12x24 Primo

One of my very favorite CNCs was the teeny tiny burly I built a while back.

It was SO much fun just mowing through whatever I put on there. But it got me thinking, maybe I’d gone TOO small at 8x10.
Then along came the primo. 2x4 was much too large for it. Super convenient to load up 1/4 sheets of material, but too many jobs where the flex led to increasing chiploads in later paths and ultimately skipped steps. So I’ve taken the opportunity to shrink him down to 12x24.

1/2 inch torsion box cut out on the other CNC.

I also cut out the base on the other cnc so I can try out t-nuts. It’s definitely not a spoilboard. Replacing it will require removing the cnc. My idea is to use a separate spoilboard on jobs that require through-cuts. I should be able to cut them out of MDF to line up with the holes in the base, so that should be fun. The odd pattern for the outer most set is because I didn’t want to buy more and that’s all I had in my drawer. I also cut holes for the legs because every MPCNC I’ve built took me WAY too much time and effort to place the legs (except the teeny tiny, where I used this same strategy).

But somehow or other, I goofed up the cam and the larger pocket for the tnut was just too small for all the “interior” nut locations. I set the plunge router and spent 15 minutes opening them back up. Since the center hole was located properly and sized correctly, I just needed to make room for the nuts to drop all the way in. Beauty didn’t matter because this was getting flipped upside down anyway.

For the legs, I left just enough tube sticking out to register the holes in the base. I don’t expect any significant extra rigidity from going the full 1/2in deep, and on the teeny tiny, they actually made it difficult to get the height right. A little gap at the bottom, a little gap at the top, and all the leg heights are defined by the printed parts.

I used some scrap to dial in the squareness. I usually just bore some holes and stick my calipers in them. This time, I decided to use some of my many broken endmills in the holes as pegs instead. I’m not using the calipers correctly here, but registering them on the scale like this makes sure that I’m consistent from one diagonal to the other and not introducing false error by being a klutz.

Here he is at home. I mounted the cable chains higher up so I can pass work through without catching on them, and I’m preparing for some air blasting with that white hose. I made some handle holes in the sides for when I have to move it, but since I’ll usually be able to ask my wife or kids to grab a side, I should have moved them closer to the edges. I also should have designed a pocket in the torsion box for the electronics. I had just BARELY enough room to mount them to the back of the box, and the base board hangs over them just an inch or so. It’s adequate. The cables come up around the left side to plug in to the PC. Set up with Grbl Mega, didn’t realize that it’s based on 1.1d (at least the version I got is). So far, no real problems with that so I’m not going to bother with it unless that changes.

My first attempt at wood errored out, and I can’t find anything in the code. It MIGHT be related to rounding in the arcs throwing off the error checks, but it ran fine on the other cnc so I ran it there (they were paid pieces). Sadly I didn’t have any other wood jobs to run, so…why not try to get after some steel? Details in the next post…


Let’s start with the obvious. I’m not a videographer. I’m not a machinist. Yes, this was REALLY aggressive.

I figure as long as things don’t break right away, I can figure a reasonable recipe and tweak from there. Plus, it’s a lot easier to convince myself to back off when things aren’t going well than to get more aggressive when they ARE going well. I tend to be more of a “If it ain’t broke don’t fix it” kind of guy.

Endmill is SpeedTiger coated 4fl, and they suggest something like
13k RPMs
0.375" DOC
25% stepover
10.6k RPM
0.050" DOC

Here, I used:
16k RPMs (Limited by Dewalt)
0.18" DOC
25% stepover
16k RPM (Limited by Dewalt)
0.0003"CPT (50%stepover for extra roughing pass)
0.010" DOC

I’ve got some small aluminum parts to mill coming up, so I’ll post those here when I get to them.

Ultimately, I think ramping down the tabs was too much, but there is a small possibility the endmill just started to pull down because it was so short (see the end of the video) and caught on one of the tabs as it backed up to start a new ramp.

It was really taking some bites out of that steel on the adaptive! Then the endmill just…broke. I had enough to grab on to, and the flutes looked OK, so I chucked it back in and started over!! Went really well for a while. Slotting took time but no real problems. The little chips were super hot when they hit my arm but the stock didn’t go over 105F and was usually between 90-95F. I thought I was showing my thermometer to the camera, but remember what I said above? Yeah, sorry.

I really didn’t expect it to cut far at all, but I think taking adaptive stepdowns to 50-80 thousandths is a good idea, and the RPM/Chipload seems adequate, given the final condition of the endmill. I think I always knew I was going to be better off with adaptive over slotting, but I really wanted a torture test here (and I got it).


12x24 is what I am building right now as well. Need to wire and belt it and it is ready.


DUDE! Another build, and that shop…awesome.

Is that other CNC self-made or is in a printnc? Either way, I have never understood the reasoning behind the shape. Why is the single rail axis the largest, in my mind the dual rail axis should be the longest (twice as strong) yet every time I see one they are the way you have it?

1 Like

It’s a printNC. The idea behind the orientation is for passing material through to the back. I almost went the other direction because I agree with you 100%, but it had to be 2x4 because I’m doing a LOT of work that size, and it didn’t seem like much fun to access that way. The Z is the weak point (at least on mine) and allows chatter way before the long axis becomes an issue.

As for the shop…it’s still just the garage at the rental place. The new build keeps getting delayed because people don’t want to show up for work. Everyone keeps asking me if it’s “supply chain” keeping us from getting stuff, but we have all the stuff. I need to update that thread with some pictures, I think. We’re so close I can taste it, and we just found out the framers f’d up some doorways and the doors that were ordered don’t fit.


And honestly, the only reason it LOOKS awesome is because I stumbled on MPCNCs a few years ago, lol.

1 Like

Oh that video is stupidly impressive. I love it. I was messing with more steel last week and realized I have to have at least good air, but went and got a cheap mist system, I hope to give it a shot soon!

1 Like

I’m seriously impressed with the primo, dude. Full disclosure, even on the teeny tiny i used a much smaller stepover. I only started with that because it was listed on the manufacturer sheet and I’m kind of a happy idiot.

I had a lot of trouble before with mist on the steel. Uncoated tooling, but the chips kinda stuck to everything when they got just a little wet, and started clogging things up. Maybe I just didn’t have enough pressure, who knows?

Here is the tool in the video. AlTiBN coated. I also tried a kyocera 2 flute AlTiN coated tool and it was good, but twice the cost. These showed up in my search list and @Joseph had just said good things about them in his transmission adapter thread so I went for it. The coating is specifically recommended for cutting dry, and I like them. Hell, i chucked the broken mill back into the router and almost finished. I only had a few psi coming from that blow gun in the video.

SPEED TIGER ISE Carbide Square End Mill - Micro Grain Carbide End Mill for Alloy Steels/Hardened Steels - 4 Flute - ISE1/8"4T - Made in Taiwan (5 Pieces, 1/8") https://www.amazon.com/dp/B07FSJLMJ5/ref=cm_sw_r_apan_glt_fabc_5SDXB73Z5K647XWJDNWE?_encoding=UTF8&psc=1

1 Like

Well, i took what I learned yesterday and ran some more. Sadly, i crushed a tool with some WHACKY feeds right off the bat. Somehow managed to more than double the chipload. Not good.
Second attempt went well until it didn’t. Couple restarts because I couldn’t be sure whether I’d lost steps in an inside corner, but replaying the video just shows how much things flexed.
My last idea was small stepdown, big stepover, but after a helical entry, it retracted, plunged, plowed full steam ahead, and before I realized what was happening my last tool was broken.


Awwww bummer.

With such high rapids I would suggest turning down the accelerations a bit more. Looks a too jerky to me.

1 Like

You are still doing some seriously impressive removal of material. Have you tried climb vs conventional. Looks like all of that is climb milling?

1 Like

It is climb. I think conventional might fix that grabbing on the inside corner of the slot, but no way to know for sure.
I think I’m disappointed in the WAY this failed. First and last breakages were absolutely avoidable. It wasn’t from pushing hard, it was from being in hurry with the toolpaths, and only looking AT the things I was looking FOR.
I’m not crying though. Still went way better than it should have.
I may have to beef up the gantry tubes at some point. Some of these tubes were left from the teeny tiny, but the two stainless tubes were from another build (lr2 maybe?).
It definitely flexes more than the teeny tiny did, but the plastic parts are all tighter and better - assembled. It would have been nice to try this job on the teeny tiny for comparison. I did quite a bit of adaptive steel on that, just never any actual parts. All messing around with scraps telling myself ‘one day I’ll be ready for car parts’ lol. But I don’t think any of those were quite this long.

1 Like

Also, a couple aluminum parts for the car. Once it’s assembled I’ll post it in my off-topic, but don’t hold your breath.
Worth noting that BOTH of those endmills are old and beat up, and the only reason I didn’t toss them was “just in case I don’t get around to ordering more”, and of course…I haven’t. The first one was so bad I pulled it out after the first pocket. The second was almost as bad, but at least it wasn’t scoring the wall (as much).
I wanted a press-fit on the bearings, but I really need better measuring tools. My HF digital calipers really helped me dial things in nicely, but after a couple years, they’d just cut out when opening and closing them. I got a replacement from Amazon and I’m going to commit to just using them for woodworking projects. I went from “no way in heck will this fit without the hydraulic press” to a really nice slip fit. Oh well, I’ll powder coat them and see how much of that clearance gets taken up. Won’t be any pressure working to slide the bearing out, so it’ll either fit just right, or I’ll epoxy it in.


You seem to be loading those up pretty good, have you tried 1/4"?

1 Like

I’m seriously thinking about it. On one hand, they definitely cost more. On the other, if they don’t break I come out ahead.

I really want to try the low doc again (without going straight into a slot, lol). The biggest reason I went with the 1/8 was to keep the surface speed low, but I’ve seen a couple videos where people are ripping through stainless at 15k+ with an uncoated single flute. That right there boggles my mind. I don’t think my primo is rigid enough to keep up with that one, but lowering the doc might still make it an ok strategy for bigger tools, too.

I don’t really have any other steel parts to cut, though, so it’s just this clamp for now, and the slot is 0.3 in, which is a win for the 1/8 size. 6mm tools are pretty cheap, so I’ll have to see if i can helical down into it. Or maybe a very gentle ramp would be good enough.


I swear there is something wrong with me. A rational person would spend the $15 on a proper coated 1/4in mill for testing, after seeing how well the little ones did. Me? Noooooooooooooooooo…
I have this really good uncoated 3flute endmill that was made to cut aluminum and shares ZERO characteristics with anything I’ve actually seen cutting steel. Let’s try that!!

In fairness, though, I have seen people blasting single flute uncoated mills through stainless at really high RiPpeMs, so why not?

After this, I think my machine is rigid enough for 0.015"-0.02" DOC all day long. The router had to do some work on the big stepover, though, so I think the MRR is about maxed. The coated 1/4 mills I found are all 4flute, so I’ll have to take the stepover down if I get around to testing them. Also found some 2 flute, but I’m not sure how I feel about that yet.

All things considered, this endmill did a really good job. I think it got dull at some point, might even have been my fault for messing with it but it still looked pretty good at the end. Might have even been OK running dry, but I’ll never know.


Since I’ve got so many fail videos up, here’s a refreshing change of pace. All the failure doesn’t bother me because I know it’s just me pushing and pushing, but just in case anyone started feeling sorry for me, I made this. I’d already cut two when I realized it would be a nice video.


Forgot to include the picture!!

As for the Primo, there is definitely a larger difference in capability vs the Teeny Tiny Burly than I thought there would be. The only other difference is that here I have the 065 stainless vs the 125 DOM on the teeny tiny. I had no problem adaptive clearing 1/4 plate all the way down before, but here I was limited to about 0.06". I could still go fast enough to help make up most of the MRR, so maybe it doesn’t really matter.

I also really don’t like the T-nuts. For starters, the machine I made the holes on wasn’t as square as I thought it was so I won’t be able to use them to line anything up like i’d hoped. But man alive does trash fall right in there! Also have to pay attention to where I drop the workpiece so I can get good enough clamping pressure from each side. I’m gonna have to figure out how I want to put t-track back in there.


So nice! That surfacing really cleaned it up!

1 Like

Thanks. I engage a lot in the aluniminumian threads because I cut a fair amount of it a while ago. This ought to help my street cred.