At 4060 steps/inch, there is ZERO benefit to increasing the Mach3 kernel speed above 25 kHz, and it often causes problems, even on fast machines. 25kHz is fast enough to run well over 300 IPM.
Regards,
Ray L.
Printable View
I guess I'd stop the Porn watching while machining then.... Haha
I run 8000+ steps at 45k.
Maybe there are processes in the background running and using up CPU time? Ctrl+Alt+Del to open Task Manager then click the "processes" tab then the CPU column to sort it and see whats running much or even at all above 0%.
"clicking" reminds me of a busy computer when trying to run my TAIG with Mach3.
I tried using a pulse width of 10us and kernal at 25k but it did not help. I also reran the driver test and was able to go all the way up to 100k without any problems. I don't think there is anything running on the PC that would cause slowdowns. It was a fresh install and I did all of the PC setup tips from Mach.
Chris,
I know this is a long shot but I found this problem a few years after I had gutted an older computer. This computer developed a mind of its own and would shut down on its own, just die. I pulled the power supply to make a bench power supply and when I opened it, I noticed the capacitors had ruptured and were leaking. Quick Google search and I found out about the cap problems on some of the older power supplies. Only certain years were an issue but that computer started doing some crazy things.
Might want to google power supply capacitor leaking and see if your power supply for the computer might be dropping out somewhere. Maybe not the main power but one of the lesser volt supplies might be dropping somewhere causing the train to slow down.
Might want to see if your CNC power supply is drooping under load as well. Since there hasn't been any new post I figure people are out of suggestions. The motor soulds like it is stalling out like you said and I wonder if voltage is an issue.
I just got my first cnc project so I will be dipping a toe into the great world of cnc. Nice big router table needing new electronics.
Chris
FOUND IT!!!
I talked to Automation Technologies this week and they suggested I try setting the drivers to 1/8 stepping. That did seem to help either.
This mornign I looked at the PC and went through the XP Optimization document one more time. There was a section about ACPI. I tried setting the computer to "Standard" from "ACPI" and suddently the glitches were gone. :cheers:
Here's a video of the machine at 300IPM:
[ame=http://www.youtube.com/watch?v=Od9Pr1nlmnQ]G0704 steppers at 300 inches per minute - YouTube[/ame]
I'm waiting on CNC Fusion to send me a new Z screw and mount so that I can finish assembling the machine. I also received my treadmill motor this week, so I'll start looking at how to mount it up.
Chris
I'm still waiting to get my new Z screw and mount from CNC Fusion. It was supposed to ship on Wednesday, but they haven't sent me the tracking number yet.
I decided to put the column and head back on the machine. After some tinkering I got the head trammed. I'm dying to make some parts on this thing and I have the oiler mount drawn up, so I'm going to try cutting it 2D with manual Z moves this weekend.
I needed a chamfer mill so I stopped by the tool shop this morning. I've been wanting to try an aluminum specific end mill so while I was there I picked up a "Best Carbide" 3/8 3F 45 degree EM. We'll see how it works out.
It's supposed to rain all weekend here, so I'll have plenty of time to tinker in the shop. :D
Chris
Is it ZRN coated? If so, you bought a great endmill, 3Flute all the way!
No ZRN. I'm sure it will be much better than the cheap HSS EMs that I have been using though.
Actually, used properly, a good HSS 2-flute will provide all the performance your machine can support, and at a much lower cost that carbide. You're going to be limited by spindle RPM, power, and machine rigidity, all of which will prevent you from getting any real benefit from carbide, and force you down into RPMs and feedrate that HSS can easily handle.
A 3/8" 3-flute carbide running 0.1" DOC slotting wants to be spinning at 15K RPM, and feeding at 150 IPM in 6061,and your spindle will have to be putting out close to 2HP to do that. And you'll need a GOOD coolant and chip removal system. Can your machine handle that? I know my knee mill won't do it.
The same cut, using a HSS 2-flute, wants to run 6600 RPM, 44 IPM, and 0.5HP. Still almost certainly beyond the RPM and rigidity limits of your machine, but not by nearly as much.
High-quality, US-made (Putnam, Greenfield, Atrax, Hertel, and others) HSS 2-flutes (I typically buy 1/2 HSS endmills for about $12 each) are not only more than adequate for small machines, but will generally last longer, as carbide is much more easily damaged by re-cutting chips, spindle runout, thermal shock and many other factors.
Regards,
Ray L.
Thanks Ray. I will keep than in mind. Unfortunately, the local machine tool supplier mostly caters to the big CNC shops. The have a very limited selection of HSS. I'll try ordering a few name brand USA end mills when I get the chance.
BTW: Where do you buy your tooling?
what butt did you pull those numbers from? gwizard gives 7500 rpm, 67ipm and .53hp for your carbide 3fl at .100 slot. your hss specs are way beyond extreme too even using the vmc machine. Chris, get gwizard and get real speeds and feeds for your machine and don't listen to someone trying to scare you.
walt
I agree with Walt. Get Gwizard! I wanted to just solve everything on my own to include speeds and feeds. Its not that I didn't have the money just thought it was an easy mathematical equation that I would easily tackle. Well it ended up costing me the amount of a 1yr subscription to Gwizard in broken tooling or lost material.
I remember Ray posting a link to a free speeds and feeds calculator somewhere. I'm sure you could search for it. But I haven't compared it against Gwizard so I couldn't give you a review on it.
SFL
GWizard....I guess if you want to pay for it, over and over and over..
This is what I use. http://zero-divide.net/index.php
He gave you speeds-feeds for a carbide tool. And i assure you i have bushed them harder than that.
Here->FSWizardfor a 0.5" 3Fl HScobalt End mill recommends about 6000 RPM, 90IPM for full slot at 0.1" deep. Should pull 1.5 HP.
On a HAAS VMC i ran the same end mill at the same 6000@90 speed/feed 0.25 Deep !
Tool lasted for a week.
Carbide tool with the same settings lasted for 2 months.
for a 3/8 HSCO FEM it gives 7700RPM at 81 IPM when slotting 0.1 deep.
So yeah that 7500 RPM at 67ipm is in the range.
If FSWizard shows overload on power you should reduce depth of cut rather than feerdate. This way tool will do more work before it gets dull.
FSWizard may be free but experience and math put into are not worse that other paid-for software.
hey BamCNC thanks for shout-out!
the specs i gave were for carbide too according to gwizard, even yours with a 3/8 carbide maxed doesn't come close to rays typical over exaggerations. my point to chris is to get the info for himself. there are tons of free s&f calcs on the web. gwizard isn't free but it comes with far more than feeds and speed. thanks for the link to your freebie.
walt
zero_divide,
Please don't waste your time responding to WaltP. He is by far the worst troll on all of CNCZone, as he's never once made a useful, helpful, or constructive post - only criticisms, snide insults, and nasty personal attacks. His only goal is to stir up trouble for his own sick amusement. Based on his posts, I very much doubt he has any actual machining experience at all. He will, of course, claim otherwise, but has never posted anything that would in any way provide back up for his claims. If he did know anything, you'd think he would have been able to contribute SOMETHING of value in his 92 posts to date, but he never once has. Most people here have had to learn to just ignore him, in the hopes that he'll eventually take the hint, get bored and go away, as most trolls eventually do when others stop "taking their bait". But he seems to be a slow learner....
Regards,
Ray L.
You are welcome.
..
No, No. He meant that for slow flimsy machines there is no reason to buy carbide tooling. That's all.
You will max out RPM long before you can realize the carbide tool's potential.
IE: At my work we have a HAAS. I run 1/2" 3Fl 2" long Hi-Helix EM at 10000RPM and 170 IPM and 0.25" depth of cut when slotting. And i go 250IPM for slotting at 0.125 deep.
If all i had was 5000 RPM i would go with HSCobalt or HSS just because i will never be able to utilize all of carbide potential at 5000RPM
BTW there ere not many free online/standalone calcs like FSWizard. in fact i dont know any.
@HimyKabibble
Hello Himy.
I dont ignore anybody and i am always willing to pick up a fight.
But thanks for warning ;)
Yes i agree these benchtop machines aren't Haas's and one would be foolish to expect similar performance. tools have a range of recommended sfms so running at the low end of the range does put it within the realm of some benchtops. only giving a newbie the maxed out specs is of little value. again he needs to get the info himself with the aide of a good calc like yours.
some people don't like to be questioned and have to resort to rants and name calling, pity.
walt
I cut my first part with the mill today. I still haven't received my replacement Z screw from CNC Fusion, so I did all the Z moves manually. I cut the mounting plate for my oiler out of .250" 6061 plate. I tried cutting the perimeter with a 2F .375 HSS at 13ipm and .110 DOC and it was a bit much so I backed off to .060 DOC. That was much better, but then the mill loaded up about 15" into the cut. I didn't have any WD40 on hand, so I just used air at low pressure to keep the chips clear. I definately need a better EM. I was tempted to use the 3 flute carbide I bought, but I'll try it out later.
You don't say what RPM you were running, but the endmill "loading up" is usually an indication of a too-slow feedrate. At 0.060" DOC, coolant/lube should be completely unnecessary if you're running the right RPM/feedrate with a sharp tool. Assuming about 2500 RPM, you should be running about 23 IPM at that DOC.
Regards,
Ray L.
Oops. Yes the rpm was 2250. I was using .003 per tooth per rev. You're suggesting ~.0045" right? I'll give that a shot next time.
Sent from my Nexus 7 using Tapatalk 2
The way to get best performance is to calculate starting RPM/feed values, then try them out on a pretty shallow cut. If it works OK, then go deeper, and try again. Keep going deeper until you get into trouble (spindle bogging down, chips welding, etc.). When you get to that point, back off until it cuts well again. To go deeper than that, you need to reduce RPM and feed, and start over.
Regards,
Ray L.
Hey Ray,
I think most of the problem was caused by using the cheap worn EM. I've had this set for 15 years and they were pretty crappy to begin with. I ordered some Hertel 2F TiN coated HSS aluminum cutting end mills from Enco this morning.
Where are you coming up with your starting feed per tooth numbers? All of the calculators and data sheets I have seen recommend ~.003" per tooth.
Thanks,
Chris
I pulled that one from FSWizard. But it really doesn't matter - pick a number you like, and start there, then to the experiment I suggested. No calculator will give you the best numbers for your machine - every machine is different, and the calculators can only give you a reasonable starting point. You need to learn to "read" the cut, so you know whether you're RPM is too high or low, your feed too high or low, by the sound and the surface finish. That only comes through experience.
Regards,
Ray L.
I got the Hertel mills today and did a few test cuts. At 14IPM I could take cuts at up to .090". I tried 20IPM and at .060" it cut fine but it was noisier than I felt comfortable with, but at .030" it sounded fine.
Its amazing what a sharp tool can do.
I received the replacement Z ballscrew and mount from CNC Fusion on Friday. I installed the screw yesterday. Unfortunately I managed to get the top AC bearing in upside down. After a couple up/down transitions the bearing came apart and stretched the coupler about 1/4". :mad: I've ordered a replacement coupler and bearings so the machine should be back together on Tuesday. Luckily it was only a $45 mistake. :rolleyes:
Chris
While I was waiting for the replacement bearings and coupler for the Z axis I decided to install a set of AC bearings on the spindle. I carefully sanded the spindle and quill to allow them to be a light press fit. I got is all back together this morning. Then I found a small flat copper colored piece of metal. It's about 1/8" wide by 2" long and it has a dimple on one end. I don't remember taking it out. I also looked through the diagram in the manual and can't find it there. It bugs me because usually I can take things like small engines apart and put them back together from memory. Oh well, I'll figure it out eventually.
I got the replacement AC bearings and coupler for the Z axis. I got the bearings installed this morning. I should have it back together tonight. Then I can actually MAKE something. :banana: First on my list is a set of small manifolds for the oiling system. I also need to make a cover plate for the hole in the column where the Z handle used to be.
The copper colored strip came from the ruler/scale/whatever you call it for the quill readout. I noticed mine fell out when I slid the ruler/scale out to remove the quill/spindle for the bearing upgrade. I can't for the life of me figure out how it goes back in. I've tried repeatedly with no luck. I don't really need it anyway as I'm going to be using Hoss's belt drive conversion that removes the ability to use the quill. Just bugs me that I can't get it back in though. You would think it would a simpler matter. It probably is and I'm just over thinking it or something.
That strip is basically a gib strip. If you look at the scale, on one of the long sides, you'll see two small screws, which are used to adjust the gib. There is a small hole in one end of the gib, and one of the screws should stick into that hole, to keep the strip from coming out. Yours was obviously much too loose, or it would not have fallen out.
Regards,
Ray L.
Thanks Ray, that's been driving me crazy for about a week and a half. I imagine it will put Chris's mind at ease as well. Since it fell out, I assumed it should just slide right back in, but I had no luck doing so. Now I know why it's not going back, I must be hitting one of the screws. I assumed that it was there to take up slack, but there being screws involved never crossed my mind for some reason.
Ray, I'm sure you're right. I know I have seen one of those strips before. I think I tried to repair a HF dial caliper about 10 years ago that had a similar strip.
Like Mystic I plan to go belt drive, so the DRO will be coming out anyway.
I'm looking forward to actually cutting some metal with it this weekend. I'm a little nervous about getting my tool offsets correct though. I want to make a touch plate and get it hooked up. That reminds me on my way home I need to stop by the electronics store and get a piece of circuit board. :idea:
I tried my first fully automated operation today. I drilled a 2" deep hole through a 2"x1"x.75" block of aluminum. I tried a peck drill with a .340" deep plunge and .040" retract, 2250 RPM 13.5 IPM. The feed seemed too aggressive so I backed off to 7ipm and then 3.5 ipm. Then the chips were getting backed up, so I tried a full retract between each plunge. That helped too. On the third part the spindle stalled, so I hit feed hold, cleared the chips and then hit cycle start. The drill operation finished, but it must have missed some steps when it stalled. The drill bit was too low and gouged the block. :mad:
I guess it's all part of the learning curve, right? :o
You don't say what SIZE the drill was.... Kinda hard to comment without having essential information. Generally, when drilling, you can go no more than 2-3 diameters before you'll need to do a full retract to clear chips. With larger drills on a small machine, you probably won't even get that far. As you go deeper beyond that, you need to start reducing both RPM and feed, AND retracting more often. Larger drills can take a LOT of power, which will limit you further still. Lubrication, rather than coolant, can help.
Regards,
Ray L.
Uggg! I keep leaving out vital information don't I? It was a .343" drill.
Afterward I tried using the G83 cycle on some .159" holes. I used a plunge of .125" for a total depth of .625" and it worked much better. I think I will try to use G83 anytime I have a hole that's more than 2-3 diameters deep.
I'm not sure why but my last operation was giving me fits. I was going to drill two holes .750" deep. For an air cut run I set Z=0 to 1" above the part, but when I ran the code it kept plunging into the part anyway. I'll have to dig into it when I have some free time again.
I think I have the Z issue figured out. When I set up the motor tuning I used the jog keys or Xbox controller to jog back and forth to test the settings. I found a post on the Yahoo forum where Art states that G0 rapids and Jogging do not work the same way. So I need to retest my tuning settings with some G0 moves. I think the X and Y are fine but the Z is missing steps once in a while. I probably just need to lower the Accel setting because it will jog at over 200IPM.
I also looked into the peck retract issue I was seeing. I'm using SheetCam at the moment and I was setting the retract distance to .050". The g code would only retract that amount from the bottom of the hole. I couldn't find an option to tell it to retract that amount above the Z plane. I found out today that you have to tell it to retract the full depth of the hole plus the clearance. The program will only retract to the "Safe Z height" so it won't waste time retracting too far above the part.
Chris
First, Congrats on getting to see the benefit of your hard work and studying. The lessons are common and if using charts that might pertain to larger machines, the noises made you slow down.
Try a few tests first. First rerun the drilling operation above the work piece a specified distance, zero 3" above the work for example, rerun it, does it return to 3" above the work after the file is complete and sent to 0?
You might be fine at this point, however I noticed your first peck was pretty aggressive IMO (this 1 move is where I suspect your steps were lost) and I would lessen it considerably. Being a hobbyist I am much more thrilled by success and the time spent during a pecking cycle is still short by any other comparison.
Thanks Fastest1,
I'm getting there. I've been using some of the online feed and speed charts and the free speed/feed calculator(the name slips my mind at the moment).
I slowed the motors to 120 ipm and 30 accel. I had them at 240 and 40. The funny thing is even "slowed down" it makes me pucker watching the tool racing towards the part.
I went back and editted the drill operations. I used 2250 rpm, a feed of 1ipm , full retract, and a peck depth of .125". It worked much better.
I also spiral milled pockets for the screw heads. I used a 1/4" 2 flute, 2250 rpm, feed of 3ipm, 3 degree slope. It was slow but it worked and it didn't wake up the kids either. :^)
Thanks,
Chris
Sent from my Nexus 7 using Tapatalk 2
Those are pretty extreme accel rates. I think I am running velocities of 80-120 depending on axis and usually accels of 1/10 of that. I have run higher but was suffering a loss of steps during a multiple hole pocketing routine when using CamBam. I can get it to repeat the error. I can even identify the line where it goes askew but there is no reason. The speeds aren't too fast and it will skip during an air cut too.
I broke and will continue to break end mills and parts in my journey. It is the way I learn unfortunately. Better than not learning at all, a little more expensive though. It is hard to listen to a machine as it cuts properly. Many times to me or other noobs or novices the sounds are very counterintuitive. The more is better approach is not always the answer here. I used to like high rpms in the spindle and if feed rate was proportional it is a good thing most likely. I have found trying to slow the spindle speed down works better because my tendency is to go too high and burn up the tool. Me and the local metal recycler are good friends now.