New to stepper motors..

[Lord Banshee]

OK well i want to learn how to use motors so i decided to start with one i had around my house collecting dust, floppy drive .

Ok well i looked up the model number and i found these two spec sheets:
http://www.vitech.com.tw/type_stepping_...stepping_motor_pdf/M20SP-GW15_SPEC.pdf
http://www.vitech.com.tw/type_stepping_.../stepping_motor_pdf/M20SP-GW15_DIM.pdf

it is a bipolar 18*/step motor DC:5V.

Now my questions, as the second spec sheet shows how the pins should be activated to make the device move, looks easy enough.

1) Would you use a parallel port dataline D3-D0, pin 5-2 to control the stepper? Would the stepper ask for too much current out of this port?

2) I also wanted to control this with a FPGA which is 3.3V. So i know it requires 5V so i should make a control circuits with transistors to operate the device at 5V and not worry about the motor asking for too much current and kill my FPGA right?

My plan is to make some software that connects to my FPGA though the serial port to controls the stepper motor.

Ideas, Thought, bombs on my plans?

Thanks,
Chris

 

[f95toli]

You need a driver chip (or build something yourself with a few transistors).
Stepper motors are really nasty loads and they use a lot of currents; ususally a few hundred mA per phase.
Hence, there is no way the parallell port could cope with a load like this.

 

[Lord Banshee]

Thanks thats what i thought, spec says 333mA per phase.

 

[harrkev]

You can make due with some sort of H-bridge circuit (build your own or an IC). Somebody also makes dedicated stepper controller chips.

 

[Lord Banshee]

ok i got it working by just using the floppy driver controller, did have enough of spare parts to build the H-Bridge

Another question if i wanted to create a device which needed to be precise and move in very small increments for example.

If i have a last pointer hooked to a motor i would like to rotate it the smallest angle possible and make sure when the motor is 360* around it is exactly back at the same angle would using a stepper motor for this be best? i found 0.9 degree steppers but none smaller than this.

Or would using a DC motor, never used one of these?

Also i notice there were feed-back encoders on some of these motors what exactly do they have the ability to do? IF an encoder can i set it to a certain angle from power up, as if i just plugged in my device would i have the ability to know it is correctly at 30* so i know i need to move it to 180* the default starting position?

As you can see i am new to motors, i've used a small servo for a very in-expensive object avoidance robot but thos things were not precise at all as far as i can tell.

Thanks

 

[Ben]

I don't know if it's the same for the micro steppers and motors like they use in a floppy, but in the larger industrial units, the encoders can be either incremental or absolute.

An absolute encoder will give it's position even if it's moved without power. Typically absolute encoders in the industrial field are limited to 1 revolution or so.

An incremental encoder will loose it's position if it's moved without power, but some of the better units have a "z" position indicator that is like a zero position reset. The down side is that you must rotate the encoder after power up until the z channel triggers a reset before you can know your position.

Typically incremental encoders in the industrial field have no limit on rotation and often times they are used for precise RPM detection.

 

[Lord Banshee]

Ben,

Thanks for the explanation on some of the encoder types. I am not messing with the floppy motor anymore as i just wanted to use to make sure it was easy to control and i could control it correctly which seems to be depended on the driver/controller for the motor. Seeing that you stalking about industrial field motor i assume you've worked in the field with motors so i have a question about part searching to meet specs.

I am trying to design a device which rotates say a laser pointer along the z/y plane with one motor and use another motor to rotate the previous device in the x/y plane so i can map a spherical area. The thing is i am looking for < .9 degree motors that max can move 3-5 lbs. I found some motors which were small that did 0.1 degree step with the help of gears but was included in one package. So do you have any suggestions on looking to find geared motors what could support that weight of 3-5lb and have such small steps. And i like the idea of the zero position reset as that is exactly what i need it to do move back to a default location.

Thanks,
Chris

 

[Ben]

Your talking about smaller components than I typically deal with so I'm not sure about specifics.

From what I've seen on ebay and robotics sites, your probably going to end up using a geared reducer to get smaller than 1 deg increments, like you said. The problem with a gearbox is that it has "play". Meaning, if you reverse direction, it may take a bit of rotation of the motor before the output shaft on the gearbox starts moving the opposite direction. It sounds like your trying to get very precise with your positioning so this may be a problem.

With the larger stuff, we would use a servo motor with an encoder on the output shaft with extremely fine increments. I've seen them as high as 5000 pulses per revolution. Maybe a 3600 PPR would do for your application? That'd be 1/10th of a degree. Since we are measuring the movement of the output shaft rather than the motor, the gearbox backlash isn't as much of a problem.

 

[Lord Banshee]

Thanks Ben

i found this geared motor: http://www.portescap.com/product.cfm?Product_ID=156 with a step angle of 0.125 degrees. I am just not sure if it will handle the weight, i have to figure out how to do the calculation, been some years since the last time i worked with "torque".

I was thinking about the encoder and maybe i can just get some sensors on the devices it self and if the sensors are not set up right than rotate till it is aligned, but then again i do not think i can get sensors to be as accurate as 0.125 degrees.

I have a lot of work ahead of me

Thanks again for the advice

 

[Ben]

Cool site you found. I browsed for awhile.

You know, for this small stuff, I really don't know much compared to others than are in the field. You might try this question over at Ars Technica and see if you can snag a robotics guy to help.

 

[jagec]

If you keep the stepper motor powered up, and don't try to drive it too quickly, you can probably forgo the position sensor, and just use dead reckoning. Calibrate it on startup, and let it run. Of course if you have the budget, it's more robust to measure the position directly.

As for the hysteresis problem on step-down gears, that can be resolved by either always approaching an angle from the same direction (that is, you either turn clockwise to get to the angle, or you turn past the angle counter-clockwise and then go clockwise again to reach it), or you could try putting a spring in the system to keep a constant tension on the gears. Of course, if you do have the position sensor, you don't need to worry about it since you will be measuring directly.

 

[Lord Banshee]

About the positioning,

I was thinking about limit the rotation by only 180 deg so i have a little block on the stationary dock that touches another block on the rotating disk. The rotating block will have a contact which is ground. The stationary block with have a contact that is 5V using a pull up resistor. So when the two contacts are not touching the MCU will do 5V, when they do tough it will see GND. So on a reset the the motor will move in one direction till this sensor block is GND. At this point it knows exactly where it is so it will move it so many steps in the other direction to a default location.

It seems like a simple fix that might work, but other ideas are always helpful.

I do have a question about find the right motor for the job.

One motor will need to rotate a 1" dia shaft holding a 1lb object.
Second motor will need to rotate all of weight of the first motor plus a structure, thinking fiberglass, holding everything in place. So maybe 2lb-2.5lb.

So the motor listed here: http://www.portescap.com/product.cfm?Product_ID=173 has a holding torque of 84.0 mNm converting 1 ft.lb / 1.356 Nm x 0.084 Nm = 0.062 ft.lb. From this document: http://www.portescap.com/Repos...ack_Geare_01032007.pdf i see the shaft is 0.1875".

So for movement of 1lb on a 0.1875 shaft:
Torque = inertia x angular acceleration
inertia = m x r^2 = 1 lb * 0.1875"^2 = 0.352 lb.in^2
speed = 0.1 deg / 0.5 sec = 0.2 deg/sec

ang acc = not sure how to do this so i looked in my physic 1 book and try to fudge something
speed/time = acc = 0.2/0.25 = 0.8 deg/sec^2, i choose 0.25 as t because angular displacement would be 0.5 x ang_acc x time^2, so i want to travel 0.1 deg per 0.5 sec therefore 0.1=0.5 x ang_acc x 0.5^2, solve, ang_acc = .8 deg/sec^2.

Now back to torque:
Torque = inertia x ang_acc = 0.352 lb.in^2 x 0.8 deg/sec^2 = .281 and some werid units i assume makes in.lb lol ? So i am lost here, and if this is 0.281 in.lb than that motor will not even be close to do what i need it to do correct?

IF this is all correct need to go searching again for that 1 in a million motor i am looking for lol.

thanks,
Chris

 

[Nathelion]

Just a tip for you there... keeping everything in metric from start to end usually makes things easier

 

[jagec]

Another tip is that for most hobbyists, it ends up being simpler to design the system around the motor, and not vice versa, since finding a cheap and readily available motor is usually a greater concern for them than, say, NASA. Just get a motor that's overspec'd for what you need, and fit the system to match. Remember that you can always gear down or up to get the speed/torque you need, as long as the motor itself isn't being overstressed.

 

[Ben]

Both good suggestions.

I haven't done the math, but do you really need a 1" shaft?

 

[Lord Banshee]

Ben,

The 1" shaft isn't the motor's shaft, sorry for the confusion. For example this is the an example of what one of the motor will be and do.

||--[1lb]--||{motor}

{motor} = the stepper motor
[1lb] = the device that weighs a lb that need to be rotated
-- = a mount in the form of shaft that connect the 1 lb device and the motor's shaft
|| = vertical mount with holes for the -- mount to run though and rotate in.

Nathelion,

Yes good tip, that was a pain in the butt

Jagec,

I would design a system around a motor, but i do have specs i have to meet. 1) small steps, 2) can turn what i need it to turn in weight. So from the (1) spec i found that above motor until i did the calculations and found it did not meet (2).
You stated i can gear down or up, do you have any links of the hardware to do this. I was looking at a much cheaper motor before those small portescap motors but it's step angle was 0.9 deg. So if i can gear it to at least half that would be great. Truthfully i might just get the 0.9 deg model and maybe later on get something much more precise or find gears set to do what i really want to do.


So i guess my real question is did i do the math right?