Sewing Machine Control - motor source

sesquatch

I would like to control a sewing machine by replacing the motor (most likely with a stepper?).· I am looking for a source to purchase a single motor for experimentation with the Stamp as the controller.· Any recommendations?

The existing external singer motor is a universal motor, AC/DC 85-135V, 0.5A and about 1750 rpm.· It is driving the machine flywheel through a 4:1 reduction via a belt drive.· I would like to be able to control speed, count stitches, stop with needle up or down, etc.··
Replacing with a DC motor is not a problem (I do not need to keep·the existing AC 115V supply)

Bruce Bates

sesquatch -

What would be the point of replacing an AC-DC motor with a DC-only motor, unless the only point is to switch to a stepper motor?

Counting stitches is probably best done with an encoder, and determining UP or DOWN position is probably best done with one of the following techniques:

2 - electro-mechanical limit switches - one for UP and one for DOWN

or

2 - photo-electric (PE) detectors - one for UP and one for DOWN

The life of the PE method is probably near infinite, so long as the path is kept free of dust, thread and other hinderances to light. If the truth were to be known, if one used a co-processor, one cold count the stitches very easily using the PE method as well.

I hope for your sake that reverse (backwards) stitching isn't going to be a part of this set-up as that brings in a whole new batch of considerations!

Regards,

Bruce Bates

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

Phil Pilgrim (PhiPi)

This really is an interesting app! From my very limited exposure to sewing machines, it's apparent that there's quite a bit of stiction in the drive mechanism and, when the universal motor finally overcomes it, it has a tendency to race. The typical triac speed control requires quite a bit of finesse to obtain a gentle and predictable start. A DC servo motor or stepper would make it possible for the electronics to do the finessing, leaving the operator the luxury of a true speed control, rather than just a voltage control.

But...assuming AC/DC watts and stepper watts provide comparable horsepower, you'd need a 60W stepper. That'd be a real beast, as steppers go...

-Phil

D Faust

I definitly agree about the precision of a sewing machine.· It is so hard to accurately control it with your foot.· Why would you need an encoder, couldn't you count the number of steps the stepper motor·makes.· You are hooking this to·a microcontroller right.· Whatever you decide to do it seems very practical, and I would love to see it at the end.· Mabey you should put some line followers, and· a servo with an omni wheel·(http://www.omniwheel.com/omniwheel/omniwheel.htm) to turn the fabric and make it really automated.· Just a thought.


▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
D Faust

Bruce Bates

D Faust -

If you don't use an encoder how will you know if the stepper made the right number of steps, trust it?

Regards,

Bruce Bates

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

bobledoux

The textile industry has servo motors that replace regular sewing machine motors for special applications. Check out how that technology works before you reinvent the wheel.

It can take significant force to poke a needle through fabric. Unless the mechanism is moving with some speed before the needle hits the fabric it can come to a dead stop. Even commercial sewing machines, with half horse clutch motors, sometimes need a manual hand to get started. So don't underestimate the stepper power you need. The other option is to stop the needle bar as the needle exits the fabric. The sewing mechanism has lots of mass so use a soft start.

sesquatch

Thanks for the responses - many of them I have considered (optical encoding is already in place for revolustion counts - but since precision in counts isn't extremely important I figured trusiing the stepper would be adequate - provided I account for stalls). I've also ordered the pwc module to drive a triac for the existing motor.

This is PURELY a hobby application (quilting) and it is somewhat important to be able to stop with needle up (or down) so I thought it would be fun to play with either stepper of servo technology without breaking the bank.

I was considering a motor from http://www.usdigital.com/products/ms23c/ which seems like it should have the drive - but was wonderingif anyone on this forum had any favorites (maybe I should try the robotics group).

Thanks

bobledoux

I wonder if a hall sensor mounted on the hand wheel would meet your needs. The hand wheel rotates once with each stitch. The bobbin cases on many lock stitch machines make two rotations for each stitch so be careful if you decide to place your sensor below the pressure foot.

bobledoux

Stepper motor torque drops as step rate increases. I'd try to use a motor with few steps per rotation, like a 15 degree per step unit. This reduces the number of steps needed and increases the stitches per minute. A simple routine would simply look for the stop command by the operator and then step forward until the hall sensor was read. This would provide a consistent needle position at stop.

Of course, by using counters the sewing machine could be commanded to make a sequence of a fixed number of stitches. That leads to all kinds of stitch routines.

D Faust

To account for stalls you could monitor the current becuase don't motors draw more current when they stall. Bruce, that was what I was thinking, just trust the motor. Just wondering, but what would be your user interface, a potentiometer? I also like the hall effect sensor-hand wheel idea. that could work for stall sensing and rotation counting. You are just trying to get better speed control, not overhaul the machine right?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
D Faust

sesquatch

I would like to accomplish several things:
1) Stitch regulation. I already have encoders on relative fabric motion, so speeding up / slowing down the machine relative to the fabric will keep the stitch length uniform (quilting machines do not use traditional fabric feed)
2) Braking. This particular machine (selected because of the "throat" capacity), continues with 6 or 7 stitches when power is removed because of inertia. I need to stop it within 2 or 3 stitches to avoid knotting - wanted to use motor braking but have considered caliper brakes or....?
3) Programmable needle position. One touch operations to lift needle of put needle down for different motions (for example, the Juki TL98 commercial machine includes this).
4) Other nice features such as estimated bobbin remain - if stitch regulated and count of stitches, bobbin remaining can be estimated. # of stitches on needle, stitches since last oiling, etc.

and 5) an interesting project to play with microcontrollers....

Really I was just looking for sources to buy motors - the Web is a bit overwhelming.

Lawson

cool project. First off, the stock belt drive will slip and cause a stepper to loose sync with the machine. So a big stepper (if you can find one!) is a good idea, but the belt drive will have to go. The stock motor with a position sensor and electronic controller would be sufficient to do speed regulation, but braking the sewing mechanism would require modifying the motor.

www.banebots.com has a selection of motors and electronic speed controllers that might interest you. (www.surpluscenter.com is another good target)
www.banebots.com/pc/MP-36XXX-540/MP-36036-540 with the encoder option could work very well direct driving the sewing machine. (i.e. no belt) A cheaper source of DC motors is to go find cooling fan motors from a local auto salvage yard. They should have the power you need at 12v and are usually happy running at 24v if extra power is needed.

Beware, using a speed controller with an "RC" interface will limit how fast you can run the speed regulation servo loop. A motor driver that's just a large H-bridge will be better for this application. Item 1) is likely to require VERY fast speed regulation. Braking, Item 2), comes automatically with a stepper or servo control. Needle position, item 3), should also be trivial with a servo or stepper.

Best of luck with this,
Marty

Post Edited (Lawson) : 6/9/2007 2:33:47 AM GMT

Lawson

Just an additional note.

http://www.usdigital.com/products/ma3/

I found this interesting little magnetic absolute encoder on the web. With the PWM output option it'd be much more BS2 friendly than an incremental optical encoder. As long as the BS2 can read in a pulse three or more times a revolution the BS2 will be able to count revolutions of this encoder. (i.e. for a 1000rpm speed the BS2 will need to check it at least 1000/60*3 = 50 times a second)

marty