[Whizbang]

I just purchased a used Novakon NM-200, I got it moved, set-up and running over the weekend. I now have lots of questions....

1.) There is a port on the back of the head (see photo) What is it for? 4th axis ready?
Attachment 197130

2.) The electrical port on the top for the Z motor, its one of the plastic screw down ports. Where can I find a replacement? Or at least the replacement screws? (picture coming soon)
-Long story short....I didn't quite have a enough clearance with a lip on my garage door and broke the plastic mount to the motor for that port. I tried to check the thread on the screws because I can repair the port base but I mangled a couple of the screws. However the tiny long screws don't match the pitch of anything I have on my thread checker in eng or metric.

3.) The max travel proximity switch do not work (see photo), is this corrected by setting the soft limits through Mach?
Attachment 197132

4.) I have the original Novakon with the yellow box controller. Can I hook up and use my ethernet smooth stepper to t he computer? Are there advantages to doing this other than having the ability to hook up multiple breakout boards?

5.) So far I have only used a flycutter and was very happy with the results, it was very accurate. Are there any recommended parts/projects to complete that give a good gauge of all around accuracy of the machine?

6.) Does anyone have any booth enclosure plans? Mine didn't come with the side tins and after a nice hot shower of aluminum chips I clearly need to build an enclosure.

7.) Finally I have a light that is labeled 50W 220V but no matter what I do it will not power on. I thought maybe the blub was burnt out but honestly I cannot figure out how to get to the bulb without likely breaking a part of the light some how. Glass seems to be sealed so I can't get to the bulb. Any thoughts?

All photos: https://plus.google.com/photos/10520...KP5uoXrsvjqkgE

Thanks,
Whizbang

[kvom]

1) I'm pretty sure the extra port on the rear is for 4th axis. My mill has nothing attached there.

3) My limit switches don't work either, so I can't auto-reference the axes at startup. What I have to do is use the ref buttons on the mach3 diagnostic screen to zero the machine coords. I tried using the soft limits but since I make small parts I gave up doing that. The main risk is that the table can over travel in the Y direction and hit the front door frame if you have the Novakon coolant setup.

4) I also have the external box and just use the supplied cables. The processor is dedicated and it's unlikely you'd miss steps. I don't know if the smooth-stepper will work or not.

5) Best way to test accuracy/backlash is with a digital indicator. Definitely check the head tram.

6) To avoid chips in a cheap and easy manner I use a piece of clear plastic (lexan) with a magnet glued to it. I attach it to the vise. I do have the side chip trays from Novakon.

[UniqueMachining]

Whizbang

1. Yes, 4th axis connection. Verify internal hardware is present.
2-4 Email Novakon support for assistance.
5. Calibrate entire machine as per factory specs. Tram, perpendicularity, repeatability, backlash etc.
6. See various threads for sample enclosures.
7. Email Novakon, I remember it was tricky, just don't remember the trick.

Time spent calibrating the machine to as precise as possible will pay many future dividends.

Enjoy your new toy.

John

[Whizbang]

1) That is what i thought I saw 4 Geckos in my controller so it must be 4th axis. I need to find out the name of that connector so i can retrofit a 4th axis.
2) Emailed a few days ago, but haven't heard anything. I electrical tapped it all up so that moisture and dust can not enter the motor housing.
3) still working on this issue although hints would be nice. I think I will just use soft limits. however the real limit working would be nice for the z axis (worried im going to
travel right off the rails.
4) This was mentioned by Ray L to possible add a smooth stepper. besides the extra parallel ports does it really make that much of a difference smoothing the
movement.
5) Put pro-tram on it and within.00025 in the x and about .002 in the y. I plan on shimming the y however the machine is going to be moved again in a few weeks to a
permanent location where i will level the machine and really dial things in.
6) Definitely looking for plans and have one contact who may be getting rid of their pans. just looking for something cheap.
7) Still trying to figure this out. Thought I was going to break the glass if I pushed and pulled any harder

Does anyone have a genealogy of what has changed from nm200 to nm200 rev 2 to torus to torus pro. Is it all electronics and the spindle motor? The castings look identical. Was there any production changes due to quality or is the torus pro just a beefed up nm200 due to the electronics?

I am trying to justify buying another novakon if this machine performs well but i want the machines to be interchangeable for if i buy spare parts.( thinking long term)
I am looking for some nice small projects to cut to show the worth of the machine off. My parts i normally cut are bland. This is also why I am asking about the 4th axis cause if i could retrofit a 8" or 10" it would allow for some pretty unique projects.
Thanks guys.

[Novakon]

Hello Whizbang,

I have not seen your e-mail in our support mail. Please try again using [email protected] and I will keep an eye out for it.

We have many of the connector and parts for your 4th axis. We can work on any of this as you need.

The basic machine is very similar to the Torus PRO as is the ball screws and miscellaneous spare parts are much the same. The biggest difference in the Torus PRO is the electronics cabinet where we have included many commercial grade components and capabilities to be easily upgraded with the new PDB and pending ATC. Of course the servos are the main attraction for the serious user with production in mind.

Looking forward to your message.

Regards,
Novakon Team

BTW... are you the customer who bought the used NM-200 mill from a local customer near Toronto? If so, they have upgraded to the Torus PRO servo last week.

[Whizbang]

So I have the machine up and running. I have check the accuracy and backlash with gauges and so far am able to hold up to 0.001" which is awesome. I have not machined any material as of yet to check and see if the machine can hold that tight of tolerances under a load. I haven't machined any material yet because I have a few issues that really need to be address first.

First my E-Stop wouldn't work, My wife suggested the connector could have come loose or broke in transit. Sure enough one of the snaps on the plug was broken off and the E-Stop was unplugged. That was a nice quick fix. I keep a few e-stops on hand for the cnc router and printers

My next problem arose when I actually tried to machine some wood as a quick trial to make sure I was generating G-Code correctly using HSMXpress and SolidWorks. In doing dry runs each axis homed, then I moved the axis to the point I wanted to start and manually zeroed all 3 axis (X, Y & Z). When I would start the program the head would re-zero them come back and start cutting. The X & Y always move as they should but the first time I tried, my head moved up instead of down to cut. Now I noticed that after homing the Z axis that the down direction is negative. In the past and with most machines I have used down is positive for the Z axis, would most agree with this? Anyways I switched the positive negative orientation for the Z axis in MACH and tried to run again just cutting the air, that time the head moved down. When I finally decided to cut the wood I repeated the starting process and then hit start. The head re-zeroed itself then plunged all the way down into the material. THANK GOD is was wood and no damage was done but I cannot figure out what is going on. I don't know if it is in my Gcode or if it is a MACH setting or what. I have attached my Gcode for anyone who is better at looking for errors than myself. The part I was trying to cut is the image below. I set it up (at least I think I did) to to cut out the hex shape and hole on the inside first (starting from the left in the picture) then cut out the actual exterior contour.

My Routers that i have built both move positive down as in progressing + into the material. I typically use cut2d but think that many members results with hsm look fantastic. This is why I switched to HSM and to be able to do more complex 3d parts. With Cut2D I would zero the xyz and then cycle start and it would start cutting from there, however hsm is doing something differently. Also I can only export(post process) as a mach2.tap. I have no options for Mach3 Post process options. I understand this is outside of this specific novakon forum however I though I would mention it. I will probably also post it a HSM related forum.

Thanks guys

Attachment 197582

[MRM RCModels]

Are you using the post processor made by hsm for mach3? If not, give them a call.

Check your mach3 tool offsets. See what the offset is for tool one.

The code is a bit weird. It starts with a g91 incremental touch off to z0 and every z movement is in the positive direction (away from the work piece).

Check the tool offsets first. I had the head crash in the table once due to an incorrect tool height. It has to be said, did you set z0 using some kind of gauge block?

Sent from my Galaxy Nexus using Tapatalk 4

[Novakon]

Here is a little history on the your mill we had seen in the past. We presume your machine is the one you bought local to us.

We had done a field visit to the company relating to the reversing of the Z axis. This problem apparently came about after several years of use and, out of the blue, it would randomly start reversing the Z axis. Now, in looking deeper into this, it was totally isolated to the CAM and Mach3 when used on the PC. When the post code was loaded into the program screen, all seemed well. When viewing the geometry in the Mach3 graphics screen, the Z axis was reversed on the screen! The mill followed the course of the graphics precisely in reverse. I verified the control signals were correct from the BOB for direction of the hardware. There seemed to be no problems with the BOB to the motors...it followed the commands from Mach3.

We had reloaded the latest Mach3 program and still had the same problem. I don't remember the CAM package used, but after some of their work on re-posting the code, it all seemed to be working well. We had not heard about this problem again from the customer ever since. It seemed to be totally isolated to the post code generated by the CAM package, using the Mach3 post processor in the CAM package. We would look at the tool path representation in the graphics window and see if it looked reasonably correct for up and down direction. For what ever reason, it was reversed, and Mach3 did represent the graphics also in reverse.

I thought this might help you in isolating this problem you are having. This was the only time we have seen something as strange as this.

I hope it helps.

Regards,
Novakon Team

[Whizbang]

Yes, I am the one who bought the used machine local to you.

Just to clarify, their problem was the GCode that was output from their CAM software? Once they modified it they no longer had the issue?

Everything is moving in the right direction now, however before the program get to the actually cutting the head plunges down really far (beyond the set zero), then shoots back up (above the set zero I believe, however to avoid damage I am cutting air so its hard to tell how far above zero it is going). Once the head is back up it starts to cut.

There will be more to come as I try to problem shoot this. I have attached the most recent GCode in case anyone thinks they may have a bright idea. I did find an option in the "post processing" to NOT use g28 and am about to try running that right now. The attached code is what I already tried before finding that option.

Thanks
Whizbang

[Novakon]

Their problem was identified about 1 1/2 years ago and I believe it was corrected using a different post processor. Usually, once we hear nothing back, we presume things are working well until we get notified again. We have not heard anything more on this problem so we presume it was working.

The one thing you can do is to trouble shoot is to step through the program a line at a time in mach3 while watching the results of the mill. Somewhere, you will find the gremlin code causing this. I have used this method before and found several posting errors. It takes a little concentration, but it will reveal itself.

Regards,
Novakon Team

[TiagoSantos]

Just a couple of thoughts, although I have no experience with a Novakon and have only tested HSMXpress very briefly.

Most machines that I have seen or used (milling machines, at least) use the standard axis orientation. Z+ is up! Yes, this makes it so that cutting is always into the negative quadrant, nothing wrong with that. But that's just the way the coordinate system works. Positive is up, negative is down..

On a normal program out of HSMXpress or most other CAM packages, the machine will start with a G28 to Z0. Then for each operation, it'll move to a set safe rapid height, rapid down to the clearance height, and then move at cutting speeds from there. The code on both examples you posted looks like normal Z+ up code.. I recommend you set up your machine the same way!

Looks like on the first example, the machine zeros itself fine, moves to the starting position, applies the tool offset, then does this:

Z0.74
G1 Z0.6524 F20.

It's still in G0 mode, so it rapids to Z0.74 - normally, that would be a safe clearance height, in your case, since you flipped the Z+ to be down, it plunges straight into the wood at rapids speed. Then it feeds to Z0.6524 - again, normally that would be a down feed, to start cutting the material, but in your case, it's feeding up!

Anyway - My recommendation is to reset Mach3's axis orientation back to normal and get used to that.. You could probably work things out in HSMXpress to work with the negative up idea, but I bet it'll take a while to get through all the quirks and problems. It really is built to work with Z+ facing up, so I'm sure you'll keep bumping into problems for a while..

[LeeWay]

That also makes inputting your offsets much easier. That is unless you have a tool that is shorter than your Zero Tool. All offsets would be in the positive and never require a - except for tools shorter than your Zero Tool. Both my mills, router table and plasma cutter are setup this way.
In fact my lathe is set that way too.

[Whizbang]

Alright guys here's a bit of an update.
Regarding the cam software I believe I have it fixed, there was an option in hsmexpress not to perform a g28 z0. So what I did was bring the tool down to the material then touch off and zero all axis, I then cycle start the program and it rapids up to its clearance height before tracking to the first pocket that cuts. This method seemed to work well. I started off by cutting air, then I cut MDF, then aluminum. The air cut great who'da thought. The mdf cut great too. then the aluminum, it cut great up until the finish pass at full depth of the pocket. The tool snapped off immediately at the first corner of the hexagon pocket. I am thinking of greatly reducing the finish pass I think it was .012 which seems aggressive at 18ipm and 4000rpm for a .125 endmill. I am going to 1/4 that finish pass. Currently I have the finish pass set to 1 pass but if I select 2 passes does it 1/2 the finish pass or just perform 2 passes at the finish pass offset?

I hope that I can perform multiple tool changes using my current operations for homing i.e. home the first tool to zero cycle. Then have the machine pause at clearance height allow me to manually jog the z upward perform a tool change then lower back down and re zero the z axis followed by a cycle start and the continuation of the program. I am hoping to continue this procedure for the rest of operation so hopefully this method I am describing is possible. Eventually I plan on getting a z touch plate for these tool changes and even farther down the line I plan on purchasing the tormach tts or something similar if available.

One thing I noticed when taking tool out was how hot the spindle and drawbar were. Is there a lubrication point some where? It feels hotter then I would expect it to be after a 20 min run. I believe I read about hoss using "kubler grease" or some thing like that. Any way more to come on that and feel free to comment.

The next thing I noticed is that I don't believe that my oil is really getting to all points on the machine. Unfortunately I cant say for certain but I do have a small leak right next to the one shot. I have considered using just a touch of dye when i get low on oil again just to see when i can get it creeping out. I may overhaul this system down the road. My wife fortunately is a hydraulic and fluid conveyance engineer so can get me just about anything and any kind of fitting needed so maybe I will upgrade the system in a month or so. Has anyone else done this or noticed some points being more lubricated then others. pictures that I took during tonights trials are on their way. I am having some uploading issues at the moment. There are pictures of my 3D printed parts I used for prototyping as well as my first test cut in mdf and then my aluminum comparison.

Thanks for reading and commenting

attached are pictures of the 3d printed part, the mdf and the failed aluminum.

Attachment 197924
Attachment 197926

[CROSSHATCH]

Whiz- That's my typical method for touching off. I like to set my zero with a feeler and let it rapid to the height *(safe Z) before it comes down. That finish cut pass to me on .125 end-mill is way to high. Despite the fact that your using HSM toolpaths, I can only imagine what your tool deflection is. I'm a new advocate of something I wanted to try for a very long time. When my trial is done, I am purchasing it... download GWizard from CNCcookbook. I have been using it to estimate my feeds and speeds and you really cannot go wrong with it, I'm usually a little more conservative so i'll bump down the recommended. I'm sure if you input that data in Gwizard; it would have spit out a high deflection which is instant killer on a small end-mill.


I use my oiler too, but get yourself a nice hand-pump action oiler. I have one with a braided adjustable end and use it to get into any areas the one-shot does not. I used it on my old mill and I use it now on the Torus Pro; it's a simple upgrade.

BTW- What printer did you use for your part?

Hope this helps somewhat.

-Jason

[kvom]

When machining into a corner the tool engagement angle increases greatly depending on the radial engagement. You can reduce the roughing clearance, but your feed will always be restricted by the limit imposed by the corners. One option might be to pre-drill the corners. Doing so could also allow you to use a larger EM for the pockets.

I also use G-Wizard. My tool breakage went to 0 after I started using it.

[LeeWay]

I too think your finish pass was too aggressive.
I like to go 50% of the width of the end mill when pocketing. For regular shaped pockets, I can go like 75%, but odd shaped ones will tend to leave a hair of an island in some places.

For finish passes, I like to crank up the speed and take no more than 1/5th the end mill size. That would mean .025" in your case for both width and DOC. I slow it down too. This greatly reduces tool deflection, which is what you are looking for in a final pretty pass.

The other day I was running some jobs that I have ran thousands of times. I do use Gwizard as well and have those jobs optmized. I kept getting really terrible cuts and tool breakage. 1/8" carbides and even a couple 3/16" HSS EM's.
I found out that the material I ordered, 6061 aluminum is not what I received. They sent 6063 instead. It was like trying to machine chewed bubble gum.
So material is a big factor in how well something machines. I didn't even see a type for this kind of aluminum in Gwizard to input. Rightfully so. You don't want to machine that stuff.

[rlockwood]

When machining into a corner the tool engagement angle increases greatly depending on the radial engagement. You can reduce the roughing clearance, but your feed will always be restricted by the limit imposed by the corners. One option might be to pre-drill the corners. Doing so could also allow you to use a larger EM for the pockets.

I also use G-Wizard. My tool breakage went to 0 after I started using it.

Especially a "corner" like the ones shown, which ends up effectively being a slot.

[rlockwood]

Whiz- That's my typical method for touching off. I like to set my zero with a feeler and let it rapid to the height *(safe Z) before it comes down. That finish cut pass to me on .125 end-mill is way to high. Despite the fact that your using HSM toolpaths, I can only imagine what your tool deflection is. I'm a new advocate of something I wanted to try for a very long time. When my trial is done, I am purchasing it... download GWizard from CNCcookbook. I have been using it to estimate my feeds and speeds and you really cannot go wrong with it, I'm usually a little more conservative so i'll bump down the recommended. I'm sure if you input that data in Gwizard; it would have spit out a high deflection which is instant killer on a small end-mill.


I use my oiler too, but get yourself a nice hand-pump action oiler. I have one with a braided adjustable end and use it to get into any areas the one-shot does not. I used it on my old mill and I use it now on the Torus Pro; it's a simple upgrade.

BTW- What printer did you use for your part?

Hope this helps somewhat.

-Jason

Check out FSWizard at zero-divide.net

The web version is free, the downloaded version is about to become not-free, and has a few additional features.

[Whizbang]

UPDATE!

I have successfully made my first part. Yes it is only a vise wrench and drawbar socket holder, but none the less it is something I have always wanted to make. Better yet it is with in a .001 yes I baby'd it with a slower feed rate but none the less it is awesome. Only one problem I couldnt do the exterior finish pass because the part fell out of the bottom, I just had it in my vise. Next up get surface finishes like Ray L.. He seems to be really pushing limits of the torus/nm200. I am looking into G-wizard but price is still a bit steep for a lifetime subscription. though I suppose the 3 year would be good enough cause of the power that the nm200 has my spindle hp is with in the 3 year hp range. Do you guys purchase the life time or 3 year? G-wizard editor looks great too but it has a steeper price. I'm just a hobby user for right now. I would like some sort of quick gcode generator like what g-wizard or mach3 wizards offer just cheap and fully and furturely supported and developed.( p.s. I made futurely up)

Last main concern is wow the head and spindle sure get hot. Any grease points I'm not seeing. I definitely don't want anything wearing or burning up in the spindle assembly.

Attached are some pictures off the part before it was thrown in the parts tumbler.
Attachment 198278Attachment 198280

[SCzEngrgGroup]

Take a look at HSMAdvisor and FSWizard (both at CNC Tricks Home), instead of GWizard. Pricing is far more reasonable. And the BEST part is, they are NOT built using Adobe tools that force you to do an update about every other time you start it up. That was my #1 reason for steering clear of GWizard.

Your spindle should not get actually hot. It will be warm after running at higher speeds, but mine is only warm even after running all day at high speed.

What tools, WOC, DOC, feeds, and speeds did you use to cut the outside of the aluminum part? The finish should be MUCH better than that, even on the roughing pass. Try roughing at 3200 RPM, 0.250" DOC, 20 IPM, and finishing at 4500 RPM, 35 IPM, 0.010" WOC, full depth.

Regards,
Ray L.