[kvom]

After a couple of weeks of emailing Khai at Novakon, I have decided to "pull the trigger" and order a NM-200. This will be for hobby use primarily, as my goal is to build a working radial engine (probably Edwards 5).

I am ordering the CD-100 upgrade with an integrated motherboard, which will hopefully avoid any Mach3 install and configuration issues. Because of space considerations I decided not to order the coolant system, so initially I will be using the machine without coolant.

I will need to decide on a CAD-CAM package. The mill comes with BobCad V21, so I will do some experimentation with that. I'm currently taking MasterCam training, so I am pretty new at CAD-CAM. However, I have taken courses in g-code programming, and most of the small engine parts I have made manually in the past I can program in g-code. So I'll get acquainted with the machine via hand-coded programs at first. I do want to be able to do some engraving and similar 2D work, so a package will be a necessity. CamBam looks interesting.

I'll try to update this thread with some photos when the machine arrives.

I currently have a manual mill with an R8 spindle and an air-operated drawbar. I suspect that changing tools manually will get old pretty soon, so I will be considering a quick-change system at some point.

[urbanimports02]

i would strongly consider the bobcad v23 upgrade. at $350 it is a killer deal. it is much much better then v21. i am not to saavy on the cad side of it yet, and so far do not like it much, but the cam side is very nice for the money. for cad i have both solidworks pro(very spendy) and alibre design standard. i love alibre for general use. it is very easy, and very quick. i use it on my lap top then export files to the cd-100 into bobcad and works like a charm. for more complex stuff i use my solidworks. i use it for everything from simple parts up to race car body work and chassis design, but it sounds like you will not need anything like that especially when it costs almost as much as the nm-200! if you look around and download the free trial version of alibre, they will start emailing you offers. they eventually offered me the full program for $99, and recently offered me to upgrade to there newest version for $99 as well.
also, request that novakon test your machine with your controler. i suspect they did not test my machine with my controler as it did not have a novakon xml intsalled and the y axis was running everything upside down which was fixed in mach3 telling me my cd-100 was not used to test my machine.

[fourwheeler]

i would strongly consider the bobcad v23 upgrade. at $350 it is a killer deal. it is much much better then v21.

Very good advise...V21 is not to my liking either but I am starting to grasp it better. I downloaded a trial version of V23 and it is much better to work with.

[CROSSHATCH]

I plan on upgrading an purchasing the nm-200 within the next year and half. Will be saving for it. I can't wait for the day

[kvom]

It's supposed to ship today. We'll see if it gets onto the truck.

[ihavenofish]

congrats. i use an nm200 sometimes and its a pretty nice machine.
i increased the speed to 125ipm on it and it runs very well.

[urbanimports02]

do you mind sharing your settings to get 125ipm? i tried that, and it was stalling and losing steps big time! do you use more conservative acceleration numbers or something? also, how is your circular interpolation? my machine produces circles that are out of round by more then .005" that is using 3 flute solid carbide cutters and i have even tried rerunning the program to clean it up, and still out of round. otherwise i like the machine. i ran it hard all day yesterday and produced 7 of my parts, running at 4000rpm pretty much the whole time, hope the spindle is up to the task!

[ihavenofish]

do you mind sharing your settings to get 125ipm? i tried that, and it was stalling and losing steps big time! do you use more conservative acceleration numbers or something? also, how is your circular interpolation? my machine produces circles that are out of round by more then .005" that is using 3 flute solid carbide cutters and i have even tried rerunning the program to clean it up, and still out of round. otherwise i like the machine. i ran it hard all day yesterday and produced 7 of my parts, running at 4000rpm pretty much the whole time, hope the spindle is up to the task!

the one i use is the older blue one with the gecko controller. i dont know how different it is form the new one, but heres some tips...

1: you need to increase the mach core clock for more speed. i think i set it to 45k, possibly 60k.
2: 125ipm is tops. above that the motor torque drops to 0. so 150ipm for example doesnt work... at all, motors dont even want to turn with no load.
3: acceleration on mine is at 40, which is HIGHER than stock, and lower than i wish it was. faster speed is useless without the acceleration to get there.
4: the stalling is due to pulse timing in the motor tuning panel. i had to adjust them and test a longish program over and over with straight lines to get stall free running.
5: the timing settings arent going to be the same on every pc / machine, so giving you mine wont help. i tried changing khai's CMTS demo machine at the show and it stalled with my settings. you will need to experiment to find the perfect match. i had to do the same exercise on my gecko g540 controlled nm-070. it seems to be the nature of paralell port and drive variability at higher pulse speeds.


6: if you are getting out of round circles, this may be due to another separate issue, especially if the "flat" sop is always in the same axis. try running mach in sherline half step mode. this fixed a similar issue on my gecko g540. the nm200 cuts dead perfect circles (within .001"). this is aparently a difficult to track "bug" in the gecko drives. if you manage to isolate it to a single drive, gecko will warranty it through novakon, but theres no absolute guarantee the new drive wont do the same. the only "perfect" fix ive found is the half step mode, which doesnt seem to have any reall effect on performance and accuracy, though theoretically the motors are moving less smoothly.

also make sure theres no backlash in the system of course. check the screw bearing supports, gibs, and such.

[kvom]

I just got an email from Khai that my mill was shipped today.

[fourwheeler]

I just got an email from Khai that my mill was shipped today.

Ah the anticipation....:cheers:

[urbanimports02]

congrats, when your machine arrives, try the settings below from ihavenofish. i owe you big. changing the kernal speed to 60khz and increasing rapids seems to make everything happy. i test ran a program with rapids at 120ipm(just going a little more conservative) and accel at 40, and walla! it is moving a lot faster, almost scary on short rapids, it sounds like the table is jumping, thought for sure it was going to lose steps. finished the test run and rezeroed with edge finder, and it was bang on. ran the program on eight parts with a work stop on the vise, so no rezeroing, and all parts cam out perfect and the circles are now within .0005" i guess the 60khz must have done it as i did not do the sherline half step mod? whatever it is, i am pleased as punch and appreciate the insight. oh, and i have the current novakon cd100 control with internal pc. i will also try to post some pics of my $40 flood enclosure, it is working great, just need to improve the "door". i made a simple upper square tube frame and attached .020" marine vinyl(for boat canvas windows) using office depot 1" capacity paper binder clips. the vinyl just hangs inside of the coolant containment tray and works like a charm. it is kind of the shower curtain idea, only heavier material. between this and the other mods i made to the flood cooling system, i have this thing working pretty good, and filled a 8 gallon trash can with aluminum chips in 2 days time without any clogs or anything. i enjoy this nm-200 more everyday, but already looking forward to a sharp 2412s.

[ihavenofish]

woot. glad that all worked well.

the sharp.. mm.. yeah. i actually saw the new haas tapping centre at the cmts show and have been wanting one since. at $50k though... its more than i can justify for such small travels.

maybe we can all hound khai to make a novakon 20x40 tool room mill, hahahah.

[urbanimports02]

hmm, i think i spoke too soon i ran one of my more complex programs, and at just one spot in the program, it will stall at the end of a rapid. the program cuts a final pass on a center round pocket, then rapids out and over to start an outside profile, at the end of that rapid, it stalls and misses steps. i stopped it just in time, and it looks like i can save the part, but i ran the program with no spindle, and you can hear it stall at the end of the rapid movement, almost like something put the brakes on. even turning it back down to 75ipm rapids makes the same noise, and seems to be off a couple thou.(at 120ipm, it ends up off .250 or more) funny thing is that the program makes the same series of rapid movements several times as it makes the next pass, and the rest of the rapids no matter what speed i set it at sound nice and smooth, and they are all the same rapid code???
sample of code
finishes center pocket at about X1.45 Y2.17
G0 Z.1
G0 X-.25 Y2.2(stalls at end of this rapid along x axis about .25" shy of destination)
G1 Z-.25
G2 X2.85 R1.65(something like that, too lazy to go out and read it)
G0 Z.1
G0 X-.25 Y2.2(smooth as silk???)

i ran this code on a couple parts last night at 75ipm rapids, and all looked well. upon closer inspection i see that it is off .007" or so left to right matching this stall i see in the x axis. other funny thing is that at first when i went to run this program tonight, it stalled at the end of the rapid and then the x axis would not work. i shut the machine down and fiddled with wiring, and nothing seemed to be off. when i would jog the x axis, the dro would show movement, but the table would not move in the x direction, only y and z. i tried restarting mach3 and still no x axis. i restarted the cd100 and when it came back on, the x axis was back to working. could this be one of those gecko drive problems you speak of? funny thing is it only does this at this one rapid, i ran several parts today at the faster rapids and all came out perfect. this one spot in this code is giving me problems, and it looks like it did on this part yesterday as well. maybe i better start a seperate thread, sorry.

[ihavenofish]

hmm, i think i spoke too soon i ran one of my more complex programs, and at just one spot in the program, it will stall at the end of a rapid. the program cuts a final pass on a center round pocket, then rapids out and over to start an outside profile, at the end of that rapid, it stalls and misses steps. i stopped it just in time, and it looks like i can save the part, but i ran the program with no spindle, and you can hear it stall at the end of the rapid movement, almost like something put the brakes on. even turning it back down to 75ipm rapids makes the same noise, and seems to be off a couple thou.(at 120ipm, it ends up off .250 or more) funny thing is that the program makes the same series of rapid movements several times as it makes the next pass, and the rest of the rapids no matter what speed i set it at sound nice and smooth, and they are all the same rapid code???
sample of code
finishes center pocket at about X1.45 Y2.17
G0 Z.1
G0 X-.25 Y2.2(stalls at end of this rapid along x axis about .25" shy of destination)
G1 Z-.25
G2 X2.85 R1.65(something like that, too lazy to go out and read it)
G0 Z.1
G0 X-.25 Y2.2(smooth as silk???)

i ran this code on a couple parts last night at 75ipm rapids, and all looked well. upon closer inspection i see that it is off .007" or so left to right matching this stall i see in the x axis. other funny thing is that at first when i went to run this program tonight, it stalled at the end of the rapid and then the x axis would not work. i shut the machine down and fiddled with wiring, and nothing seemed to be off. when i would jog the x axis, the dro would show movement, but the table would not move in the x direction, only y and z. i tried restarting mach3 and still no x axis. i restarted the cd100 and when it came back on, the x axis was back to working. could this be one of those gecko drive problems you speak of? funny thing is it only does this at this one rapid, i ran several parts today at the faster rapids and all came out perfect. this one spot in this code is giving me problems, and it looks like it did on this part yesterday as well. maybe i better start a seperate thread, sorry.

might be a good idea to start a thread on its own about motor tuning. what youre seeing seems to be the pulse timing thing. what is the timing set to in motor tuning? i *think* i set it to 2/4, but im not in the shop. i will email the owner and he can probably check on monday. as i said though, each machine is different.

whats good now is you have a program that will show a stall predictably. the stall should be random, as in not precisely the same spot every time, and likely not on every single run.

if its in the exact same spot every run, it may be something else, like a tight spot in the gib.

[kvom]

My mill arrived today; it was shipped last Wednesday, so one week end to end.

The truck delivered 2 wooden palletized crates:



The smaller contains the stand:



I set it up to lift off the pallet with my auto lift. The base was attached to the pallet with a single bolt.



With the stand inj the air, attaching the 6 feet was easy. I adjusted the corner feet to be equal with the two centers a little short. Once the mill is in position I can level it more precisely. In any case, measuring with an electronic level after it was on the floor showed 1 degree off crosswise and 3 degrees lengthwise.



The chip pan is banded to one side of the mill crate.

With the sides and top of the crate removed, here's the mill as packed:



And with the packaging removed. Notice the steel bar that's protruding from the sides of the column. That's one I placed through the holes provided and that I will use to lift the mill off the pallet.



Here's the mill as seen from the rear:



I'm going to hold off until Friday for the assembly, as a friend will be available to lend a hand. I suspect the mill will tilt when lifted off the pallet. There are two eyebolts on the front of the base casting that can be used to level it off. Once in the air, I'll maneuver the stand and chip pan underneath and lower it down.

[kvom]

I bought the CD-100 controller with a motherboard, 2G memory, and Windows 7 installed. Doing it this way Novakon installs Windows and Mach3, then configures and tests the mill with the controller that's shipped. Thus, I should have to do no software configuration to get running correctly. We shall see.

I connected all of the cables to the controller, added a mouse, monitor, and keyboard, plugged it in, and booted it up. So far, it all looks OK:



Mach3 starts up, but states it's in "emergency mode" and requests an external e-stop. I'm assuming that's because the cables that attach to the mill are not attached.

The CD-100 is a hefty unit, much stouter than my other PCs. I'm estimating it's at least 30 lbs. It appears to be designed to sit as shown in the photo, given the placement of the air vents and the rubber feet. With the installed motherboard, it requires two separate power cords. So it appears that the motherboard and the CNC controller have separate power supplies.

The two cables that connect the CD-100 to the mill are about 10' long, so there's a lot of latitude on placement relative to the mill itself.

[kvom]

Having bought a length of 12ga SSO for the power cord, I attempted to wire the plug provided by Novakon for attachment to the mill. The interior of this plug is so narrow that anything larger than 14ga can't work. Even with 14ga single strand wire it took considerable manual dexterity to put it all together.

I frankly think Novakon needs to rethink this design. Most 220V circuits in the US will have 30A breakers, and 14ga wire is not capable of handling 30A current. In any case, a pre-wired plug with a pigtail would be preferable.

[izzy88]

Congrats! Please keep us posted. Looks like a really nice machine!

I'm looking at their NM-145 myself but maybe I can sneak this one passed the wife

[David Da Costa]

I bought the CD-100 controller with a motherboard, 2G memory, and Windows 7 installed. Doing it this way Novakon installs Windows and Mach3, then configures and tests the mill with the controller that's shipped. Thus, I should have to do no software configuration to get running correctly. We shall see.

I connected all of the cables to the controller, added a mouse, monitor, and keyboard, plugged it in, and booted it up. So far, it all looks OK:



Mach3 starts up, but states it's in "emergency mode" and requests an external e-stop. I'm assuming that's because the cables that attach to the mill are not attached.

The CD-100 is a hefty unit, much stouter than my other PCs. I'm estimating it's at least 30 lbs. It appears to be designed to sit as shown in the photo, given the placement of the air vents and the rubber feet. With the installed motherboard, it requires two separate power cords. So it appears that the motherboard and the CNC controller have separate power supplies.

The two cables that connect the CD-100 to the mill are about 10' long, so there's a lot of latitude on placement relative to the mill itself.

You will get the emergency mode on Mach 3 until you turn on the power to the PC (the small power switch).

David

[kvom]

Sunday I was ready to mate the mill to the base cabinet. My buddy didn't show up to help, so I used the auto lift to raise it off the pallet. Using a rod through the column holes, the mill tilts very far forward. So I need to use all 4 arms: 2 for the rod and two for the eyebolts that are provided at the front of the base. So I was pretty easy to lift and remove the pallet.

I loaded up the piece of the packing crates in the back of my pickup, and was able to put them in a dumpster that was at a neighboring house undergoing renovation; saved a trip to the landfill.

Then I noticed that the holes in the cabinet are threaded, rather than through-drilled. I didn't get any mounting bolts in the shipment, but was told the 4 bolts used to attach the mill to the pallet would work. I found they wouldn't fit. After some investigation, I determined that I need M12x1.75 bolts rather than the M10x1.50 bolts that I had.

So I will need a stopover at HD on the way to school tonight to get them.

In the meantime, I removed the wooden block that was supporting the head and installed the drawbar. The wood is through-drilled lengthwisen and supported by a metal rod threaded on both ends. One end is screwed into a t-nut in the table, and the other bolted at the top of the spindle. Once the rod and wood are removed, it's necessary to attach the drawbar collar to the threaded top end of the spindle. This is a left-hand thread. Novakon provides a crude 36mm wrench to hold the spindle nose while tightening the collar. Then the drawbar slided down the spindle bore and rests on the collar.

I tried mounting a R8 collet and had no issues. Tightening and loosening the drawbar means using two wrenchs, one on the collar and the other on the drawbar.