[adprinter]

I have just completed an upgrade to my machine- and probably should have posed this question BEFORE the upgrade. My machine features a moving gantry design, with a single acme lead screw down the center of the X axis. The "upgrade" included pretty much a complete machine rebuild to a more sturdy frame design. (FROM 1" square tube steel, TO a steel frame constructed of 3"x4" tubes for the legs, and 1.5"x3" channel for the ends of the frame, and table supports, with 1.5" square tubes for the linear rails of the X axis which also saw an expansion FROM 48" length lead screw TO 72" lead screw).
The current lead screws I am using, are single-start acme thread 10 TPI, with 906 oz in steppers driving them on the X, Y and Z axis. The 4th axis is driven by a 1200 oz in stepper. The drive is a Gecko G540 with a 48 volt power supply. This arrangement has served me well, in carving mostly laminated Oak (recycled tractor-trailer flooring) for the past 3 years.
However, I want to be able to drive the machine at a much faster pace than the average 60 IPM I am getting now. After some comments I saw on other posts, it made me realize that my design is inefficient for it's main use.
I am aware that my current steppers and drive is capable of much more than I am currently getting out of them. What recommendations would you offer to increase speed, using the above listed steppers and drive?
(I.E. different lead screw ratios, rack and pinion drive, etc.) Please include suppliers with your recommendations. Thanks!

[uli12us]

With a stepper you have always the problem of the best compromise between speed and accuracy. If you double your speed, you earn half of the accuracy. If you want to increase both or increase speed and the accuracy should be at the same level you need servo motors. But first of all, you should invest in Ballscrews.

[louieatienza]

In my opinion, I don't believe you'll see any real benefit from having a very high resolution in wood. Wood does seem to machine better at higher feeds and spindle speeds. Multi-start ACME (5-start, .5" lead) would tremendously improve your machine's peformance, and you still have .0025" native resolution, which for wood is more than you'd ever need.

If you consider that the lead error on your screws, flex in your machine, even thermal expansion can exceed that, you'd see that chasing a higher mechanical resolution may not yield the results you want.

In the other thread, I showed a typical stepepr torque curve chart, which demonstrates how using a multi-start screw can give you more holding torque at the same speed compared to a single start screw. And that torque is available at a wider range of feeds.

Rack and pinion would be a great solution but would require far more work than a leadscrew swap. Ballscrews are great but you have to consider the cost factor. Plus I think you might run into problems with the high efficiency of the ballscrew coupled with the relatively light weight of yoru machine and the large steppers.

The fastest method? I'd have to say direct-drive linear servos! But obviously a very impractical solution! R&P and belt are capable of very high rapids, as well as ballscrews/servos.

[OCNC]

What steppers are you using? I'm curious to see what their spec's are.

Regarding the lead screw I agree with Louie. Roton is also a good source for lead screws.

[Dylwad]

High lead multi start leadscrews will do the trick, I use 1/2-10 5 start with 380 oz steppers and G251's and can easily rapid at 400+ ipm (have taken it to 700!), leadscrews are tensioned a bit to prevent whipping. .00025" per step is many times more accuracy than you need with wood, I hold .001" on Delrin on a daily basis. Y axis is 50" travel, x is 25", all 8020 and steel, moving gantry and heavy 2.2kw spindle.

[007CNC]

If you go with leadscrews, make sure you get something thick enough to avoid whipping. Maybe around 3/4" or 1" for a 0.5" lead screw.
Anything smaller will probably whip at higher speed.

Roton seems to be a good source for reasonably priced acme screws.
I bought a set of 3/4" hi lead screws with 0.5" lead (2 turns per inch of travel) early this month from them two 48" L and one 36" L on a 3/4" in diameter with 6 corresponding threaded mount plastic nuts: 3/4 X .500 Right Hand Hi-Lead Lead Screws & Nuts for Power Transmission - Roton Products, Inc.

I later changed my design to use a differen drive system, so I plan on throwing this on e-bay in the coming week or two for a fraction of what I paid for them. They are brand new still in he box. I know the 48s are too short for your X axis, but if interested in the nuts, let me know. I will still be posting on ebay, but may do it sooner if interested.

[ger21]

If you switched to 380oz motors you'd probably double your speed, as the G540 ca't drive a 906oz stepper nearly as fast as it can a smaller one.
Having said that, I'd either switch to a 1/2-10 5 start (or 1/2-8 4 start, which is comparable) or 1/2-8 8 start screw. The 8 start will be faster, but with lower resolution. The 5 start should be plenty fast. After you switch tio the 5 start, I'd be curious to try a 381oz stepper from Keling, because it might be even faster than your 906 oz motors. Unless every part of your system is a perfect match, larger steppers alone will rarely be faster.


At 72", the 8 start may be a better choice, as you'll possibly run into whipping issues with the 5 start.

Going to a larger diameter will cost quite a bit more, require more torque to accelerate, and are usually less efficient. However, you may find that you'll get whipping with the smaller screws.

YOu can get the screws at McMaster Carr or Roton, and nuts from DumpsterCNC.

[louieatienza]

If you switched to 380oz motors you'd probably double your speed, as the G540 ca't drive a 906oz stepper nearly as fast as it can a smaller one.
Having said that, I'd either switch to a 1/2-10 5 start (or 1/2-8 4 start, which is comparable) or 1/2-8 8 start screw. The 8 start will be faster, but with lower resolution. The 5 start should be plenty fast. After you switch tio the 5 start, I'd be curious to try a 381oz stepper from Keling, because it might be even faster than your 906 oz motors. Unless every part of your system is a perfect match, larger steppers alone will rarely be faster.


At 72", the 8 start may be a better choice, as you'll possibly run into whipping issues with the 5 start.

Going to a larger diameter will cost quite a bit more, require more torque to accelerate, and are usually less efficient. However, you may find that you'll get whipping with the smaller screws.

YOu can get the screws at McMaster Carr or Roton, and nuts from DumpsterCNC.

The only thing with the 8-start AB nut is, the last I checked, they were listed as currently unavailable from DumpsterCNC. You can get one from Kerk or Thomson, but they'll be a LOT more. My X axis screws can go 500+rpm with the accel at 40 before whipping, and that's only on one side where there is a little bend. I get over 720ipm with the accel at 100 on my Y, and probably can go faster if my computer would allow it.

I use 8-start on my machine; and hope Tim and DumpsterCNC starts producing them again...
[ame=http://www.youtube.com/watch?v=d6MBKm9Ujbo&feature=plcp]GeckoDrive G540 Installed in my DIY CNC Router - YouTube[/ame]

[louieatienza]

Another view of the 8-start leadscrew in action. The higher accel of the Y axis is apparant here; and I now try to take advantage of this whenever possible when porgramming toolpaths.

[ame=http://www.youtube.com/watch?v=Z8d-V2Hh0GA&list=UUh02xArFBxUbWIhO2HPsxfQ&index=9&feature=plcp]DIY CNC Router, Aluminum Milling High Speed Motion - YouTube[/ame]

[adprinter]

After reading all of the replys, I went to the Roton site to research availability. If I am comprehending correctly, the 5/8" 2 start should yield a 150% overall increase in speed (over my current 5/8" 1 start 10 TPI acme screw). Which Roton has rated at 0.100" travel per revolution, the 2 start is rated at 0.250" travel per revolution. The only 8 start screws they seem to offer are 7/16" diameter. Which I feel would just be like driving the 200 lbs. gantry using a wet noodle. I regret that I did not take this step, prior to the upgrade. I also researched their prices for ballscrews and nuts. (Since friction on the acme threads was an issue with the new screw and nut overheating and trying to seize).
After some thought, it appears that the best overall improvement could be had for around $550 plus shipping. (5/8" ballscrews and nuts on XYZ axis). At least now I have some idea of the budget I will need, in order to increase the speed of the machine (without the need to COMPLETELY redesign from the electronics on up!). Of course, this assumes the speed increase would be realized from the increase in travel from the same RPM output of my current motors.
However, I am aware that this thinking is in theory only. Since the motor tuning which would no doubt be required in Mach3 may actually result in different motor outputs in terms of actual speed realized.
To answer the question about the current motors I am using, they are the Keling KL34H295-43-8B wired Bipolar Parallel connection with 3K ohm resistors to limit current to 3 amps. These were used on the XY and Z axis of my machine. I went with the KL34H2160-62-8B motor on the 4th axis. Wired Bipolar Parallel, also with a 3K ohm resistor. All are powered with the 48VDC power supply sold by Keling, and driven by a Gecko G540 and Mach3 with the pulse rate set at 35 Khz. 1/2" single start 10 TPI acme lead screws are used on the Y and Z axis.
While the 4th axis is a direct belt drive (Thus the reason for the larger, 1200 oz in motor). I forget what the actual belt ratio is, (the 4th axis project was completed over 2 years ago) but I think it is something like a 64 tooth belt, with motor gear of something like 14 teeth, and drive gear of 36 teeth. It was just something I threw together, after several failed attempts at finding absolute center of the motor shaft, with absolute center of a 10-1 ratio worm gear box I had in the shop. The belt drive is much more forgiving, in terms of shaft alignment. I toyed with the 4th axis briefly after completing that project. But it has remained largely unused since then. And in fact, is currently unbolted from the machine, and slid to one end out of the way (until I can afford to buy two 1.5"x1.5"x80"x1/4" angle steel for permanent mounting).

[louieatienza]

After reading all of the replys, I went to the Roton site to research availability. If I am comprehending correctly, the 5/8" 2 start should yield a 150% overall increase in speed (over my current 5/8" 1 start 10 TPI acme screw). Which Roton has rated at 0.100" travel per revolution, the 2 start is rated at 0.250" travel per revolution. The only 8 start screws they seem to offer are 7/16" diameter. Which I feel would just be like driving the 200 lbs. gantry using a wet noodle. I regret that I did not take this step, prior to the upgrade. I also researched their prices for ballscrews and nuts. (Since friction on the acme threads was an issue with the new screw and nut overheating and trying to seize).
After some thought, it appears that the best overall improvement could be had for around $550 plus shipping. (5/8" ballscrews and nuts on XYZ axis). At least now I have some idea of the budget I will need, in order to increase the speed of the machine (without the need to COMPLETELY redesign from the electronics on up!). Of course, this assumes the speed increase would be realized from the increase in travel from the same RPM output of my current motors.
However, I am aware that this thinking is in theory only. Since the motor tuning which would no doubt be required in Mach3 may actually result in different motor outputs in terms of actual speed realized.
To answer the question about the current motors I am using, they are the Keling KL34H295-43-8B wired Bipolar Parallel connection with 3K ohm resistors to limit current to 3 amps. These were used on the XY and Z axis of my machine. I went with the KL34H2160-62-8B motor on the 4th axis. Wired Bipolar Parallel, also with a 3K ohm resistor. All are powered with the 48VDC power supply sold by Keling, and driven by a Gecko G540 and Mach3 with the pulse rate set at 35 Khz. 1/2" single start 10 TPI acme lead screws are used on the Y and Z axis.
While the 4th axis is a direct belt drive (Thus the reason for the larger, 1200 oz in motor). I forget what the actual belt ratio is, (the 4th axis project was completed over 2 years ago) but I think it is something like a 64 tooth belt, with motor gear of something like 14 teeth, and drive gear of 36 teeth. It was just something I threw together, after several failed attempts at finding absolute center of the motor shaft, with absolute center of a 10-1 ratio worm gear box I had in the shop. The belt drive is much more forgiving, in terms of shaft alignment. I toyed with the 4th axis briefly after completing that project. But it has remained largely unused since then. And in fact, is currently unbolted from the machine, and slid to one end out of the way (until I can afford to buy two 1.5"x1.5"x80"x1/4" angle steel for permanent mounting).

Keep in mind the ballnuts from Roton are NOT anti-backlash, so you will need two per screw, plus some form of adding tension between the two, to remove play. Also if you look further, Roton makes both a 5/8"-8, and 3/4"-8, 2-start (.250" lead) screws. McMaster sells a wide variety of ACME scrwes; i'd check there too.

It isn't theory, it's fact: You WILL see an increase un performance, and it may be BETTER than you expect, since you're utilizing more of the stepper's potential. I originally built my machine with the same screws as yours (just to get it up and running) and maybe topped out at 50-70ipm rapids.

There are many machines here that run .2" and .5" leads and achieve high rapids and have plenty of available power, with gantries way heavier than yours. Heck my gantry weighs about 90lbs, and I run it with 8-start scrwes and a puny 425on-oz stepper. I can't stall my machine b leaning on the gantry when at cutting speeds.

[007CNC]

After reading all of the replys, I went to the Roton site to research availability. If I am comprehending correctly, the 5/8" 2 start should yield a 150% overall increase in speed (over my current 5/8" 1 start 10 TPI acme screw). Which Roton has rated at 0.100" travel per revolution, the 2 start is rated at 0.250" travel per revolution. The only 8 start screws they seem to offer are 7/16" diameter. Which I feel would just be like driving the 200 lbs. gantry using a wet noodle. I regret that I did not take this step, prior to the upgrade. I also researched their prices for ballscrews and nuts. (Since friction on the acme threads was an issue with the new screw and nut overheating and trying to seize).
After some thought, it appears that the best overall improvement could be had for around $550 plus shipping. (5/8" ballscrews and nuts on XYZ axis). At least now I have some idea of the budget I will need, in order to increase the speed of the machine (without the need to COMPLETELY redesign from the electronics on up!). Of course, this assumes the speed increase would be realized from the increase in travel from the same RPM output of my current motors.
However, I am aware that this thinking is in theory only. Since the motor tuning which would no doubt be required in Mach3 may actually result in different motor outputs in terms of actual speed realized.
To answer the question about the current motors I am using, they are the Keling KL34H295-43-8B wired Bipolar Parallel connection with 3K ohm resistors to limit current to 3 amps. These were used on the XY and Z axis of my machine. I went with the KL34H2160-62-8B motor on the 4th axis. Wired Bipolar Parallel, also with a 3K ohm resistor. All are powered with the 48VDC power supply sold by Keling, and driven by a Gecko G540 and Mach3 with the pulse rate set at 35 Khz. 1/2" single start 10 TPI acme lead screws are used on the Y and Z axis.
While the 4th axis is a direct belt drive (Thus the reason for the larger, 1200 oz in motor). I forget what the actual belt ratio is, (the 4th axis project was completed over 2 years ago) but I think it is something like a 64 tooth belt, with motor gear of something like 14 teeth, and drive gear of 36 teeth. It was just something I threw together, after several failed attempts at finding absolute center of the motor shaft, with absolute center of a 10-1 ratio worm gear box I had in the shop. The belt drive is much more forgiving, in terms of shaft alignment. I toyed with the 4th axis briefly after completing that project. But it has remained largely unused since then. And in fact, is currently unbolted from the machine, and slid to one end out of the way (until I can afford to buy two 1.5"x1.5"x80"x1/4" angle steel for permanent mounting).

I think you can get even better performance by going higher on your travel per revolution, at least 0.5" travel per revolution, which I believe is a good ballance between speed and resolution.

My very first built was with a 0.5 dia 0.1 lead acme screw, 36V supply. I maxed out at about 70ipm. I swapped those out with 5/8 Roton ballscrew and 48V supply, and I was up tp 300 IPM before it starts whipping.

0.25" lead will give you some improvement over what you have now, but I dont think you will be getting the full potential of your setup in terms of speed.

For acme screws, Roton has information on efficiency for the difference sizes and leads. Some are counter-intuitively more efficient than others. I would take a look at that too.

[OCNC]

To answer the question about the current motors I am using, they are the Keling KL34H295-43-8B wired Bipolar Parallel connection with 3K ohm resistors to limit current to 3 amps. These were used on the XY and Z axis of my machine.

So you're running those motors at 1/2 of their rated current and 83% of their rated voltage (48/(32*3.3^1/2)).

I use a Roton .25dia x .25 Hi lead screw on my Z axis with the single lead nut. I haven' t had a problem with backlash but am only cutting wood. I find that Z is more limited by acceleration than top speed. It's such a short axis.

Chris

[adprinter]

Keep in mind the ballnuts from Roton are NOT anti-backlash, so you will need two per screw, plus some form of adding tension between the two, to remove play.

Louie,
Have a look at the attached Zerobacklash.PDF and tell me your thoughts on this. The prototype build (on a smaller scale) exhibited ZERO back lash, when reversing travel directions. (I.E. there was no hesitation of travel movement between rotational movement of the lead screw). However, I have not yet built one to size to test on the machine. Ideas are cheap, putting them into practise is the expensive part!

[louieatienza]

There was a guy on the mechanical forum who built a similar nut based on this idea, for a lathe. The thread was about roller screws, which are a different animal entirely. While you havve less friction, I don't think this would work well with a higher lead screw, since it would tend to "slip" rather than "roll" on the screw thread. So if you used a finer pitch (like what you have) you still end up with the same problem of screw lead/stepper mismatch.

This concept is also used in "threadless" lead screws, where the bearings are skewed and ride against a smooth hardened and ground rod.