[JerryBurks]

Thanks for the update Jerry, I have a 4th axis on my wish list and your find looks like a great product. I want to make a chess set using one of these would make the process much faster.

That would be an interesting project (do you have models for the chess pieces?). If you clamp the short and stubby pieces with spigot jaws you may not even need a tailstock and save finishing off the bridges/supports. But for some reasonable detail I suppose a 1/16" ball nose wold be required and accurate zeroing will be an even bigger challenge.

[JerryBurks]

I want to make a chess set ......

Hey Drools, I found something.... look at this . Unfortunately not cheap at $279 but that would be a worthwhile CNC project. The usual chess pieces are rather designed for the lathe and there is no much to gain with a CNC router.

The license prohibits of course redistribution of the original file, but it does explicitly allow posting, selling or giving away the model as "Incorporated Content", for example as image rendering, movie, computer game or physical part. I would guess that includes a rendering as CNC tool path. Maybe that would be a way to share cost if other people are interested.

[Drools]

Thanks Jerry, that is a nice chess set. I expected to pay something for the models, I do have some bookmarks but not on this machine and they would have been from a while ago. Basically you just need 6 models (King, Queen, Rook, Knight, Bishop, and Pawn) and the board.It works out to $46.50 per model not including the board.

[JerryBurks]

Just for fun I purchased the King of this chess set to play with. I can not use the native 3D-max format but it also exports .obj file. That I can use with my CAD system to extract the figure from its environment (chess board and walls), scale to proper chess piece size and export as .stl for the CAM.

Then I imported this .stl into DeskProto and started experimenting with various machining strategies. See picture below for a "waterline" tool path strategy with 7/1000" steps in continuous X-Z-A rotary mode. Not sure yet what would work best and is most time-efficient but all I need to do now is find a piece of 1.5x1.5" lumber and try it out.

Well, that means once I am back from a 1-week business trip to Hong Kong. :tired:

[Drools]

Please make a video
I forgot to ask, do you play chess?

[JerryBurks]

Please make a video
I forgot to ask, do you play chess?

O.K. I will do that after I get back home.

I consider the chess set more an interesting project and would not have thought about it if you had not mentioned. I still know the rules and may remember some basic strategies but I have not played chess for at least 20 years. But if I can convince myself to finish such a set and if I find a partner who knows.

JB


Sent from my iPad using Tapatalk HD

[JerryBurks]

The continuous 3d rotary machining is more complicated than I thought...
Not so much the machine part since the gearbox and electronics work quite well.

But the CNC controller has a bit of trouble getting the proper surface speed for the commanded feed rate dependent on the distance from the rotation axis (I use the Planet-CNC USB controller). Not sure if that is an issue with Mach3.

Anyway I have been working with Andrej, the guy behind Planet-CNC to help him improve this. I can already achieve a decent roughing cut on the little chess figure. The finish cut is still in the works. Here some video (try full screen):

[ame=http://www.youtube.com/watch?v=-gfLKKssa50]ChessRotaryMilling.avi - YouTube[/ame]

I wish I could turn faster for smaller diameter but with the supplied stepper the reliable speed limit of the rotary axis is about 8000 degrees per minute.

[Drools]

Looks good Jerry, what kind of wood are you working with?

[JerryBurks]

Looks good Jerry, what kind of wood are you working with?

Curly Maple. The local Rockler store had a special of 1.5x1.5" turning blanks ($2 each for a 2 foot piece) and I bought a few. Kind of waste to chop up such nice wood for my experimentation, though.

[JerryBurks]

I am giving in.....Since I finished the machine I was unhappy with the high speed performance of my huge 1600in-oz steppers (you can see some of the original concerns at the very beginning of this thread). This looked like a pretty good deal on a 4-axis kit from Longs Motors before I joined this forum and learned about the limitations of big steppers. While the torque is vastly more than needed at slow speeds these high inductance motors run completely out of steam over 150-200 ipm (even with 15mm pitch lead screws). Reliable machining is only up to 100 ipm.

So, I decided the anniversary of my build start would be a good time to upgrade:
- I want to keep NEMA34 because of better mechanical fit with the machine. But I will go for 465 in-oz steppers from Keling (only 1.65 mH/6amps vs. 21mH/3.5amps before)
- Upgrade the power supply from 48V to 72V and make better use of my Gecko 203V
- Swap the tapered roller bearing pairs on the lead screws to angular contact bearings (much less friction when pre-loaded).

Unfortunately it also means to replace the screw couplers (shorter motor with 1/2" shafts instead of 14mm dia).

Let's see how that works. If successful I will post the comparison. If not, oh well.

[CarveOne]

Get a Keling 72vdc linear PSU. This one change will be more improvement than you would think. Try that before you commit to buying new motors. It could well be all you need to do to get more power and cutting speed.

[ger21]

72V is too high for those motors, and they'll get really hot. Use 32x sqrt of 1.65, or about 41V. 48V might be OK, but they still might run very hot.

[CarveOne]

Yep, Gerry's right. I didn't check the specs for those motors. My Keling 495 oz N23 motors don't have the heating problem.

[JohnZ]

72V is too high for those motors, and they'll get really hot. Use 32x sqrt of 1.65, or about 41V. 48V might be OK, but they still might run very hot.

Is that formula to be used by multiplying mH or V? What he had listed was 1.65 mH, and I thought the formula above was to calculate max voltage based off of the designed voltage. But I may be completly wrong here, so please correct me if so.

I can state that with those big 1000+ oz-in steppers that the jump in ps voltage makes a HUGE difference in rapids and torque.

Thanks

Sent from my DROID4 using Tapatalk 2

[Drools]

Is this your kit Jerry? Stepper Motor/Stepper Gearmotor/Stepper Motor Driver wholesaler and supplier
If it is the mH of the stepper is 22. Plugging that into Gerry's formula 32 * sqrt 22 = 150VDC. The DM860A will handle 80VDC maybe just a new PS might give you the performance you are looking for.
Just a thought.

[JerryBurks]

Is this your kit Jerry? Stepper Motor/Stepper Gearmotor/Stepper Motor Driver wholesaler and supplier.........

Pretty close, except that I have dual shaft motors. Also I already swapped the Chinese drives for Gecko 203V very early and it improved the situation significantly. But no matter which one they both take 80 Volts.

I have the Keling 72V/20 amp unregulated power supply already and will try that first. The new motors are my fallback if that is still unsatisfactory.

I am aware of the power supply formula but as far as I understood this is more a rule of thumb and mostly aimed at NEMA23 motors. NEMA34 steppers have better heat dissipation because of the larger surface and my steppers are anyway mounted on sizable 1/4" aluminum plates/heat sinks. The big steppers I use now get barely hand warm.

I believe the voltage/heat limitation comes into play when the motors are mostly run at high speed. I may try one of the low-inductance motors on the Y-axis in parallel and serial mode and see if that works better than the old motors without overheating.

[ger21]

I am aware of the power supply formula but as far as I understood this is more a rule of thumb and mostly aimed at NEMA23 motors.

I thought is was for all motors, both Nema 23 and 34. I looked at Geckos site, and it just says never exceed 25x the rated voltage, or 32(sqrt(mh)).

The rated voltage is 1.62V, so 25x that is 40.5, which happens to be almost the same as the other formula.

72V is almost 45x the rated voltage.

[JerryBurks]

I thought is was for all motors, both Nema 23 and 34...........

That is not clear to me, at least is does not make sense from an engineering perspective. The thermal resistance of one-stack NEMA34 is about 2.5 C/W and for the higher torque NEMA23 it is 4 or 5 C/Watt (from the Xylotec data sheet). That means the NEMA34 can dissipate twice the power for the same temperature rise (all other things equal like copper and iron losses). I just assumed Gecko would target the formula for NEMA23 because they are much more common. Maybe I should ask Mariss sometimes.

In any case I will stick one of the adhesive film thermometers on the motor to watch what happens.

[JohnZ]

Hmmm, this is very interesting. I have nearly the same nema 34 stepper motors with slightly different specs. (IE 2.6v vs 1.62 and 1200 oz in vs 1600 if I remember right ) and I run them with an 80 v power supply and they never even get warm. So, 2.6*25=65v max, but they are always cool to the touch. But this would mean that I am using too high of a voltage by about 25%

When I went from a 60v switched supply to an 80v torroid power supply, I nearly doubled my rapid speed and greatly increased the high speed torque where if I leaned into the machine while moving it could travel a lot faster before stalling.

Sent from my DROID4 using Tapatalk 2

[JerryBurks]

I hooked up the 72V supply today and the difference is indeed significant. I did a dry run with some random tool paths and it appears I can get to my target feed rate of 200-300 ipm with the old steppers but that still needs to be proven with a real machining run in meaningful depth. But so far....pretty good and maybe I don't need the smaller/faster stepper motors. However, I do have to set the acceleration to 10-15 in/sec^2. It takes quite a bit of torque to swing my 200 pound gantry around and at higher acceleration it will stall with these speeds. That is how it looks like at 250 ipm:

Dry Run