[jid2]

^^^ It's funny because it's a switch I have to forcefully turn off. I want to try and run everything at full tilt, often just as a chance to learn more about the "limits". It took some work to make myself turn the feeds down from 90 IPM to 60 IPM and then I also started with 3 of each part knowing I needed one good set.

Trying to figure out a whole bunch of what if's, but this specific project/product is looking to be the green light for a new VMC in the garage. Looking at new or used Super Mini Mills, Sharp 2412's, and Hurco VM10i. I plan on creating a typical part with roughing and then surfacing, with a hand full of tool changes and then seeing what the cycles times are for each machine - running the same code. I know the new Hurco control runs HSM and surfacing toolpaths really well/fast. I think Haas with their HASM module is just OK, and Sharp with the basic Fanuc control I fear will be the slowest. But I really want to see how they stack up.

[rlockwood]

Our Mori's are all capped at about 240IPM, and I'd bet most fanuc's are about the same.

Keep in mind, alot of "high speed" toolpath speeds revolves around the actual bandwidth available to the control (processing power.) So make sure you're using tool path smoothing and have realistic accuracy settings, or you wont get a realistic result. Data starving can be a PITA.

Also keep in mind, speed and accuracy go hand in hand, so faster isn't necessarily better. The reason the Fanuc will be slower is because its smart enough to know it cant hold position faster. Unfortunately, they're not necessarily smart enough to realize that accuracy doesn't always matter. Also, I should say "out of the box" the fanuc will be slower. They have settings (not just parameters, actual user settings) to allow switching accuracy priority- I rarely make use of them, we don't do a ton of surfacing, and 250ipm is generally fast enough to allow usage of all available horsepower.

There's always Fadal's available, and while they wouldnt compare to any of the machines you're looking at (maybe the Haas, but about half the rapids), but they're considerably cheaper

[jid2]

Rob what do you use for your typical tolerance and smoothing settings? I seem to plug .0004" into both for everything. But for roughing I could likely open those up. I do seem to remember reading that keeping those settings tight can sometimes make for less code. I know that I need to get a good handle on how much I can let the machine and the code deviate, make sure I leave enough stock to accommodate that and then let things rip. In aluminum with a 3 flute 1/2" tool running 1.0-1.5" deep at 10Krpm what sort of stepover and feed should I expect to get out of a Super Mini Mill or a Hurco? Seems like I should be able to run between .10-.20" stepover and 100IPM maybe 200 IPM? What about surfacing with say a 1/4" ball end mill at 10KRPM, what sort of feedrates do these real machines want?


I know that the Haas control has an M code that will toggle the machine into the high speed mode where it gives up some accuracy to improve speed. I've talked with the HSMWorks guys about creating a post that toggles that M code when using adaptive clearing so you leverage that for roughing.

Also, have you seen this video of the Hurco control. Same code, generated by Solidcam - but the new Ultimotion is better suited to deal with the acceleration and data rates.

CNC Machine Technology Reduces Cycle Time | Hurco UltiMotion - YouTube

[rlockwood]

Rob what do you use for your typical tolerance and smoothing settings? I seem to plug .0004" into both for everything. But for roughing I could likely open those up. I do seem to remember reading that keeping those settings tight can sometimes make for less code. I know that I need to get a good handle on how much I can let the machine and the code deviate, make sure I leave enough stock to accommodate that and then let things rip. In aluminum with a 3 flute 1/2" tool running 1.0-1.5" deep at 10Krpm what sort of stepover and feed should I expect to get out of a Super Mini Mill or a Hurco? Seems like I should be able to run between .10-.20" stepover and 100IPM maybe 200 IPM? What about surfacing with say a 1/4" ball end mill at 10KRPM, what sort of feedrates do these real machines want?


I know that the Haas control has an M code that will toggle the machine into the high speed mode where it gives up some accuracy to improve speed. I've talked with the HSMWorks guys about creating a post that toggles that M code when using adaptive clearing so you leverage that for roughing.

Also, have you seen this video of the Hurco control. Same code, generated by Solidcam - but the new Ultimotion is better suited to deal with the acceleration and data rates.

CNC Machine Technology Reduces Cycle Time | Hurco UltiMotion - YouTube

Hadn't seen that video, but it is a good example-- though its really a better comparison of the effect of accel\decel curves on part length.

I usually rough with a much looser tolerance, roughly .002" in both tolerance\smoothing tolerance, and about .0002 on finish surfacing. Though, I pay alot of attention to what tolerance is actually needed, as I am rather conscious of program size.

My 1/2" aluminum endmill, as saved in my library, is at 9000rpm & 100ipm, and barring anything incredibly stupid (like forgetting to clearance over a bolt and nailing it in rapid, etc) I change this tool every few months. I typically shoot for .15 WOC & .75 DOC, and would consider this "safe"

When I move something from the Mori's that I run to the older fadal's (newest one being 2003), I reduce WOC by 30%, and otherwise run it the same.

I'd guess either of the mills (though, i've never seen a mini mill in person, and have come to understand they're relatively light duty) you're looking at would fall somewhere in between there.

I generally use 5/8" diameter endmills wherever possible though, as they seem to make better use of my available horsepower, and as such are able to achieve a much higher MRR. The cut I described above is somewhere around 13cuin, my similar "safe" 5/8" cutter is at 25cuin.

Break's over. I'll try and expand more later if I find the time

[jid2]

Cool - thanks for the insight, it's good to have someone who actually knows this stuff!

I know the Mini Mill guys try and keep it to 1/2" tools based on their relative rigidness, and I "think" a typical cut is .125" at 100IPM. I believe the Hurco or the Sharp could handle the 5/8" tool like you said.


What about the surfacing stuff with things like 1/4" or even 1/8" ball?

[SCzEngrgGroup]

...and I "think" a typical cut is .125" at 100IPM.

For what they cost, I would hope they'd to better than that! I can do that on my knee mill or my Novakon Torus Pro which, combined, cost way less than a MiniMill.

Regards,
Ray L.

[rlockwood]

For what they cost, I would hope they'd to better than that! I can do that on my knee mill or my Novakon Torus Pro which, combined, cost way less than a MiniMill.

Regards,
Ray L.

Fairly certain he's talking about a relatively full depth axial, .125" radial-- which isn't bad at all.

Surfacing is so dependent on so many factors (toolpath style, desired finish, desired accuracy, spindle speed, tool type, engagement) its tough to give an estimate of what "should be doable"

I regularly surface at between 150ipm & 240ipm, and none of our machines are anything close ideal for surfacing. But by the same token, there's plenty of times that i'm running SIGNIFICANTLY less than that.

[jid2]

No practical experience for me to confirm, but for a Mini Mill, 1.5" deep .125" wide, 6K RPM and 100IPM were settings I got from a user as nice safe ones. On a Super you could likely do 150IPM. Here's video of a super doing 350IPM at .5"deep and .22" wide.

https://www.youtube.com/watch?v=zrql...e_gdata_player

If your Torus Pro can do 1.5" deep and .125" wide at 100IPM that is pretty killer. I'd program a basic clearing operayion and send you code to see video.

[SCzEngrgGroup]

No practical experience for me to confirm, but for a Mini Mill, 1.5" deep .125" wide, 6K RPM and 100IPM were settings I got from a user as nice safe ones. On a Super you could likely do 150IPM. Here's video of a super doing 350IPM at .5"deep and .22" wide.

https://www.youtube.com/watch?v=zrql...e_gdata_player

If your Torus Pro can do 1.5" deep and .125" wide at 100IPM that is pretty killer. I'd program a basic clearing operayion and send you code to see video.

No, I misunderstood what you wrote. You only gave one number, and I was thinking depth while slotting, but you were talking deep peripheral milling.

Regards,
Ray L.

[jid2]

Here's some really little stuff.

tapping 0-80


.065" slotting



.020" drilling



Finished part with .032" slots, next to a 9mm bullet for reference.

[FannBlade]

I would break that drill bit just chucking it up! Nice work.

[rwskinner]

I think all the pictures and videos are awesome so please keep up the great work.

Richard

[annerajb]

I think all the pictures and videos are awesome so please keep up the great work.

Richard

Agree I am lurker and I get instant emails when you update the thread. Don't post because I don't have anything to contribute but your progress and good quality videos and info are great!.

[gmh265]

jid2,

Love your machine, has been a great inspiration for my own build which I am busy researching and designing (also RF-45 clone). Your signature quote suits me perfect as well.

Couple of questions if you don't mind, are you using an XL belt profile on your 3 axis drives? and I see you have 2:1 belt ratio for x & y, is that also the case for the Z? Are you happy with your speeds?

Cheers!

[jid2]

Yes on the XL belt profile. If I did it again I would likely use a profile with better repeatibility. I've had no measurable problems with the XL's but there are profiles that support higher torque and precision - I didn't know that at the time. I'm 2:1 on X and Y and 3.3:1 on Z. For me Z axis has been tricky and even with the gas shocks I had to really knock down the accel and speed to keep it from faulting.

I think the speeds are great for what they are, and it's hard to go fast on a machine with dovetail ways. I ended up running 200 IPM on X and Y with around 50 accel. And 100 IPM on Z with some stupid low accel, I can't remember - but the thing doesn't fault anymore. Others have managed faster, Gary got 300 IPM out of each axis, as did Ray on his Knee mill (except the heavy knee). But I couldn't, and didn't want to trouble shoot it further.

Lots of people say that they don't need fast cutting or rapids speeds on a hobby machine.... Well, I'm not retired, and I have way too much crap to do than I have time for. So guess what, I want fast cutting and rapid speed, it doesn't "scare" me - bring it. The Novakon machines with linear ways and servos are doing 500 IPM rapids - that would be sweet. I'm to the point that I'll just buy a real VMC next, but that is because I have some stuff that I want to manufacture myself. I need more speed, a tool changer, good flood coolant, a real control system, and a 4th axis. And I don't want to build it all myself anymore.

[SCzEngrgGroup]

Take a look at GT2 belts and pulleys - they're more accurate than XL, for not much more money. And they're incredibly strong. I have 3/8" GT2 belts running all four axes on my knee mill, include the knee. X/Y/Z will all do 350 IPM, the knee will do 75 IPM, and it's been operating on the same set of belts for almost 6 years now, with no signs of wear or distress.

Regards,
Ray L.

[gmh265]

Thanks for the detailed reply jid2, and the gt2 recommendation Ray!

I actually managed to find both a gt2 5mm and an XL type belt here at work and the GT2 looks much more robust.

Half of the fun of this project for me is the actual build, but know where you are coming from with being over building it yourself. We have a handful of haas machine where I work so I would like my mill to have half decent speeds. I think I would be happy to get over 200 IPM on the x & y at least.

cheers!

[gd.marsh]

For the record, I used HDT5 belts & pulleys on my machine & have good repeatability ..

I initially was able to get over 300ipm as shown on the video, but have since tuned things down to 250-X, 225-Y, 175-Z
After a break-in period I re-adjusted the double nuts on the screws & snugged the gibs up to improve finish. At that point I started seeing the same issues Jid2 described on his Z axis. My machine seems to have some slight flexing in the width of the Z gibbs through the center of travel .. which means to keep from binding at the top & bottom they have to run loose through the middle. No doubt it's related to the big spacer I put between the head & Z-slide which increases stresses on the column considerably.

In my case I tweaked & tweaked to find a happy medium on the gibbs & backed away from the more aggressive settings for reliability.

[jid2]

Picked up an important element in the "desire to have a real VMC" department. This popped up on Craigslist yesterday morning and I had it by afternoon.

20hp Phase Perfect digital phase converter, and it came with the 3 phase panel with 3 circuits and enough wire and cable to setup a shop with 4 machines. Pretty sweet deal for $3500.

[rlockwood]

Good buy; let me know how that chev' looks.. I might take the other one