I wish i had all the skills you have. I've watched 3 videos on this machine so far, and this is amazing . I wish I had the skills and space to build a CNC mill of this size and capability. For a homemade machine, this thing is absolutely amazing.

I had no problem. Got the notification of the original video, and watched it in its entirety.

Could not wait for the next video after seeing the cnc build… what a weath of knowledge packed into one video. I love the use of scrap and discarded material in your projects, making somthing from free stuff is always that much more rewarding. Again, wonderful video.

Friend, thank you so much for watching. Part 2 is here! kzfaq.info/get/bejne/hZqjfbd0yZjRmZs.html The 3D model of the cam-clamps and a bunch of other stuff is here (it’s in the shop but free): ko-fi.com/floweringelbow/

We do know how magnets work. Its the alignment of the intrinsic angular momentum of electrons aka spin. But this is certainly not common knowledge. Your videos are amazing and you’re a genius!

Missed in your absence ... busy Summer? Good to see you back!

So invested in your stories. I can empathise with all of it.

You're at 3N/micron stiffness...which is actually really good for a router and certainly capable of cutting steel at reasonable feeds and speeds :) well done!

so glad you fixed the glitch! i've been in similar situation - this stuff can get really frustrating.

I remember my first convertion on a suregrave engraver to Mac 3 and having so much joy when I got everything dialled in to the sub mm on the steppers.

a tip for your cooling tip... you can use cheap needles with leur lock on your mister. Just 3d print an adapter to screw on the flex hose thread. Then you can change the needle gauge out easily and you can buy a pack of various gauges from Amazon. This can reduce the CFM and give you better control.

Dude, I frickin love you man!

Yay, an update! Love all your projects!

Congratulations with getting it to move somewhat reliably. Your concrete dampening inspired me to make my mightymill which is UHPC (durfill concrete) filled in its extrusions. Awesome content!

Love your videos, thank you.

Love it!

you should use these round cam clamps inside an outer square frame with the round hole in the square towards one of the corners. you then get a cam clamp with a a flat edge and 4 different thicknesses, Ive made some out of wood and they are soooo solid.

I find you incredibly talented, to have abilities from rough farm work, tree felling all the way to diagnosing cnc errors. You should be very proud of yourself. Although I suspect you're quite a humble guy. Just take the compliment, well done mate.

"...Someone that can read the chips..." Pure gold.

Stefan Gotteswinter was showing using Isopropyl Alcohol as a lubricant/coolant for alu machining. it has the advantage of zero residue (not like the crap coolmist leaves behind), doesn't affect MDF too much, but requires good ventilation as well as spark control.

I feel your pain with the missing steps. I built a CNC using very cheap drivers. They seemed to work but I started using it to do some complicated carving and it started losing steps. I chased it for a while but ended up replacing all the drivers and the problem was resolved. It was a bummer to have to replace all the drivers, but a great relief when the machine stopped losing steps.

I'm glad to see you back to putting up videos again. I was afraid there were no more. I wanted to know how you are progressing. I'm sorry that you have had those problems. I'm glad that you were able to solve the positioning error. I will need to view the video again because I did not catch what was the cause or the solution. I want to make my own CNC router. I had planned on making a smaller, less robust implementation, using unsupported linear rails, in order to get something to play with and learn from. I have used Japanese THK linear rails (34" and 54") that I obtained from dismantling an old surface mount pick and place machine for the planned 2nd machine. But last week I picked up four 26" THK rails for a great price, so now that throws a monkey wrench in my plans for the quick and dirty first machine, by using the new high quality supported rails. It changes the structural design. I'm thrilled that you are posting videos again.

13.50 is nothing to do with the part getting hot . you use an upcut bit that pulls the part out of the wooden socket

OMG EPIC GLITCH HUNT 😂 U have the patience of a shaolin monk ❤😊

Very cool machine. Glad it's working so nicely. With drill cycles in ali, i use 60 m/min surface speed. Feed per rev is drill diameter/150 and use a peck depth of drill diameter/20. This way you don't get long stringy chips. Also use a centre drill that has a wider point angle than the main drill. I have one with a 142 degree point. This means the main drill always centres to the centre drill spot. With regards to chips sticking, don't use end mills that have a TiAlN coating as the Al in the coating attracts the Al being cut. Use uncoated for ali. Well done again!

i was wondering what those "bean can alloy discs" then i saw the mounting thingys they became wonderfull. so cool

Raise your workpiece up from the table to keep the cutter and the z- travel as short as possible. At 12:50 your cutter is blunt, and you are polishing the surface with the end mill. This will not be a flat surface. Your incorrect position at the end of a long program, was missing (or gaining- unlikely) steps. The faster your cutting feed the more stress you put on the machine, and the more problems you will have. Slow your feed ( I know you have a high speed spindle) and use only carbide cutters. Hss tools will blunt very quickly if your rpm is too fast 😊

Still my favourite CNC machine on KZfaq. Have been waiting patiently for your next video and hoping it was CNC machine content. Success. Glad you got the settings dialled in. I reckon MAC BOUNCER is right about the rails, but if it's doing what you need, all good. Wood vice had no show of holding that smooth surface, but if the bottom of the disk had a lip to cut into the wood on clamp it'd hold fine. Machine the lip off afterward. Someone else I watched recently machined button cap screws eccentricly and they rode in a pocket rather than acting directly, so no "bug" so to speak.

Another huge comment but I feel your pain here, I started with cnc diy retrofitting a brigeport interact1 devoid of electronics, and the lack of a enclosure and closeness to the tooling but with cnc speed after manual machines was scary, I did mist coolant but disliked the fog, fitted workshop extraction, then flood coolant with suds. Which without a enclosure meant getting covered with suds, waterproofing sensors/wiring etc. Missed steps drove me mad, but mine were the parallel port not being fast enough to keep up. A mesa FPGA motion controller fixed that as then the pc just fed the mesa the co-ordinates and offloaded dealing with motion to it. Also chip removal & packing in aluminium is a issue unless your stood there poised with hoover tube babying the part as youve found, and steel is even more dependant on cooling. I eventually made a telescopic thick perspex enclosure around the table itself but had to put lots of M00 in the gcode to come in and clean out the chips periodically. Also using the original machine spindle which was QC30 was a mistake, so I was the ATC watching for the toolchange dialog box to pop. I went to auto touchoff cycles working to a sensor on the bed so it touched off every tool on change and put the offsets in for a cut. Limited Z axis with the cnc quill but manual knee which to fix would mean bringing the knee under machine control & the complexity of a summing Z axis and controlling same in gcode. Without this when it came to using very mixed length tooling in a single program risked crashes with clamps etc and a permanent juggling act. I still use the bridgeport, occasionally with a rotary A axis made from a old lathe headstock mounted on the bed which is magic for cutting gears, yes it has inherent limitations & I boxed myself into several corners but really it was about learning and doing it all myself, which is fun and has a value you cant put a figure on. We're all just working on some bigger project ultimately, ourselves. For my limitations eventually the answer appeared on ebay, a Cincinatti Arrow 750 VMC for 500e, enclosure, 21 postion atc, pumped lube, suds, chip extraction auger, huge travel etc, all the things I ultimately wanted but didnt realize at the begining. Cheap because nobody else bid on it because it was so difficult to extract & absolutely had to be out by 2 days time but theres always a way when you are mad enough. For now it all just works and is lovely and can just rattle off hundreds of something unattended in that soul-less industrial metal machine eating way, but when it breaks expensively I'll retrofit it with linuxcnc using lessons from the bridgeport. Maybe make a diy trunnion table for it, turn it in to 5 axis. Whatever. I wont be scared to tinker and I dont have to worry about downtime not paying for it.

About glitches: make sure that you use a good heavily shielded cable for the spindle with shield connected to control box ground and not connected to the spindle itself. That will eliminate a bunch of other potential issues in your system due to noise from VFD.

Hi Chris. couple of suggestions if I may; measure the deflection at the tool tip not up the spindle! :-) The reason you were getting better numbers with the Z axis raised was because of the orientation of your linear rails... They have the highest load rating through the top, then the sides, then the bottom. so the Y axis rails facing forwards is one of the weak points of the design. (its a common error, a lot of so called quality makers do it the same way. the rails should face up and down to get maxximum rigidity and load bearing. That's harder to do right, so they do it the easy way.) have a search for "sweet dreams CNC" you'll see what I mean. You are correct in thinking that ideally you want the footprint of the bearings within the footprint of the other bearings, in your case, raise and lower the bed while keeeping the Z axis bearings within the Y axis footprint. Done this way, 5 micron deflection is a realistic number. I know because thats' what my design does.

I need to come learn at your studio

you are just awesome

You need to use polished HSS for aluminium alloys, any coating such as the black oxide or TiN makes it more likely there will be a build up on the cutting edges.

You may add a flat part to the release side of your cam clamp disks.

Love the videos another great one. Don't know if you've used them but shopapt is where I get all my tooling. Really high quality at an amazing price. And the holders are balanced.

If your metal is super hot after milling your probably rubbing to much and not cutting is the way I understand it. The cutting of chips is to remove the heat, so you need to adjust your feeds and speeds even after hours of cutting I can place my bear hand on the metal and barely feel heat .. but I am also knew to cnc .. great work though👍

this porosity was the result of a small riser, which was unable to store the gases generated in the foundation of its part, to solve this in the next foundations make the riser bigger

As a machinist you want your heat put into the chips and not the cutter or part. Blue/golden chips are usually perfect.

Something went wrong with your upload. I had to find this video manually after receiving the notification

Have you heard of the concept of heat partition? It has to do with how the chip is formed and how the heat is split between the work, the tool, and the chip, and (I suppose) the coolant. I've never gotten to the bottom of it but I have one of those slow-speed carbide circular saws for cutting steel----the chips are blue but the work stays relatively cool. I've been thinking about heat partition ever since.

I Use denatured alcohol to machine Aluminum and other non ferrous metals. I probably get about 12 hours of machine time if not more from a gallon. Since it evaporates so quickly there isn't any mess afterwards. Just need to make sure you keep steel away from it. If it ever catches fire it burns clear. On hold downs... Don't waste your time with cams and blah blah waste of freaking time. Go vac hold down for bigger parts or vice for smaller. Heck even simply using masking tape and crazy glue is good enough for most metals. I used the tape and glue method to make one of my vac tables. Having reliable hold down is 90% of the job. Trust I've been there. It's cheap and easy. Get it done.

Very good job, you remember me to Marius Hornberger facing with similar problems. Could be of help to watch some of his vids.

Hello there ! Very cool video. Did you also do a conversion on the jcb or just mixing with diesel ? I've watch a lot of different setup for filtering i'm curious to see your take on that. I did my first recon mission for WVO last week, I might be collecting soon ! Best wishes for you too, cheers ps: congrats on the 50k subs, well deserved

Great video - thanks :) I like the way you explain things. So what was the root of the problem? I thought it might be the relation between the step and direction pulses when the direction changed. (I do know the joy that you get when you send the machine off to do loads of freaky forwards and backwards stuff and it comes back to 0,0 afterwards! )

Do you have an EMI filter installed for the spindle power cord? I absolutely love this channel 😁

This looks great! Just wandering... Q: Would submersing the workpiece in coolant / cutting fluid be possible / helpful? Or would it just exchange one type of mess for another...?

excellent project well executed, however I didn't understand what the problem was in the end. I had a similar problem using cheap motor drivers. When I upgraded it went away.

So it was a timing issue with step and direction on those drivers? How did you fix it?

Comment for the algorithm

You are taking a much too conservative cut in terms of chip thickness if your part is hot (or tool is dull). The chips need to be thick enough so they take most of the cutting heat.

precisely because you cannot mill in steel and certainly not with carbide end mills, high stability is required

Part 2 part 2!

Hey, Mr. Elbow. So I'm planning on building my own cnc router and I really like your gantry design but I've got a question and it'd make me quite happy if you could give me your two cents. I want to use aluminum extrusions similar to how you've done your gantry, but I had the idea of filling the extrusions with ultra high performance concrete and using pre tensioning or post tensioning techniques to stiffen the member. Does this sound like a viable option?

And I thought my cnc was heavy. Mine is 12 ft long, but I bet yours weights more

Hi, I noticed your Walls in several videos, did. you finish Them with clay? Did you Build your House with clay so it breates. That Might be an interesting Video

Why not flood cooling?

You need a second, slow speed spindle!

The heat build up and the chips you are getting make me think you are taking too fine a cut and getting a lot of rubbing. Are you using a feeds and speeds calculator to work out what feed-rate and RPM gives you a chip thickness of 0.1mm? I suspect the RPM you are using is far too high for a large cutter. Rubbing will wear your cutters faster as well. My feeds and speeds calculator gives the following for a 4 flute Ø25 HSS cutter in aluminium 1940 RPM and 80mm/min and 0.01mm chip thickness.

I don’t drive 😂😂

The chips tell me there is a romance coming in your life. Oh wait... That is the pattern for tea leaves... Sorry... I do not think they use the same patterns.

I figure if you can tune servo driver vs drivers, would you please call Putin and end this insane war.

These videos are making me hot and bothered

1. MQL is cool ... and an excuse for a design flaw: not considering the flow of the coolant ... where it should go. Like a massive and for that purpose optimized machine bed (cast iron for small ones, as a screw-transport sump-system, including for chips, for bigger ones). Yeah, fancy. But all that counts is Kelvin per Watt dissipation:P 2. In professional machines there is always some -cm- room sideways X/Y or a special place up there in the "machine heaven", where the tool-changer can resist. That is ... not something bad (when it is designed in from the beginning). This also leads to completely separate tool-change compartments with spindle-cleaners and fancy doors (... see? I'm not against fancy stuff:P). 3. Great response from your spindle guys (YOU CANT DOOOO THIIIIIIS, hehehe). This kind of caring is something that money can't buy:) 4. Okay, not really. I hope you wear your safety-googles when performing such stunts (sure you do! ... I'm just egoistic, wanting future videos!:P) 5. Okay, really. You can buy that. Greetings from Germany, neighbor: Siemens ... Hehehe 6. Thanks for the exciting video (I always loved Sherlock Holmes, Miss Marple), for the "machining thriller":) Edit: reside, not resist ... (my full blown bias up there **g** )

Horizontals ftw

Hey check out youtuber "Marius Hornberger" they built a very impressive automatic tool changer for their CNC. I think you could get some great ideas for your setup from the work they did.