Got servos on my TAIG... SPEEEEED!

[Smertrios]

Had a xylotex stepper on my taig and although I have not used it much I remember how I hated the missed steps at what I would consider to be very slow (less than 12 ipm). I've had some gecko 320's, kelinginc nema23 servo motors (the smaller ones that they sell now) and a 36v torroid power supply for 4 years but due to circumstances out of my control I have not been able to get them working until now.

Watching the taig move at 12IPM and now seeing it move at 120 is really impressive to me =>.

[cameraguy]

Congratulations! I remember the days when I thought managing even 8 ipm was doing pretty well also. Xylotex-driven machines were just unusable. Those controllers were probably the single factor that drove more people away from benchtop cnc than everything else combined.

Glad to see you are up to speed now. Your will notice your tools will last a lot longer as well. I cut at 100+ ipm on my (stepper driven) Taig as well, and even now it still looks fast to me! Its just a matter of getting decent voltage and a decent controller, whether stepper or servo.

[fretsman]

Forgive my stupid question here as I do not have servos, but are you guys actually able to really get up to a full 120 IPMs? I would have thought that would be awfully hard on the screws and nuts on the Taig?

Thanks for sharing, I may look into this, can you share the part numbers/website pages of what you use?

Thanks so much-
Dave

[Smertrios]

G320's from www.geckodrive.com
KL23-120-36 (NEMA23) from www.kelinginc.net
300CPR usdigital encoders from www.kelinginc.net (1200 steps per rev)
36v Antec power supply from www.eBay.com
and a level shifter from cnczone member
oh... also have 3 CPU fans bought at www.tigerdirect.com (with heatsinks)

And I definitely do get 120IPM as for how hard it is on the screws I'll have to wait and tell you about that when I know more about wear but it doesn't sound bad on the mill at all.

I read another post about a user moving a TAIG mill at over 250 IPM but he said the vibrations were really bad. With my mill its smooth at 120IPM.

[cameraguy]

same here. taigs are built like tanks. smooth as silk at 100-120 ipm. i have moved it as fast as 150 ipm, but my computer isn't stable that fast. because of that i can't make it go faster simply due to the number of pulses per second it would require. so, i can't tell you what it does any faster than that.

its actually quite a bit smoother and quieter and has less vibration at 100ipm than trying to make the same cuts at slower feeds, just due to using the tooling at a more correct feedrate. that makes things better, not worse, for vibration.

i am using pacific scientific stepper motors and a 48v supply. the motors are wired for 2.1 mh inductance, so are matched well to a 48v supply. motors with any higher inductance than that need even more voltage to operate at full capacity. that one factor is probably the most important thing for performance, especially on a machine with such a high thread pitch as it requires such high rpm operation. that gives plenty of mechanical advantage though, so motor size becomes almost irrelevant, but higher voltages and matching motor inductance becomes critical instead.

[Fixittt]

as with anything else, the faster it goes the more wear it will accumulate.

The draw back about such speeds and the Taig are the dovetail ways and the metal backlash nuts. I know the taig is more of a machine then my old maxnc I had, but when I was able to get really fast speeds out of it, it really damaged the ways.

I now have machines with ballscrews and thompson shaft/bearing ways and reliable run 118/3000 mm a min. on steppers. One of the bad draw backs about such speeds and a small machine is that you run out of travel very very very quickly!!!!!!

Good job on the taig, and I too had the xylotex drive for 2 years. Chased gremlins on that damn thing!!!!!!

[cameraguy]

I know the taig is more of a machine then my old maxnc I had, but when I was able to get really fast speeds out of it, it really damaged the ways.

I can see that with a MaxNC, but not a Taig. IIRC, the MaxNC was a machine that didn't even have bearings (or at least more correctly relied only on the stepper motors bearings).
The stepper was unsupported and simply hung down under the end of the table by its top two mounting holes, and the leadscrew was a piece of common 1/4" allthread attached directly to the stepper shaft. That was the whole axis. That kind of setup was fragile at best, and 'rapids' - even on a Xylotex - could cause visible flexure of that system.



The Taig does has dovetail ways as well, but they have large brass gibs and an inordinate amount of bearing surface compared to the MaxNC. As well, the aluminum ways are anodized hardcoat. Not ways made from a painted, unmachined extrusion!
As well, the nuts are also metal, but they are long brass split nuts on 1/2" threads designed for being on a mill, not standard 1/4"-20 hex nuts running on allthread from the hardware store.

You aren't going to hurt a Taig at those speeds. The two are totally different beasts. It isn't possible to damage them at those speeds, and it is even damn hard to wear out one of those things without production work for a long time at those speeds (Ask Taig how many replacement saddles they sell. Not very many. Its damn hard to wear them out even at that speed.)

[Smertrios]

Here is a thread I started years ago when deciding to get servos...
http://www.cnczone.com/forums/taig_m...nginc_net.html

[Smertrios]

All I had hooked up yesterday was the X now I have all 3 axis setup and working (but not tuned). The X and Y have no problem moving at 120 IPM but the Z can't go much over 110 on the "up" stroke without faulting.

This is my setup. The enclosure is pretty nice I have a sheetmetal box with an acrylic window that sits in the front. The control box is non-existant the components are just sitting on the board (a few are screwed down).

[Smertrios]

I did the tuning "by ear" and the x,y can go upto 140IPM before faulting. I locked the x,y speed at 100. The Z was faulting its set to 80. I may drop the speeds a bit more as the movement is a little stiff but thats how I want it. That means machine temperature can have effect on the maximum speed before faulting so I may need to loosen the gibs or lower my max speed. Dont matter either way I am glad I have a machine capable of not loosing steps!

[fretsman]

Thanks for the replies to my question, fellas, I may look into building one of these with servos sometime.

Does anyone have any video of these running?

Thanks again,
Dave

[Jeff-Birt]

I like seeing folks who like to tweak their machines. Most folks think bigger is better and that is usually not the case. My mantra is "The key to getting good performance is to match the motors, drivers, and power supply to the machine." With my 'tiny' steppers I can get as good or better performance as any servo system that I have run across. As machine size increases servos become more of an advantage.

Here is a video of my Taig doing 250 IPM. This is not a practical speed by any means, lead screw whip is a problem as you ramp the speed up.

http://www.youtube.com/watch?v=tYn9lCgZPEw&list=UU6n7eWUxKSi6NPsycpa17iQ&index=11&feature=plcp]250 IPM SS - YouTube

Maximum speed on the rapids is often less important than acceleration. I had someone, who cuts a lot of lithographs, send me a file this past week to test out. This type of 3D profiling results in many thousands of very short movements in all three axis. Due to this the machine will almost never reach its commanded feedrate as all the axis are constantly changing direction.

Starting with fairly conservative acceleration setting of 6 in/s/s and a max velocity of 85 IPM (on all three axis), CV turned on and a largish look-ahead setting it took about 2:20 to run the 460,000 line program. Setting the acceleration up to 10 in/s/s dropped the time to 1:57. The commanded feed rate in the program is 65IPM but it will almost never get to that. The X and Y axis have more room to have the acceleration raised. The Z axis will not be able to be tweaked as high.

The point is a high maximum feed rate is not always the best performance metric. Tuning for acceleration can provide better results depending on your application.