Build thread: submicro brushless motors

[mrk_ls]

Hi'  Where can i find the IC chip for the motor and  do i real need one for motor to spin?

[shagrug]

      I buy the A1442 chips from Digikey, but other online vendors (like Mouser) sell them too.
http://search.digikey.com/us/en/products/A1442EEWLT-P/620-1210-1-ND/1642789

      The motors can also run on AC current (signal generator), but I have no personal experience with this yet.

     

[dick]

Where do you get the magnets with holes drilled in them?
Dick

[shagrug]

Hello Dick-

      I buy most of my magnets from George the "supermagnetman" at his website here:

http://www.supermagnetman.net/

      Make sure you only use magnets that are magnetized through the diameter- the others will not work.

[Amuse1988]

Hi shagrug,

I am very interested in the motors you build and want to build one myself. I was wondering how the coil is connected.  Is the coil one strand of wire wrapped around the magnet multiple times and the each end connected to the + and - of the controller or battery? and also do you need the controller to power the motor?

[shagrug]

Hello Amuse-

      You are right, the coil is made from one strand of magnet wire with each end connected to the output pads on the controller chip (marked "vout1 and vout2" on the A1442 datasheet).

      The controller chip must be used- it turns the power to the coil on-off at the proper time, using the hall sensor inside.

[Amuse1988]

Shagrug,

Awesome! Thank you i have been trying to figure this out for quit some time now. How to you power the chip? I see that you use a DC supply. But how do you connect it to your RC receiver?

[shagrug]

      On a receiver you would connect the chip to the "brushed" motor output connections, making sure to connect positive and negative correctly. It will not damage the chip to hook it up backwards (reverse polarity) but it won't run if the polarity is incorrect. 

      You hook it up exactly like you would a normal brushed motor, only two wires are involved. (the pos-neg) Remember it is just like most small motors (pagers for example) but the chip is doing the work of the brushes. It will start running a motor at about 1.0 volts and maximum input is 6 volts. So any DC power source over 1 volt will work, like a 1.5 volt AA battery for example. 

      If you don't have a AC-DC variable power supply to power the chip you can use a rx motor output. No problem with that, unless you have a lot of testing to do. Running it direct off a 1.5v battery works fine too, but directly from a 3.7v lipo can be way too much speed.

      Below is a vid showing how you can use a cheap $4.29 wire wound rheostat as a speed control on a small pager motor. It's not perfect- it lets some voltage through at the lowest setting (1.2v or so) and restricts a bit a full power/speed. (by 0.1v or so, the lipo had a 4.12v full charge) But with a on-off switch added in you can make a simple speed controller/variable power supply for testing motors, without using a tx-rx.

      The rheostat is a Radio Shack item listed here and found in most stores:
http://www.radioshack.com/product/index.jsp?productId=2062299
It can run bigger motors, I used one on a small indoor electric control line plane a couple of years ago with no problems. (N20 motor) I remember the hand controllers for model slot car racing are also wire wound type rheostats, so they would probably work as well.

using a rheostat for a speed controller on a 6mm pager motor

[Amuse1988]

Shagrug,

Awesome! thanks ill give this a try!

[shagrug]

      Here's a motor I finished a couple of months ago that taught me a couple of new things. Not meant to be a model motor (or submicro), but as a desk fan and display motor. Also I'm interested in seeing just how long it can run before the bearings or the chip fails.

      The magnet is the biggest that George (the supermagnetman) stocks suited for making a motor (magnetized through the diameter). The dimensions are 3/4" x 1/2" x 1/4" or about 19 x 13 x 6.5 mm for the metric fellows, grade type N50 and weight of 24.5 grams. I can't describe how strong this magnet is easily- it is just amazing. It will make another magnet dance 11-12" away, and pull in bits of metal from 6" or more. The rated pull is over 26 pounds, and when something gets stuck to it you say "wow". I had the thought that with the right coil it would need very little current to run, having all that gauss strength and mass for flywheel effect. The magnet is listed here:
http://www.supermagnetman.net/product_info.php?cPath=41&products_id=398

      I thought of making the frame from delrin plates and carbon fiber rods press fitted together- that would make it easy to rebuild if needed. I ordered a couple of ball bearings from McMasterCarr (3/8" x1/4") thinking they would be just right for the size/weight of the magnet. And in the photos below you will see me fitting up a 1/8" cf shaft with the ball bearings, as I start to assemble the frame. No problem, I thought...

      But there was a problem- a incredible amount of drag and friction. And it didn't take too long to figure out why. The strong magnet was pulling the steel  balls in their cages inward towards it, binding them up from rolling. The steel cages for the balls mabe added drag too. Most unexpected, and I tried thinking of different ways around it. First was to space the bearings farther away from the magnet- but the problem was still there even with a 1" gap. At a 2" gap not so bad- but I didn't want the motor to be 5" or more long. So I tried smaller ball bearings next- using a 2mm diameter shaft instead of the 3mm (1/8"). Smaller balls did help some, but not near enough. Drag was still extreme- just like you had a brake on the shaft. I had the thought that the fluxing of the coil as the motor ran might weaken the pull on the balls, but decided not to try it. I wondered if the balls might also "rattle" a bit with the coil pulses, but that was just a thought. I also thought of non-magnetic stainless steel  bearings, but couldn't find any the right size. Since I didn't see any good way apound the problem I decided best to switch to virgin teflon bearings. I had been wanting to test more motors with virgin teflon anyway, so you see the change. The smaller 2mm shaft is made from nonmagnetic titanium by the way, a leftover upgrade for a Piccolo heli made years ago. The black plastic 5 7/8" diameter prop I'm using also came off a Piccolo, and puts out just the right amount of wind.

      I didn't have any problems with bearings after changing to the virgin teflon, but the coils took three tries to get right. I used a small lathe to wind them faster since so much wire was involved. I started out with too fine a wire gauge (.05mm), and the coils endied up at over 3,000 ohms each. (running too weakly) Then too thick, at 8 ohms...next the 22 ohms each that were just right. (from #34 wire) I wanted the motor to run silent and just ticking over, and I was happy in the end. Speed is right at 1,000 rpm with a full charged lipo (4.2v) and that gives a nice quiet 8' breeze. I've got a nice 6 volt x 150ma solar panel I'll eventually use to power it. But for now I've got a parkzone lipo socket soldered on so I can run it off some of my flight batteries. I've got a bunch that need cycling  after sitting too much last year.

      I'm going to discuss the teflon bearings some more in this thread- they have some drawbacks too. But the drag problem I found on this one I found quite interesting. It's normal to find steel ball bearings in most the bigger electric motors nowdays... but due to having such a big hunk of magnet in this motor design it creates problems. So you might want to consider my experience before trying ball bearings- they sure didn't work so well on this one

desk fan motor test run

[shagrug]

      I have been meaning to mention some new tools (and old). A fellow in another thread has asked for info on pin vises.

      First is a "small-hole drilling adapter" (part number 30505A5, price $7.05) as listed in McMasterCarr's online catalog. This is nice and lathe true with jaws that close up perfect, and can double as a small pin vise. This is actually made a lot better than two others I have that cost three times as much, so I bought a couple more after seeing the first. The 1/8" shank fits a mototool collet perfect, something else I like.

      Next I wanted to mention pin vises- I have been getting them from McmasterCarr too. They are the common "General" brand (# 8455A16), which is not too expensive and reasonable quality. I have two in the photo because they are slightly different- one is made better than the other. If you look at the handle end you can see "General-made in the USA" stamped on the bottom pin vise. The pin vise handle in the middle just says "GENERAL" stamped on it, and it is a import not so well made. (corser knurling, not true in a lathe and jaws that don't close as well) I have bought a total of 4 of these vises over the last couple years, and the first three were all the USA version. But the last one from McMasterCarr I received was a import, so mabe the better ones are gone now. I put the drilling adapter in the photo to compare, adding a handle to it isn't really needed.

      Another good item from McMasterCarr are tweezers... one of my two Didel brass ones broke at the tip and needed to be replaced, so I tried their high grade 5669A3 and 5669A7 nonmagnetic stainless versions, reasonably priced at about $6 each. They are in the middle of the photo, and I'm very happy with them They are of course not as nice (or sharp) as the $34 Swiss made one at the top (from otto Frei) but still very good- and far better than what you will find at our local hardware or hobby shops. I'd say they are better than the Didel brass ones, at about half their price.

      On the last photo, another interesting clamp at McmasterCarr is the "Thompson English piiers" part #5097A54 and sort of like a small hemostat clamp- but the jaws are different. The smooth jaws clamp flat and have quite a grip, holding the micro garolite and teflon sheet bits well. I'm using it a lot, and will probably get another as a spare. Also in the photo is a micro oiler bottle available from Peck-Polymers (in the indoor specialties). I like these and have been using them for many years, as the #25 gauge tube will fit in tiny spots and deliver micro amounts without flooding. It's easier to use than a hypodermic needle, but with the same size tip. For small motors like we are making here it is just right in size. i usually use 3-in-1 oil in mine.

[shagrug]

More tool mentions-

      Since I mentioned the rheostat speed control/power suppy above I'll mention my desktop power suppy I use almost constantly. You have seen it in almost all of my vids running the test motors, and I use it for countless other projects. (saves money on buying batteries for sure, on my Bicha charger) Not a expensive model, I think it cost $55 or so. If you are heavy into making motors or other micro RC electronics it would be useful to have one.  I have seen the prices vary a lot from vendors, and there are many models to choose from. (so shop around) Mine came from here:
http://www.kitsusa.net/phpstore/html/Test-Equipment-Assembled-86-1.html

      Back during the summer I picked up the SMD hot air rework station from a ebay vendor, thinking it would be handy since I'm doing so much soldering work. It is also for other projects, being a very handy controllable heat source with the hot air. I have already been using it to try and extrude virgin teflon tube from block for example.  And for getting the bigger ic chips off-on it is the way to go- I have always had trouble getting a chip off a board if it had more than 3 or 4 legs. Now I can do 40-50 legs no problem. (Romeodz- it is just what we needed to do that cellphone jack a while back) Along with the hot air gun it has a soldering iron built in, which I also like. I don't have to change tip sizes back and forth so much on my Velleman iron, so it saves me time. The stock needle tips that come with the station are small enough to do the A1442 chip, and mine had extra heating elements included. Cost of the unit was under $65 with shipping, as I remember.  The price is low since it is a older model, the diaphragm type air pump in it has been replaced with newer fan in handle types. So again I'd say shop around if you are interested in getting one, prices vary a lot.

      Another thing picked up this winter is a really cheap usb type oscilliscope, off a Chinese ebay  vendor for about $30 with shipping. (I won't list the vendor name since there are so many similar) I'm hoping it will work for what I need, which is testing the rx motor outputs. (I will explain in detail why later on) The box on the left has all the scope parts- basically it's just a small case with two bnc ports on the front and usb cable out the back. Works well so far, but I am really just learning how to use it. It is limited to low voltage circuits (under 5v) so I may end up getting a better desktop unit later on. (I've been looking) But for now it should do what I need for now, and I'll post my tests with it. Since it does not come with a probe I had to shop around for a cheap one- the 60 mhz one on the right of the scope came from here:
http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_2130781_-1

[shagrug]

      The photos and vids below were testing something I'd been wondering about for a while... It has to do with getting more current into the coil using two A1442's wired in parallel to the same coil. (different from the two chip-two coil setup I tried a long while back- see it on page 5) I was not sure if it would even run wired up this way, and since it did I tried the chips in two different spots timing-wise. I was thinking by staggering them it might give two separate pulses to the coil. It's interesting how much the amp draw went up and speed down with that arrangement.

      But overall performance went down compared to just using a single chip. I can really only see a improvement in power by using two separate coils as I did  before.  Good to know, I've been wondering about this ever since making that dual coil motor, so I'm happy now.  You might notice that I used a old 4x5mm test motor I had for this (no.4) rather than wind a new one. I had one more thought to try on it- mabe adding diodes. But I'd like to ask someone about it before I try it.

      So far my conclusion from this is that I'll only get more amps into the coil by using a H-bridge setup or prehaps a different chip than the A1442. I've been studing other datasheets of late and there are a couple of other chips I'd like to try, if I can get them. Some of these  new chips may work better with a H-bridge too, something else I'm looking into.

      First the test run with one chip-

4x5mm test motor no.4 with 1 chip for comparison (to 2 chips wired in parallel)

      Then with two chips placed inline (triggered together at the same time I hope)

4x5mm test motor no.4 with 2 chips in line placement (wired parallel)

      And again with the two chips now staggered (one triggered before the other I think)

4x5mm test motor no.4 with 2 chips staggered placement (wired parallel)

       

[Keith55555]

Hello. Joining in a little late on the topic. Much more of the thread to read.
I'm designing a similar motor for a different purpose. Goal is small height and width in that case (not low weight; length is negotiable). For my purpose, I need to be able to reverse direction. This concept was mentioned early on in this thread. Mentioned was to use two controller I.C's which doesn't sit too well with me, if I understand what is suggested since as I understand this entails tying two 'outputs' to each other, most often a 'no-no'.
I'd even consider the suggestion of using two motors on the same shaft but one of those motors are going to generate an output to an output once more.
So my real question is, what exactly is causing the motors inclination to move in a specific direction? If all was balanced, it would seem the direction would be a 50-50 call on start up.

I built this special motor using a pancake/disc vibration motor's innards. The only thing not milled out of the nickel for support and containment was the shaft in the center which wouldn't have been as slick if I tried to make it myself. First pic is the nickel partially milled. The second is partially populated and the movie shows the nickel spinning. The movie stinks because the nickel is spinning so fast the cam couldn't keep up with frame rate.
Note:Looks like movie is too big for the server. I'll cut it down and post later

Also wanted to mention, that i read the Allegro part is being retired and Melexis has what looks like an exact replacement.
Also not. Jin-long now has 3mm diameter motors. From the drawing, it looks like some support of the motor creates more girth but the body reads 3mm. They're 5 bucks a pop for hobbyists.-Keith

[Keith55555]

Okay, you can see the movie here. That is a real nickel milled.
http://www.lightlink.com/keith/nickelspin.wmv