using coolant 10-15% faster, if using carbide rpm x10. Rule of thumb is 0.1 mm per tooth feed. Unfortunatly the above is great if you have brand new machines so, if it chatters slow the spindle increase the feed, make sure your tool bit is as short as possible and sharp, ali is really abrasive, if using HSS don't let it rub as it will take the edge off the cutter making it a paper weight, don't climb mill unless you have a rigid machine with 0 backlash or your tool will fracture in seconds, if you can climb mill it cuts far better. when selecting a cutter for side and face machining make sure the trailing cutter starts the cut before the previous finishes. the only other thing i can think of is the collet, but the only way to fix that (apart from a clarkson chuck) is mabey using a larger collet and machine a nice long snug slieve out of brass and put a small slit down its axis, it works quite well. OK woffled enough. g luck
You're absolutely right! I'm totally going to rezero my MR-1 with the positive all available, although obviously the z axis will still go negative, but I think I will like it much better also.
Props for pulling out the moment of inertia calculation... I would just have said "ya... it's thicker... therefor stronger :P"
Smol things weak
big thing strong
oh ah ah!
My caveman mind as well.
When cutting aluminum, in order to prevent clogging of the cutting bit, you don't blow air into the bit! You absolutely need to lubricate often the bit. Nothing replaces lubrication.
I bought a 4180 machine of similar quality and even added the same 500 watt spindle . It's fine for wood or plastic but yeah for aluminum it's a struggle . I saw a fellow who said the run out in the spindle was bad because there is a rubber bushing around the drive shaft bearing to cut motor vibration but of course it causes runout as the shaft can bend inside the rubber bushing! He was a machinist and replaced the rubber bushing with brass bushing he made on his lathe. Thanks for the great explanation of hardware and software for milling , cad and cam too. ☺
We are looking forward to watching the deep review of the CNC 4040-XE.
I want onee! May order soon. I want to know the capabilities. I’m a tool and die maker and I want something to play with at home!!!
Hello, just as an FYI if you are cutting aluminum. Use Varsol as a lubicant and coolant. Thats machining 101. Aluminum is very Gummy and a low melting temp thats why its gumming up the tool. No charge by the way.😁 Do not use Carbide with alluminum. Carbide likes to work at high temps not good for alu. Use HSS (High Speed Steel) 2 fllute.
You can do a little more to make this machine much more stable and even more accurate. I filled all of the lightweight Alu profiles (i.e. these hollow aluminum parts) with a lead / tin mixture - you can also use pure cement but a lead / tin mix is better. That makes the whole machine heavier and it runs smoother and more stable and is a little more accurate when milling. The steper motors are powerful enough to handle the extra weight.
That part where you used a frequency generator to "tune" the motor speed was 👌 That's awesome
Thanks for the recognition man. I hope you have yourself a wonderful day.
My first comment to you is about the topic of starting in the logically appropriate position on your workspace (and physically on your work piece). This is your piece of wood or something more expensive (like metal or stone or whatever). I used to have a CNC plasma table and I told the builder of my table that I don't want my table to start in the middle of the work piece because of waste and he changed the focus to the lower left position. I don't remember any problem with +,+ vs -,- because my tool could run a dummy or sample of the program so this would visually SHOW where the head of the machine would travel. This would therefore validate if being too low is a problem. This is the right way to start anyway (in my opinion) I agree with you.
The major reason I was not satisfied with my machine was because of one feature in how it was built. I know now that it was built using the wrong technology (the most important part) that affected the accuracy of the head. The X and Y and Z movement on my machine was being controlled by straight cut gears, the worst choice. No matter how the adjustments were made these gears would affect the quality of the curve or corner of each movement. It was after this that I was then noticing so many ads for so many machines were now becoming available.
Clickbait title:
"How to CNC cut metal for under $400"
Then proceeds to say it cannot machine metal worth a damn even after attempted upgrades.
An hour long video that could have been condensed into 10 minutes of explaining why it cannot
happen with that machine...
Great detailed video. One suggestion. It is not just about rigidity it is about dampening vinration too. There is no enough material in z axis to damp vibration. You will need a aluminium plate to hold spindle clamp. It will improve performance 200%, there won't be chatter and cleaner cut.
This is why i go on with my slow EDM = Electrical Discharge Machining ... no forces at all in XYZ.
May the force be with you !
Where do you get one for under $1000?
I had a cheap 3018 that I upgraded to a 500wat spindle and circuit board to.
And found that harder Aluminum cut better as the softer stuff gums up the tooling and causes the tool to wonder.
if you cant buy harder Aluminum then uses a lot of coolant. don't use oil as it will cause it to thicken up and go gloopy
Also use a single flute bit,
If you have a good cam software that supports adaptive cutting then it will definitely help the chatter problem
If not then look at your ramps
Meaning the angle of let say 2 degrees will help a lot
And be aggressive as to the cut that that the machine can handle and then use a nice finishing pass
In the world of machine shops either the work is held still or the tool is held still and sometimes both. If it is rigidity you want switch all motion functions to the bed to which you fix your work, so the bed moves X and Y and Z and the tool is rigid and solid. Just like the big boys.
I think the gantry style of router is a problem for harder metals.
What i've done is up the feedrate to about 700mm/s, depth of cut 0.1mm. Spraying it with cutting oil, if you don't have cutting oil wd40 will do the trick. Spray some more when you start to hear chatter.
Use a endmill with a coating and small diameter, 2mm would work for your requirements. The number of flutes are the least of your problem because the chatter does not come from the tool itself, i would use at least 2 flutes, but i've used 4 flute endmills with the same setup with no issues, just don't cut completly dry with such small endmills
I have the FoxAlien vasto. It’s a great machine for aluminum so far. Ball type Leadscrews all around. Motor couplings for easy upgrading. Heavy duty Dual linear carriages with rails for extra rigidity. 3 different sizes spindle holders.
Thank you for making such thorough videos consistently
Aluminum is hard to work with. I think you are using a bad alloy. It smears so much. Also you should be using a coolant. Then again you need the right tools to machine alumium. And at the end of the day, the right speed of rotation and the right feed. There are tables that recommend all these settings. I think it is not the machine that caused the problems, but primarily the material that was being used. Greetings from Germany!.
Great point!
@@DesignPrototypeTest Flood coolant is impossible for this setup. Use WD-40. It is a perfect cheapo coolant/lubricant to prevent gummy material like aluminum from sticking to your cutter, which eventually cause your tool to fail. Works very well for me. It works for plastics too, preventing heated chips from binding to itself to form a big blob.
Use a spoon
@@DesignPrototypeTestwho would have thought that if you have 0 idea of machining it doesn't work out first try 😂 and yeah there is a steep learning curve about 40% of this vid is bs
single flutes are superior on small machines
dump the rhino stuff and use fusion like everyone else
and suprise suprise a 400$ mill doesnt perform like a professional one
1310 mills do work, with knolege and limitations,
dont make sawdust, make chips
and so much more
Aluminum is hard to work with? No son, you just don't know what you are doing. Aluminum is easy as fuck. Try inconel, titanium, or any high nickel alloys. Lmao. Aluminum is hard, that's the easiest metal besides brass 😂
You have switched the x and y axis I think. Normally, the X axis goes left to right, and the first coordinate is X. This means the the upper left rectangle is actually (-,+), not (+,-)
As a machinist I work from a center all the time! There are many benefits that I didn’t notice starting out, especially if the part has symmetrical features. For me its easier to visualize one side as positive and the other side as negative in the same direction. As opposed to having it just look like random numbers in the G code. But if the part has a feature like a hole or a bore I like to indicated and set the origin to the center of that!
One of the most real world reviews I've seen, thanks.
I could see (0,0) being a useful center of my piece when its going to be a round geometry, because then I could make the measures for one cuadrant, and then adding negative signs to the X or the Y or both would turn those into Cuadrant 2 - 4 coordinates. At least that's my believe coming from someone who's not done any CnCing.
Moving (0,0) to center of the x-axis, and keep y-axis at the edge of the bed, makes it easier to design bisymstrical models, by changing the X coordinates , to -X coordinates. Place (0,0) at the center of the bed, and do the above to both X and Y, to make quadsymetrical objects super easy.
Thanks for posted this. I wont buy one lol. I started watching this thinking I could give this $400 setup a shot. Not many people post perceived 'failures' but it's probably saved a lot of people some money and head to desk banging.
I'm into one of these as well in fact I bought the 3018 as a test model to design a better one using SBR linear rails and 12mm ball screws and a 500 watt motor......3/4 finished but still on the back burner due to another project taking more priority........I would have to say that the one in the video is totally adequate as long as you don't push it ....scaling is a factor you have to experience and it won't like forcing the cut..........aluminium must be cut wet or you'll always get a cutter weldon with the chips.
I await the upgraded version with a duet board, sensorless homing and a part cooling duct 😛
I am an everlasting font of optimism. I keep watching, hoping it will get better.
Good to see a review on this machine. I didn't see if there were any drivers needed to install the CNC to your computer. Sorry I'm a total doofus when it comes to IT. I think the stage with the drivers was skipped. It's where a lot of people seem to have trouble with the other small machines. They can't find the drivers for their machines for this or that OS. Mostly the drivers are for Windows machines but people may want to use Mac's or Linux computers. I've seen people using both Mac's and Linux systems to control the cheaper CNC machines so it must be possible etc. But as far as I can see it's not clear what's needed to make them work, so the stage of using UGS can be reached.
The Benbox board on the 1310 uses a plug-in Arduino Nano clone.
If it's a faithful clone then it should work out of the box on Windows, linux and macOS using the standard FTDI USBSerial drivers that ship with those operating systems.
More likely, though, it uses a CH340x or CH341x USBSerial chip so you'll need the appropriate driver which you can download for Windows, linux and macOS from www.wch.cn/downloads/CH341SER_ZIP.html
@@anthonyrich1592 That's extremely useful. Thanks. I've got a little Windows 10 netbook I'll be using to control the machine myself but I've seen more than a few people posting about not being able to find drivers. It might be that if this board is easily recognised, it may give the product an edge over boards which need proprietary drivers etc. Hopefully your comments will help others who come looking.
Part of the problem you’re having with cutting aluminum is flex in the machine but part of it just comes down to speeds and feeds.
....And single fluted mills are required.
@@pentachronic yeah i dont understand why people think more flutes cuts better. the more flutes the faster you have to cut with a deeper cut to get the chips the right size. some people think dust is cutting and all producing dust does is wear out bits. i started using all O flute bits for my wood and plastic work and the cuts have been much faster with a far better result
I think machine rigidity is being underestimated in the comments. Feeds and speeds only do so much when the machine allows vibration no matter what.
Thanks for continuing to do videos, I do much prefer your channel, than that other guys.... Keep it up.
You're right about "milling bits" not liking to drill, that's why there are two types, end mills which don't have drilling capability to speak of or swarf clearance, and slot drills which do like to drill, do have only two flutes to clear swarf and also can cut slots full width due to there only being two flutes. Mix those two up and you'll break tools you needn't.
WD40 to reduce prevent that aluminum from adhering to the carbide.
welcome to the rabbit hole that is called cnc-ing metal. I fell in it too. Build myself a momus a couple of years ago. You wont stop upgrading once you see what others can do on youtube :-).. I can recommend the momus though if you like diy. You can build it using simple tools only (I know, I did) and its relatively cheap (In CNC, cheap is always relative :-) ) . Your cnc looks nice too although I fear you will one day upgrade the spindle motor. If you ever decide to spring for one of those chinese high speed spindles, do pick one with an ER16 collet though. Much more choices when selecting cutters as the ER11 only fits a maximum of 7mm and the choice for 8mm and up is really big. Don't ask me how I know...
Thank you. You just saved me a lot of time, money and frustration on buying or DIYing such kind of machine. It turns out that for my needs it's cheaper and better quality just to order required parts (tested with 3d print) in some kind of CNC cutting workshop.
Title: *How To CNC Cut Metal For Under $400*
Video: *It’s impossible to CNC cut metal for $400.*
For real! Every time I try and give this guy another chance, he does this kind of stuff again.. I'm going to just block him from my search and feed at this point.
Actually, he didn't talk about the thickness that works and at which point it breaks.
And why did he keep saying how sturdier this version was compared to the previous one... if he then ended up printing a holder for the spindle out of PLA? Why are we even speaking of sturdiness, making a bunch of changes, if you then use a holder made of soft plastic? Are we kidding?
@@xhorntail would you prefer he lie to you? You don't want to know the truth, and I hate clickbait as much as the next guy, but you wouldn't have listened if he'd titled it "you cannot cut metal for under 400".
@@brahtrumpwonbigly7309 this and other machines CAN cut but not for pro jobs where precision is a must. He printed the holder, because of the new motor. I have news for you, the other one for 1K is also no good too. Must go up in quality, sturdiness, etc. meening >$
Check Alex Moon's videos, he figured out how to mill aluminium with a CNC1310. He uses a much faster spindle. I have the 500W spindle you have and only goes only to 11000 rpm . I think we need more speed.
Thank you for the reference to Alex Moon, interesting. He is using a WindWard S4225-B24 spindle which is around 400$ on its own...
The soft aluminum you used is too soft for proper milling without cooling. I tried that once during an internship with a professional milling machine and it was just chewing gum.
I do alot of aluminium fabrication, I drill, countersink, circular saw with a tungsten tooth blade, jigsaw and router alloy, if I don't use a penetrating fluid or a stick wax it will only get hot and pickup on the bit as you have shown. It's a friction problem. Try slowing down the rpm to disperse the heat, not to slow to overload the bit. WD40 or a low vis cutting fluid. Hope it helps. Like your vid, thanks.
I think you should have tried to mill your intermediate spindle mount out of delrin. Even if you still have some play, you should get more rigidity, equivalent or better tolerances, and a proof of concept for your aluminum design (and you could have shifted that spindle hole a couple of millimeters to accommodate the fan)
Its hard to tell what the issue is without a video of the actual cut, but generally when you see aluminum galling up like that, its due to a surface speed that is too fast, or full slotting with too high of a feed rate. That being said your spindle only does 12k rpm which isn't terribly fast. I would try doing a shallow depth of cut start at .01mm with multiple passes, it looks like you were trying to make these cuts in a single pass, which just isn't possible at 12000 rpm. full slotting metal is difficult for an endmill, and it typically needs to have a machine that's very rigid, has coolant, or can evacuate the chips, and also done at low rpm, maybe around 2000 for a 1/8in endmill.
Also I would highly suggest getting into fusion for this type of work, the cam in fusion is awesome, and allows you to control a lot of variables to the tool paths.
I had that same spindle. It was terrible, the shaft is really thin and flexible. The chatter was unavoidable, no matter what gcode you use and even when cutting wood. When i upgraded to a 1hp watercooled spindle it made a huge difference, now i can cut aluminum in 1-2 mm passes
Thank you for scouting ahead on the path. As someone who hopes to follow, you have my gratitude.
I believe most machinists would not opt to take full width , full depth cuts with this setup. Rather, they might use smaller diameter bits to "peck" at a rough cut of the initial pass till the desired depth is reached. Once most of the material is removed to the final depth, a single "conventional climb" full depth cut would be used for a final, accurate cut.
I really appreciate the effort you put into this video. I feel like a got a year's worth of experience in less than an hour.
"That's a thick member"
"That's what she said"
You can remove the aluminium on the end of a bit soaking in an acid such as sodiem chlorate, can help prevent damaging the bit by scraping.
wait do not throw in the towel.... you are doing great... I can make the adaptive gcode for you, and you can try it again. 3mm single flute looks good but better to enter the material from outside rather start inside, chips will go out faster.
I remember LibreCAD when I played with pen-plotting! Neato
EDIT: I split this watch-through into two segments and it's very informative to see how BAD these things chatter! Wowza
If your bits came from China I can tell you that's gonna be a problem. I couldn't cut crap with the carbide china bits, got a brand name carbide bit and solved most of my cutting problems.
A better way to cut a slot if not to use the wrong bit, use a slot drill with two flutes designed for that job. Same goes for plunging cuts.
There are a few machining strategies that should be used for weak machines/setups. HSS tools, even 4 fluted tools, are preferable to carbide. You want to always conventional mill, never climb. Climb milling kills machining passes when the rigidity and backlash get involved. I feel like a very short 2 flute HSS low helix tool conventional milling with small step downs and high feed would work best here. Also, I doubt know why everyone thinks soft aluminum cannot be cut without coolant, that isn't true. It just isn't, and there isn't much more to say about it. You encompass that factor into your feeds and speeds. Aluminum only gums up if you heat up the material during the cut. If you don't, it will not be a noticeable issue.
It’s normally called a tool not a bit. CNC software has backlash compensation and lead screw compensation. Look at your feeds and speeds. Usually the electronics are in an enclosed box. You need way covers to protect your lead screws. You need a motion control algorithm with acceleration and deceleration.
this video wins most RUclips adverts ever
At 17:00 I believe you have reversed the (X, - ) and ( - , X). The horizontal line is X axis, and the vertical line is the Y-axis.
Thank you for explaining things in a way that most don't care to
Hi Matt (is that your name),
Bam, I subscribed. Great video. It shows the value of failure for education, which you make entertaining too.
If you agree (that failure is important for success), then I’d ask you to carry on with this project and not give up altogether.
Instead, learn from the failures, take on some of the points made in these comments above, about topics such as metallurgy, coolants, etc and continue to explore what’s needed to get this machine to do the job. You’re far from finished with it - I hope.
Looking forward to an update later this year 😄
Hey man, what a great video packed full with quality information. Maybe you said it and I missed it, but did you use conventional or climb milling? Or both? The difference in outer dimensions vs. inner dimensions at around 41:40 could be explained by the mixed use of both milling types. I'm sure you know this already but climb tends to push the bit away from the material whereas conventional tends to pull the bit into the material. Also spring passes (100% overlap with the previous toolpath) could be used to get better dimensionality.
At 43:14 you mention the problem you're having with chatter. In my experience with a similarly built cnc, it helps to experiment with different (scary low in my case but that of course depends on the machine) spindle speeds until you find a speed where the frequency of chip creation doesn't positively interfere with any of the machine's resonance frequencies. But that quickly becomes a shitshow ;). As for the optimal (literature) feeds and speeds: I tried them and they didn't work. My gut feeling is that those work on tormachs that actually have the required rigidity for those settings but they are incompatible with hobbyist-machines.
To the cooling: I tried compressed air, but the airflow required to cool my parts was so high that it would bend the bit/shaft/motor-mount which would create even more heat and chatter. I had good luck with isopropanol as a cooling agent because it evaporates really fast and doesn't leave a mess.
And about the adaptive clearing toolpaths: With a small workaround you can get Fusion360 to do those perfectly fine. Even with the free version afaik, but I understand that Fusion isn't as cool as it used to be due to their new pricing model.
The last thing you shouldn't forget about is that aluminium is terrible at absorbing vibrations. It's light and stiff which is great because it will make the machine light and stiff ;) But at what cost? Professional machines are made mainly out of cast iron in part because of its dampening characteristics. In this case I think dampening the frame by different means might be an option. The easiest idea would be to print 0.5 to 1mm spacers out of petg which also has great dampening characteristics and put them in between each of the aluminium-aluminium joints on the machine. For that same reason, I think that those cnc machines like the foxalien or the shapeokos with openbuilds-styled delrin or nylon rollers are a great compromise because of the dampening characteristics of those rollers.
some of the larger hobby machines are definitely capable of cutting aluminum but you really want at bare minimum 1.5 kw spindle, 4-600 oz in steppers and a machine using cast parts and not extruded parts. cast iron is better than cast aluminum. but then youre talking a minimum of 2-3000 dollars for a machine like that. im building a gecko g540 machine using a completed chinese machine with cast iron parts and gutting it of all crap electronics and shitty wiring. but people hear CHEAP and they think wow i can do all that for $500 and its just a lie. I have over a thousand into my what started out as a 3018 that cost $300 and its capable of cutting wood great but wouldnt attempt it on any metal and the size is just too small for practicality.
I submerged my machine into coolant and no noise, no chips everywhere, no sticking of aluminum to bits) Thought these days I just use 3D printing
there is a button to control the spindle, just check the setting and enable the panel :)
I also prefer a positive only coordinate system. With a positive/negative central location system, if you get distracted for a moment, and you forget to sign your command, well...
I would really only recommend conventional cutting with these machines. The rigidity is nonexistent. The motor itself isn’t meant for this. The new motor is much better. But the machine itself isn’t capable of taking usefully deep cuts for thicker metal. The new motor is far from the rails, giving it a greater ability to bounce up and down. So some of the chatter is due to that.
The fan is wider than the motor body so that the flow moves around the sides of the motor. When the fan is the same diameter, you get a very turbulent flow because all of the air is now hitting the flat rear of the motor. Assuming this is designed properly.
You know that aluminum sticks to cutters. You need bits coated for aluminum. Since you know you need coolant, I’m trying to understand why you’re not using it. I often use water diluted sawing coolant for much of my work. For this tiny machine, a few drops here and there is all you need. Why not use it? A slight mess is part of the game. Blowing these fine chips away is not the solution.
I see at the very end you discuss that motor overhang. There is no way to eliminate that. My machine weighs around a ton, and it’s considered to be a light machine. Chatter still happens, though with much heavier cuts. I’ve played around with a CNC Sherline, and with anything more that a 0.002” cut, it vibrates like crazy. I don’t know how anyone gets anything done with these little devices.
Fusion 360 will do the adaptive toolpaths...I would try a 3 flute endmill also- less thump.
3 flute is way too big for a machine this size. youll destroy bits like noones business. the more flutes you have the faster the spindle has to spin and the faster the machine needs to travel to get optimal chipload. a 1 flute or O flute works best on these smaller machines. if you produce dust you are burning bits if you produce chips the temp and friction on the bit is far less and the bits last 10x longer. also never buy crappy chinese bits. if your bits are less than $15 ea your running crap. i run Yonico on the cheaper side and Amana on the higher end side depending on the job determines the bit i will use.
Design Prototype Test, I don't know if anyone committed on this before, but I want to let you know that you made a mistake when explaining the Cartesian coordinates. The top right is (+,+), the top left is (-,+), the bottom left is (-,-), and the bottom right is (+,-). This is not criticism, just clarification.
On the other hand, thank you for your very informative video.
The Path workbench on Freecad has the adaptive clearing feature you were missing in this video. I'm not sure if they had it already when this was filmed, but at least they have it now :) Ever tried it?
Great Video by the way. I learned a lot listening to your comments throughout the video!
Adaptive tool paths were available in FreeCAD 18 so I think they should have been available when this video was filmed. The only slight gotcha would be that adaptive profiling and tabs don't seem to play together nicely in FreeCAD (applying tabs to an adaptive profile seems to send FreeCAD into an infinite loop when I try it). The work-around would be to use adaptive to get most of the way down then do the last bit, with the tabs, using a normal profiling operation.
I do agree with you that it's a great video and I'm sure many people will have found it useful.
I agree with you on the lower left of the machine should be 0,0 the only reason I can think that they would make the center of the cut area 0,0 is their programer doesn't have any real world experience. Most drafters, and other trades people were trained to use 0,0 as the lower left. in the end it doesn't really mater, it still math but it would bug me too!
I don't even do any of this stuff and I found this fascinating.
Use acid to remove the aluminium from the mill bit if it isn't broken, and will be as good as new..
I suggest wiring the 500W motor as if it was direct mains connected - mains rated wiring, wiring protection and earthing of the motor and metalwork. I bought one of these sets a few months back and the power supply doesn't provide mains isolation.
Thanks for the tip
Was a really good video going through so many different aspects of your decision making and the basis of a good cnc, to software etc. felt the pain at the end. I think we all want a working machine ideally cheap to buy. How would you make your cnc or what do you think needs changing in your final machine . Your spindle holder was not great being short in height. maybe worth revisiting when the pain has gone.( not to make us want to buy the machine but build it.) See Aurora Tech for 3018 based/robo cnc metal cutting though limited.
Please consider using her spindle holder and/or drilling holes to mount your spindle to see if it helped.
Did love your video - was just really sad for your hard work ending with disappointment and you seem to have abandoned cnc.
@43:50 i can explain you difficulties. The bouncing back and forth is because of your motor holder it's made of PLA it not only gives in the long run but it has e very elastic deformation E-module all the vibration and forces bend the motor holder elastically.