Shop OBEX P1 Docs P2 Docs Learn Events
stepper advice needed — Parallax Forums

stepper advice needed

mctriviamctrivia Posts: 3,772
edited 2010-08-28 10:13 in General Discussion
I was thinking of using this stepper in a battery powered robot. But I realized that it would probably drain my AA batteries pretty quick especially keeping it on to keep the device in position. Then I thought if I just use a worm gear on the shaft of the stepper connected to a crown gear I can get a massive gearing ratio so I do not need a stepper with such fine steps and i could shut the stepper off and it would still hold its place.

So now I am wondering does anyone have any recommendations for a small light weight stepper motor I can get pretty cheap, and preferably control using only 5 and 3.3V?

Comments

  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2010-08-27 10:21
    3.5" disk drives have a tiny stepper in them that could be directly attached to a wheel and driven by very little current.

    If they are unipolar, a UNL2308 could work well. If they are bipolar, a Beam h-bridge made of six transistors could do the job.

    I've had two steppers set aside for this for quite sometime, but I've not gotten around to cannibalizing them.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2010-08-27 10:36
    It doesn't take much current to maintain a reasonable holding torque. And stepper RPMs are low anyway compared to a regular DC motor. Adding a worm gear would slow things to a crawl.

    -Phil
  • mctriviamctrivia Posts: 3,772
    edited 2010-08-27 10:46
    I need to do about 2RPM so speed is not the most important factor though using a worm gear would mean mounting it sideways which significantly complicates the task. Not to mention small steppers seem to have a very small shaft which would be difficult to connect the gear to.

    I was starting to worry when I saw this: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=403-1019-ND

    but I just realized they are showing the back of the unit that is not the shaft sticking out.

    For bipolar steppers I have always used A3982 it makes things easy and compact but does require I generate a 3rd voltage rail since its minimum voltage is 8V.

    A unipolar and 4 transistors would make the circuit a fair bit cheaper but unipolar steppers would require me to micro step or use a worm gear to get the needed resolution.

    I think I need to put together a 12V boost converter and see how long a set of AA batteries will run the stepper I have at both 2RPM and stationary.
  • ercoerco Posts: 20,260
    edited 2010-08-27 11:21
    Heck, for only 2 RPM, why not find a small solar powered DC motor that would only pull 10 mA off your 3.3/5V supply and run it direct? No transistors, nuttin' honey. Run it through a small gearbox or worm gear, put an indexing switch on it so you know your position and call it a day.
  • mctriviamctrivia Posts: 3,772
    edited 2010-08-27 11:32
    Step 1) Hook up multimeter to stepper motor before control circuit so battery life can be estimate
    Step 2) Build boost circuit to get actual time

    So I have done Step 1.

    At 12V the holing current is 600mA(7.2W) now this current does include 2 LEDS and an inefficient linear regulator to power the control chip. both things I would fix on the robot.

    Moving at 120RPM current draw is 150mA(1.8W)

    Moving at 30RPM current draw is 300mA(3.6W)

    Moving at 3RPM current draw is still 600mA


    Well that makes the math easy. If I am moving slowly or staying still I am using 7.2W of power. My power supply is 80% efficient so I need 9W coming in that is 3A off 2 fresh batteries and 6A when dead. I know a Duracel Lithium battery can provide 2.5Ah at 2A draw which if it is capable of doing the same at 3 would mean the batteries would last 50min.

    That is not very long and those batteries are 2 expensive to waist on step 2. If I go this route I need to find an affordable rechargeable battery pack that is light weight. As an alternative a Makita 18V lithium battery could power the stepper driver direct and would last 5h. But the batter are 500g(compared to 30 for 2AA batteries) and cost $100. however they are rechargeable and I already have 2.

    Now if I could cut the power off when holding a position then the battery life would go up by 10 times since it will be stationary most of the time.
  • mctriviamctrivia Posts: 3,772
    edited 2010-08-27 11:36
    erco wrote: »
    Heck, for only 2 RPM, why not find a small solar powered DC motor that would only pull 10 mA off your 3.3/5V supply and run it direct? No transistors, nuttin' honey. Run it through a small gearbox or worm gear, put an indexing switch on it so you know your position and call it a day.

    Because even though it is moving slow I do need the position to be accurate and definable. At minimum there would be 400 different positions the motor could stop at.

    Currently I am using a servo motor but I want to switch to a stepper so it can be continuous rotating and have more control over the speed.
  • Spiral_72Spiral_72 Posts: 791
    edited 2010-08-27 12:21
    Printers have some pretty nice and small steppers in them too.

    Could you do something weird like trigger a PWM output to the stepper while it's on hold and still have full power on step?

    You'd need one Stamp pin to enable power saving mode for when the stepper is holding.... ummmm, a 555 for your PWM signal..... unipolar would be simpler..... you'd need some logic AND gates, when the stepper line is high from the controller AND your power save pin is high, feed it PWM through a 2n2222 biased for 100ma.

    Actually that sounds fantastic, but I think you'd need a transistor on each of the four control lines too because a logic device won't handle 600mA.

    I'd have to draw it I suppose. Did that make any sense at all? The point of this was, when the stepper is on hold, two coils stay energized. So if you reroute those two and feed them PWM (100mA during hold), the holding torque would be somewhat less, but the power requirement would be, I dunno, half? maybe less?
  • mctriviamctrivia Posts: 3,772
    edited 2010-08-27 14:23
    I think the worm gear is definitely the best way to go for the following reasons:

    1) mechanically locks the output shaft in position so no holding current required.
    2) gives much higher position accuracy a 7.5deg stepper and a 20 to 1 gearing ratio gives 0.1875deg accuracy with half stepping(noise in the system would be greater then this)
    3) forces speed at a more reasonable rate. a 120RPM stepper with a 20 to 1 gearing ratio forces the max speed to only 6 RPM. Any faster and my robot would likely break.
    4) high gear ratio means in accurate unipolar steppers can be used which can operate at 5V meaning no 3rd voltage rail
    5) high gear ratio means smaller lighter hopefully less power hungry steper can be used.


    The only cons is I need to find a way to buy or make a 20 to 1 worm gear system with an 8mm output shaft and a 2mm input shaft. gear ratio calculations is simple 1 rotation of input is 1 tooth of output so i need 20 teeth on output gear.
  • mctriviamctrivia Posts: 3,772
    edited 2010-08-27 15:05
    hm for simplicity sake I think I may have come up with a better idea:

    http://www.sparkfun.com/commerce/product_info.php?products_id=9402 - to control the stepper in 1/8th step increments(allowing accurate slow running)
    http://www.sparkfun.com/commerce/product_info.php?products_id=9238 - slightly smaller stepper motor but still accurate
    http://www.sparkfun.com/commerce/product_info.php?products_id=10006 - to allow mounting head directly to stepper.

    If I get a thrust washer between the head and the stepper that is not supper low friction it will resist the head moving under normal conditions when the power is off and reduce strain on the stepper.

    It is much simpler and by playing with the friction level should hopefully work.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2010-08-28 07:13
    Worm gear?
    Many times I wish I were born into a machinist's family as the electronics are quite easy to build, but having access to all the mechanical fabrication is difficult to find. A real robotic enthusiast would best have use of a machinist lathe, a milling machine, a bit of welding equipment, and maybe some sheet metal fabrication equipment. (And a forklift for whenever one wants to relocate their shop.)

    I spent a lot of time looking at gear catalogs on line and small quantities of gears are expensive, very expensive. That has constantly led me to feel frustrated about steppers because they usually come in a NEMA standard that is rather chunky and never with a gear box.

    I love the precise accuracy of steppers, but I detest the fabrication realities - all seem expensive.

    There are alternative of toothed belt drives with slotted pulleys that work. But turning a 90 degree corner with a stepper is quite unusual. You may even be doing a first.
  • lardomlardom Posts: 1,659
    edited 2010-08-28 07:50
    Nice. Everyone in this thread to this point has contributed to the over current-problem I had with my stepper. My project is now complete, I have a smile on my face and this is a great time to say thank you! mctrivia, I hope you post your solution. My education continues...
  • mctriviamctrivia Posts: 3,772
    edited 2010-08-28 08:03
    you know i can't find any reference to anyone else using a worm on a stepper either. wounder if there is a reason for that. looks like it should not be to difficult to design and build a worm and worm gear.

    The worm is essentially 1 tooth that has been spun on a helix equal to the tooth pitch.

    The worm gear is the same as a sper gear except the teeth are on an angle to match the worm.
  • ercoerco Posts: 20,260
    edited 2010-08-28 09:24
    To explain my point further, imagine a small DC motor driving a worm gear engaged to a 400-tooth main gear. One motor/worm revolution (as tracked by an indexing switch) moves the main gear 1/400 of a revolution, to meet your needs. Quite simple and requires at most 3 I/Os: one for the index switch and one or two for an H-bridge, if you need forward/reverse.

    Of course you can use a 100-tooth gear and 4 PPR on your motor index switch if need be...
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2010-08-28 09:44
    When the stepper is still, you do not need to power it continuously. Just pulse it with a low duty-cytle PWM. That will reduce the current consumption considerably.

    -Phil
  • mctriviamctrivia Posts: 3,772
    edited 2010-08-28 10:13
    erco yes you are right that could be a viable option.

    phil when i get the parts I will try PWM the enable pin to see if I can reduce the holding current significantly but still get it to hold its position when external loads are applied. If i can drop the holding current by 1/10th that would greatly increase the battery life and I would really like to use AA batteries.
Sign In or Register to comment.