alternative stepper mounts

[awetmore]

I'm wondering if anyone has made more compact stepper mounts for the Taig. It would be great to reduce the size of my enclosure a bit and the cantilevered out steppers really do increase the size requirements.

Switching to a pulley system would also let me use a 2:1 drive ratio and potentially faster speeds.

Is backlash typically a problem with pulley based systems?

On a related topic, does anyone have a good technique for finding the maximum amperage rating for an unlabelled stepper? I have an early PacSci PowerMax that has a non-standard model number. It doesn't get hot (or even warm) with a 2.3amp setting on my G540, I'm wondering if I should run it even higher?

thanks,
alex

[Fastest1]

I believe when pulleys are used it is to reduce the speed and gain torque. Maybe if you are cutting foam. Are you running Mach 3? Is it set for 25Hz? Increase the Hz and it will increase the speed the way I understand it. Is it too slow now? I dont think a pulley system introduces backlash any more than any other system. Any play will cause backlash whether it is direct, gear or belt driven.

[awetmore]

That is how pulleys are typically used. It seems like they could be run in reverse though if you have a high (for a Taig) torque stepper.

I'm not using Mach3, I'm using EMC2. My machine is tuned pretty well, but I'm still limited to a roughly 45khz pulse stream. With a microstepping G540 and the high TPI Taig leadscrews (together they require 40k pulses to go an inch) that is only 1.1 inches per second, or 66ipm. That also means that I'm running my steppers close to their 1500rpm limit.

I think I'm going to try it, mostly to shrink the setup, but also to see if I can get faster rapids reliably. The pulleys and belts aren't that expensive, so it is mostly a matter of CNCing the mounts. I'll report back.

I care about the overall envelope of the machine because I'm using a 24x36" plastic tub as the sink that catches coolant under the machine. With the stock motor mounts I need about 38" on the X to use the full travel of the table. The stock Taig mounts are rather wide (around 3"), so that is an easy place to save space.

The other option would be to switch to a single stack stepper, and that might be simpler. Since my motors have a higher than ideal inductance anyway that is probably the approach that I should use. Jeff sells some nice looking steppers that aren't too expensive ($34 each).

alex

[Jeff-Birt]

I can get about 100 IPM reliably with my G540/SmoothStepper/motor combo. Don't overlook the quality of the pulse stream that your parallel port puts out as a source of problems. While Mach3 and EMC2 work a little differently both are hobbled by the parallel port.

[awetmore]

Smoothstepper doesn't have drivers for EMC2, and I don't know of any other similar products that do. I'm not interested in switching to Mach3, I've used it before and prefer EMC2.

It looks like my parallel port is average to above average compared to other results that I've seen for EMC2. The G540/Taig combo just requires a very high pulse rate due to the 10x microstepping and high tpi leadscrews.

My biggest goal is really bringing down the envelope of the machine, not increasing IPM. I just saw this as a way to do both and was curious if anyone else had tried it.

[Fastest1]

Alex, I am curious if you use a stepper driver that doesnt use the 10x stepping such as a Keling or even a Gecko (not the G540). Doesnt it move farther with less steps? This comes at the cost of resolution doesnt it? Jeff always advocates the SS. I believe there are other answers or possibilities also. I have a few machines with gear reduction drives, these machines require many steps to make any real movement. I was informed by another individual who builds machines for a living and who had modded the same model lathe I have to use a driver with adjustable microstepping. I went with the G540 which is a great set up in regards to ease of connecting but I wonder how much potential speed I lost. The machine will still cut just fine and very accurately but how about threading speed being affected?

[awetmore]

Yes, I previously used a ~10 year old SuperCam driver that was only half-step. For one revolution of the motor I'd feed it 400 steps, vs 2000 for the G540. With a 20tpi leadscrew that means 8,000 steps for the SuperCam per inch, or 40,000 for the G540.

The G540 is a better driver in pretty much every way, but it would be nice to be able to pick the microstepping level. I have to wonder if something like a 4x multiplier would give the advantages of microstepping without requiring such a fast pulse stream.

Since I just rewired my whole control setup around the G540 I'm not planning on changing drivers again.

[Fastest1]

I built a Hobby Cnc Pro kit as my first driver/controller. I remember having adjustable stepping. As stated below. The G540 is only 1/10 if I understand correctly. So if I chose a 1/1 step, does this mean I could do the same distance with 1/10th of the steps? And my resolution would be 1/10 th of what is right now too? Just being hypothetical but my resolution is probably lower than I can realistically acheive. Didnt meant to highjack your thread as much as clarify operations.

3 or 4 Axis Unipolar Chopper control.
Individual OR simultaneous control of 2/4 Phase Stepper Motors.
Accepts 5, 6, or 8 wire stepper motors only. 4 wire types are not usable.
42VDC maximum input voltage, 12VDC minimum input voltage. 24VDC Minimum Recommended Voltage.
3.0 Amps Maximum per Phase, 500ma (.5A) minimum. Each axis adjustable throughout this range.
1/1, 1/2, 1/4, 1/8, and 1/16 Microstepping.
Step and Direction Control.
Idle Current Reduction to 50% when idle for 10 seconds. Can change this time delay.
Built In Protection Circuit to help against blown chips on stepper motor short or open connections.
3.7″ by 6.8″ double sided with top silkscreen and thru plated holes and lead free solder PCB. (RoHS compliant).
Power On Reset.
On board voltage regulation for 5-volt logic with 24VDC cooling fan from motor power supply.
On board connections for home and limit switches with 10K pullup resistor provided to each. No need for a separate “break out” board.
Minimum of components to make assembly fast and easy!
Works with Step and Direction software such as DeskNC, TurboCNC, Mach3, etc

Yes, I previously used a ~10 year old SuperCam driver that was only half-step. For one revolution of the motor I'd feed it 400 steps, vs 2000 for the G540. With a 20tpi leadscrew that means 8,000 steps for the SuperCam per inch, or 40,000 for the G540.

The G540 is a better driver in pretty much every way, but it would be nice to be able to pick the microstepping level. I have to wonder if something like a 4x multiplier would give the advantages of microstepping without requiring such a fast pulse stream.

Since I just rewired my whole control setup around the G540 I'm not planning on changing drivers again.

[Jeff-Birt]

The G540 has built in micro-step morphing. It changes from 10x microstepping to full steps as the RPMs increase. This gives you all the smoothness of a 10x microstep at low speeds and all the power of full steps at higher speeds.

Smoothstepper doesn't have drivers for EMC2, and I don't know of any other similar products that do. I'm not interested in switching to Mach3, I've used it before and prefer EMC2.

There are motion control boards for EMC2. Most of them I have run across are much more expensive than the SS though. I think Jon Elson sells some sort of motion controller for EMC though.

My biggest goal is really bringing down the envelope of the machine, not increasing IPM. I just saw this as a way to do both and was curious if anyone else had tried it.

What do you mean?

[strohkirchw]

I believe he means physical dimension of the machine.

[MrWild]

The G540 has built in micro-step morphing. It changes from 10x microstepping to full steps as the RPMs increase. This gives you all the smoothness of a 10x microstep at low speeds and all the power of full steps at higher speeds.



There are motion control boards for EMC2. Most of them I have run across are much more expensive than the SS though. I think Jon Elson sells some sort of motion controller for EMC though.




What do you mean?

While the G540 morphs it's step pulses to the motor, it still gets the x10 step stream from the computer no matter how fast/slow the motor is turning. It's the pulse stream from the computer that is causing the slow down.

Yes, the motion cards are more expensive than the smooth stepper, but they also have all the bugs worked out and the suppliers don't go on vacation or get caught up in other projects leaving their customers holding the bag. The motion controllers also are more capable and work well with servos.

[Jeff-Birt]

My biggest goal is really bringing down the envelope of the machine, not increasing IPM. I just saw this as a way to do both and was curious if anyone else had tried it.

Let's say you flip the motor on the X-axis around and run a belt between the motor and the leadscrew. You would not gain much without loosing travel as now the back of the stepper could contact the machine. For the Y axis I guess you could tuck the motor down and to the side. This might save you a few inches in each dimension but that is about it.

The G540 is only 1/10 if I understand correctly. So if I chose a 1/1 step, does this mean I could do the same distance with 1/10th of the steps? And my resolution would be 1/10 th of what is right now too? Just being hypothetical but my resolution is probably lower than I can realistically acheive. Didnt meant to highjack your thread as much as clarify operations.

Resolution is different from precision. With a 10x microstepping drive and 1.8 degree steppers you have a resolution of 40,000 steps/inch. That does not mean you have 1/40,000" of precision though. Resolution only implies the number of distinct positions not how precise each position is. It is the increased resolution that gives you really smooth motion at low RPMs.

It looks like my parallel port is average to above average compared to other results that I've seen for EMC2. The G540/Taig combo just requires a very high pulse rate due to the 10x microstepping and high tpi leadscrews.

I think your assessment is correct, but I would point out that even an above average parallel port really sucks. Both EMC2 and Mach are hobbled by sticking with the parallel port. It seems like a cheap way of doing things but many times folks wind up spending a lot of money going through several iterations of motors, driver, power supplies, etc when the underlying problem is really the step pulse quality. I really wish there was a better selection of external motion control cards for both systems as I believe it is the way of the future.

Here is a list of known EMC2 compatible hardware I found: EMC Documentation Wiki: EMC2 Supported Hardware, I wish there was a better selection.

Yes, the motion cards are more expensive than the smooth stepper, but they also have all the bugs worked out and the suppliers don't go on vacation or get caught up in other projects leaving their customers holding the bag. The motion controllers also are more capable and work well with servos.

Well I'll have to take your word for it. What a dedicated bunch of manufacturers to never take a vacation or work on any other products.