My 28v power supply has an amp meter built into it so I can see how much current is being drawn by the steppers. You should be able to put your multimeter inline with your drivers and see how much they are pulling. .05a doesn't sound like hardly enough to make those steppers move, especially at only 13v.
If you have a 24v power supply handy, I would try pushing that much to them at least, assuming your driver can handle that voltage.
I'm busy designing a Prop based MosFet driver for an XYZ table I am building. Through the Prop, for each step I can vary the step duration, acceleration, initial current in the step, then PWM'd current limit with variable on/off duration or ratio, and the length of off time before the next step. About seven things to adjust for each step, and differently for all speeds.
So through this exercise I find that the 1.8 degree unipolar motors I have with a rating of 2.59 Volts, 1.9A and 55 oz.in. holding torque will run very nicely with a supply of 24 volts at up to 5000 steps per second. I have not yet determined the optimum drive prameters...... probably end up being a bunch of tables or speed based algorithms. I can get ample torque full stepping at all speeds with currents in the 100 or so mA range. The resonance issues appear to be resolvable through profile adjustments, but lots more characterization needs to be done.
Is your 1500RPM you are talking about actual or calculated? I can't believe you will get your full 55 oz.in holding torque with only 100mA going to it. That seems really low.
The 5000 steps/sec (1500 rpm) is real, in fact, in messing with the drive waveforms I was able to coax it to about 6 or 7 thousand but not enough torque left at those speeds. So unless there is some breakthrough with the algorithms, I expect with 24 volts a practical top end to be around that 5K.
The 55 oz.in. rating on the motor, I'm sure would be its stall torque; I have not attempted to measure what the pull-out torque might be at any speed. My assessment of "ample" is that it drives my (small 11 inch) table with power to spare. At no time did I want to imply the nameplate torque rating was what I had at any particular speed.
Steppers can run very fast if they are setup correctly. I've pushed my steppers to about 2500 RPM, which is way faster then the system was designed for, but at that speed, there is almost no torque and the slightest pressure on the bearings or drive screw causes it to stall, and it also needs a bit of ramp up and ramp down to get to that speed. At my normal running speed of around 80ipm (800rpm) they have a huge amount of torque when combined with 10tpi acme thread drive screws. The torque that the stepper motors are rated at is the holding torque, not it's rotational torque. I'm not sure how to calculate the rotational torque. Holding torque means how much force it would take while in a stopped position to cause it to move. My steppers are rated for 230oz.in if I'm not mistaken.
Wow, those are huge motors. Mine are a much smaller frame, for obviously smaller requirements. And of course mine have the very low voltage rating, that keeps the coil inductance low, and hence easier to push faster.
I'm actually expecting to drive all three axes of the table simutaneously off a single 24 volt switching style 1 amp wall wart.
The spindle might be another storey though. I expect about 1/4 horsepower at 60,000 rpm for pc board drilling and routing. Probably need to make an air bearing for that..... we'll see when I get further along.
The steppers are NEMA23 size, but they are high torque and rated for micro stepping. Now that I think of it, at full speed, I should be pushing out at least 20,000 steps/sec because I am using them in 1/2 step mode. I'm also pushing around a 42" wide gantry (36" usable range) with a 2.25HP variable speed router attached to it so there is quite a bit of weight to deal with. Do you use any type of encoders to keep track of your positions or anything? I have the encoders, but that will be a next winter project to install them.
But to confirm that, soon I will set the table up for a continuous run of a million end to end strokes (will take about 3 weeks 24 hrs/day) to get a sense of the life of the mechanism and bearings I've designed. After that I will also have some idea on the stepping reliability. If I lose counts, then some encoding mechanism might be called for, in which case I will simply put a line disk with quadrature opfical sensing on the motor shafts. In assembler the Prop can easily handle the extra work load.
It's a little early to determine some of the details.... I'm still in the concept confirmation stage; bearing, lead screws, motor drives etc. I have lead screws with 0.1 inch as well as 0.125 inch pitch, and will not settle on a final choice until I see the wear results after life tests. In any case, it will be not faster than the steppers can run, so possibly the answer will be in the 3 inches per second range. Kind of crappy I know, but that's the problem with screws; they give you good forces with great precision but low speeds, whereas belt drives are great speed but low force and poorer precision. The alternate of course is lead screws with servo motors, but that's another whole level of complexity that I don't wish to take on at this point, and I do intend to build it all, including motor drives myself. Perhaps after I get this first one done, I will re-evaluate for servos.
Just a word of caution about your planned test.· Keep in mind that your moves will be at a constant speed (except for acc & dec) as well as a static load.· During normal use, your machine will encounter resistance to movement from many things including cutting forces and mechanical binding.·
Steppers seem to have problems at certain velocities and not specifically at high velocity.·For my testing without cutting, I have also used a large circular motion.· That runs the steppers at various velocities which exposes them to more of a chance to hit that horrid spot where the motors go into "Spaz" mode.
It is very difficult to simulate cutting forces as they are widly varied.·
If you allow a sufficient safety margin, steppers can run perfectly reliable without lost steps.
I have been able to coax the unit to rapaid at 420 inches/min with 24 Volt drive and complex acceleration algorithms, so that 7000 steps per second is the limit of these motors at that voltage, and I don't want to go higher. I'm now using a 0.2 pitch leadscrew, and that gives me a nice 0.001 inch per step.
@Chris_D
You are absolutely right about leaving some margin, so I expect to get reliable results at 5000 steps/second. With the higher pitch (or is it lower) screw, that is still not too bad rapiding at 5 inches per second. So far no real resonance issues.... these also are quite waveform dependant. But this is all without much load, so much more work to be done.
Until these machines are capable of .001 for drilling
I used to use a donut with a open center.
This provides a centering action for the hand drilling
and is probably good to .001 with a decent operator
and a carbide PCB drill running at dremel max speeds.
Used to raise the board into the drill for centering first
as it pocks the hole, then lower the board/drill together
for the final pass through.
Downward tension keeps the PCB from climbing the
drill bit, but a spring loaded shoe will suffice to keep
from snapping the drill bits, until you can get the technique down.
Use a phenolic backer, rated for PCB drill stacks as
the Home Depot types will smear resin back into
the hole on withdrawal.
No reason to wait for the CNC to make PCB's,
even plated through holes work with manual methods.
Photo tools are really the only choice for small efforts,
and the milling has it's own problems that just don't
make it at most of today's dimensions.
I've just started using my own CNC machine to make PCBs. So far I'm happy with the results, but Leon is right - they're not likely to take you to the level of precision needed for SMT parts.
So far I've gotten mine to do a board with roughly 0.25mm traces, and 0.25mm isolation. The board was a breakout for a 0.5mm pitch ribbon connector. It worked, but it was just a small breakout board and it still took close to half an hour. I'm using a V-shaped cutter, so doing finer work means you need to make more passes to cut the same isolation width.
It -IS- very nice that it does the drilling for you, and I plan to use this method for a while just to get my board design chops up. It's much easier than soldering wires by hand, and I already had the CNC machine. I'm not sure I'd build one for the sole purpose of PCB etching.
I agree that mechanical milling of PC boards is not suitable for fine detail. In fact I still have a complete albeit rather dated full commercial board printing facility (now dormant hopefully to be resurected and modernized this year), so I'm fully aware of what's involved in the etching process.
But my application for a tabletop CNC machine is for much more than that, and since it will be available, I am pursuing that option for simple things like quick turn one-of break-out boards and the like. And also carving 900 MHz antennas for experimenting, so I can re-mount the board and carve a little more, all with some precision.
I finally got to work on this project now that school and the ACT's are over with. I have purchased the MDF, made some sketchup plans, and gotten some prop code together to drive the steppers. So far, I have gotten the prop to drive the stepper in single step mode without too many problems. However, I put together some code to do microstepping, and the current draw of the stepper kept re-setting the propeller every time the stepper moved. To get to the point, would I want to put in a power resistor to limit the stepper motor's current? All the resistors I have on hand ended up at approximately the same temperature as my soldering iron after 10 seconds...
Also, I am clueless as to what bearings I should get. Is there some criteria to look for when purchasing bearings? Finally, what type of coupling would I need to fasten the stepper shafts to the threaded rods?
Thanks.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ MCU Projects·- my new site where I will be posting all projects, code, etc.
Quit buying all those fixed voltage regulators, and·get anAdjustable Power Supply·for your projects!· Includes an LED testing terminal!
SD Card Adapter·- Add extra memory to your next Propeller project with ease!
I got the driver parts finally, so I was able to build the circuitry for driving a stepper.
Soon I'm going to be able to work on the actual construction of the machine frame, but that will be after I finish the driver. Would it be better for me to drive my steppers in half step or full step? With half step, I can get an ideal resolution of 1/8000", and with full step, 1/4000". I don't think I'll be able to microstep them, because that seems to require some crazy current modulation.
Which step sequence would be better?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ MCU Projects·- my new site where I will be posting all projects, code, etc.
Quit buying all those fixed voltage regulators, and·get anAdjustable Power Supply·for your projects!· Includes an LED testing terminal!
SD Card Adapter·- Add extra memory to your next Propeller project with ease!
I would recommend neither of those, but instead to get a motor/screw combination that will give you 0.001 inch steps, or even coarser. Your full step suggestion of 4000 per inch will make the unit very slow to operate, and 8000 to be doubly so.
Your suggested high-res configurations are not needed..... they will be "lost in the noise", as there is no way your mechanics can take advantage of those very fine positioning increments.
Comments
If you have a 24v power supply handy, I would try pushing that much to them at least, assuming your driver can handle that voltage.
I'm busy designing a Prop based MosFet driver for an XYZ table I am building. Through the Prop, for each step I can vary the step duration, acceleration, initial current in the step, then PWM'd current limit with variable on/off duration or ratio, and the length of off time before the next step. About seven things to adjust for each step, and differently for all speeds.
So through this exercise I find that the 1.8 degree unipolar motors I have with a rating of 2.59 Volts, 1.9A and 55 oz.in. holding torque will run very nicely with a supply of 24 volts at up to 5000 steps per second. I have not yet determined the optimum drive prameters...... probably end up being a bunch of tables or speed based algorithms. I can get ample torque full stepping at all speeds with currents in the 100 or so mA range. The resonance issues appear to be resolvable through profile adjustments, but lots more characterization needs to be done.
More later.
Cheers,
Peter (pjv)
The 5000 steps/sec (1500 rpm) is real, in fact, in messing with the drive waveforms I was able to coax it to about 6 or 7 thousand but not enough torque left at those speeds. So unless there is some breakthrough with the algorithms, I expect with 24 volts a practical top end to be around that 5K.
The 55 oz.in. rating on the motor, I'm sure would be its stall torque; I have not attempted to measure what the pull-out torque might be at any speed. My assessment of "ample" is that it drives my (small 11 inch) table with power to spare. At no time did I want to imply the nameplate torque rating was what I had at any particular speed.
Cheers,
Peter (pjv)
Post Edited (pjv) : 5/7/2010 9:18:05 PM GMT
Wow, those are huge motors. Mine are a much smaller frame, for obviously smaller requirements. And of course mine have the very low voltage rating, that keeps the coil inductance low, and hence easier to push faster.
I'm actually expecting to drive all three axes of the table simutaneously off a single 24 volt switching style 1 amp wall wart.
The spindle might be another storey though. I expect about 1/4 horsepower at 60,000 rpm for pc board drilling and routing. Probably need to make an air bearing for that..... we'll see when I get further along.
Cheers,
Peter (pjv)
Post Edited (pjv) : 5/7/2010 9:47:29 PM GMT
No, I expect to be OK without encoders.
But to confirm that, soon I will set the table up for a continuous run of a million end to end strokes (will take about 3 weeks 24 hrs/day) to get a sense of the life of the mechanism and bearings I've designed. After that I will also have some idea on the stepping reliability. If I lose counts, then some encoding mechanism might be called for, in which case I will simply put a line disk with quadrature opfical sensing on the motor shafts. In assembler the Prop can easily handle the extra work load.
Cheers,
Peter (pjv)
It's a little early to determine some of the details.... I'm still in the concept confirmation stage; bearing, lead screws, motor drives etc. I have lead screws with 0.1 inch as well as 0.125 inch pitch, and will not settle on a final choice until I see the wear results after life tests. In any case, it will be not faster than the steppers can run, so possibly the answer will be in the 3 inches per second range. Kind of crappy I know, but that's the problem with screws; they give you good forces with great precision but low speeds, whereas belt drives are great speed but low force and poorer precision. The alternate of course is lead screws with servo motors, but that's another whole level of complexity that I don't wish to take on at this point, and I do intend to build it all, including motor drives myself. Perhaps after I get this first one done, I will re-evaluate for servos.
P.S. Vaati .... sorry for hijacking your thread!
Cheers,
Peter (pjv)
Just a word of caution about your planned test.· Keep in mind that your moves will be at a constant speed (except for acc & dec) as well as a static load.· During normal use, your machine will encounter resistance to movement from many things including cutting forces and mechanical binding.·
Steppers seem to have problems at certain velocities and not specifically at high velocity.·For my testing without cutting, I have also used a large circular motion.· That runs the steppers at various velocities which exposes them to more of a chance to hit that horrid spot where the motors go into "Spaz" mode.
It is very difficult to simulate cutting forces as they are widly varied.·
If you allow a sufficient safety margin, steppers can run perfectly reliable without lost steps.
Chris
I have been able to coax the unit to rapaid at 420 inches/min with 24 Volt drive and complex acceleration algorithms, so that 7000 steps per second is the limit of these motors at that voltage, and I don't want to go higher. I'm now using a 0.2 pitch leadscrew, and that gives me a nice 0.001 inch per step.
@Chris_D
You are absolutely right about leaving some margin, so I expect to get reliable results at 5000 steps/second. With the higher pitch (or is it lower) screw, that is still not too bad rapiding at 5 inches per second. So far no real resonance issues.... these also are quite waveform dependant. But this is all without much load, so much more work to be done.
Cheers,
Peter (pjv)
Until these machines are capable of .001 for drilling
I used to use a donut with a open center.
This provides a centering action for the hand drilling
and is probably good to .001 with a decent operator
and a carbide PCB drill running at dremel max speeds.
Used to raise the board into the drill for centering first
as it pocks the hole, then lower the board/drill together
for the final pass through.
Downward tension keeps the PCB from climbing the
drill bit, but a spring loaded shoe will suffice to keep
from snapping the drill bits, until you can get the technique down.
Use a phenolic backer, rated for PCB drill stacks as
the Home Depot types will smear resin back into
the hole on withdrawal.
No reason to wait for the CNC to make PCB's,
even plated through holes work with manual methods.
Photo tools are really the only choice for small efforts,
and the milling has it's own problems that just don't
make it at most of today's dimensions.
jr
So far I've gotten mine to do a board with roughly 0.25mm traces, and 0.25mm isolation. The board was a breakout for a 0.5mm pitch ribbon connector. It worked, but it was just a small breakout board and it still took close to half an hour. I'm using a V-shaped cutter, so doing finer work means you need to make more passes to cut the same isolation width.
It -IS- very nice that it does the drilling for you, and I plan to use this method for a while just to get my board design chops up. It's much easier than soldering wires by hand, and I already had the CNC machine. I'm not sure I'd build one for the sole purpose of PCB etching.
I agree that mechanical milling of PC boards is not suitable for fine detail. In fact I still have a complete albeit rather dated full commercial board printing facility (now dormant hopefully to be resurected and modernized this year), so I'm fully aware of what's involved in the etching process.
But my application for a tabletop CNC machine is for much more than that, and since it will be available, I am pursuing that option for simple things like quick turn one-of break-out boards and the like. And also carving 900 MHz antennas for experimenting, so I can re-mount the board and carve a little more, all with some precision.
Cheers,
Peter (pjv)
I finally got to work on this project now that school and the ACT's are over with. I have purchased the MDF, made some sketchup plans, and gotten some prop code together to drive the steppers. So far, I have gotten the prop to drive the stepper in single step mode without too many problems. However, I put together some code to do microstepping, and the current draw of the stepper kept re-setting the propeller every time the stepper moved. To get to the point, would I want to put in a power resistor to limit the stepper motor's current? All the resistors I have on hand ended up at approximately the same temperature as my soldering iron after 10 seconds...
Also, I am clueless as to what bearings I should get. Is there some criteria to look for when purchasing bearings? Finally, what type of coupling would I need to fasten the stepper shafts to the threaded rods?
Thanks.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
MCU Projects·- my new site where I will be posting all projects, code, etc.
Quit buying all those fixed voltage regulators, and·get an Adjustable Power Supply·for your projects!· Includes an LED testing terminal!
SD Card Adapter·- Add extra memory to your next Propeller project with ease!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Lots of propeller based products in stock at affordable prices.
I got the driver parts finally, so I was able to build the circuitry for driving a stepper.
Soon I'm going to be able to work on the actual construction of the machine frame, but that will be after I finish the driver. Would it be better for me to drive my steppers in half step or full step? With half step, I can get an ideal resolution of 1/8000", and with full step, 1/4000". I don't think I'll be able to microstep them, because that seems to require some crazy current modulation.
Which step sequence would be better?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
MCU Projects·- my new site where I will be posting all projects, code, etc.
Quit buying all those fixed voltage regulators, and·get an Adjustable Power Supply·for your projects!· Includes an LED testing terminal!
SD Card Adapter·- Add extra memory to your next Propeller project with ease!
I would recommend neither of those, but instead to get a motor/screw combination that will give you 0.001 inch steps, or even coarser. Your full step suggestion of 4000 per inch will make the unit very slow to operate, and 8000 to be doubly so.
Your suggested high-res configurations are not needed..... they will be "lost in the noise", as there is no way your mechanics can take advantage of those very fine positioning increments.
Cheers,
Peter (pjv)