PDA

View Full Version : Electric motor modification



Evan
07-04-2008, 09:58 AM
In the electric bike I have built I am using an old starter motor from a 15 hp Briggs lawn tractor engine. It has worked out surprisingly well after various mods to improve the cooling. I drilled holes in one end of the case and made a new end plate plus stub shaft for the brush end of things.

Yesterday I was fiddling with the motor to change the mounting slightly and had to take off the end plate to machine it. I was thinking about how much more efficient it would be with a couple of NIB super magnet arcs instead of the standard ceramic magnets. Then I noticed several items. There was a lot of extra space between the ends of the ceramic magnets and the brush assembly. There was quite a bit of room on the commutator for the brushes to run closer to the windings, about 1/4" if the commutator was remachined to remove a slight ridge. Also, since I had removed the closed bearing on the commutator end and replaced it with a ball bearing and stub shaft I could easily move the armature 1/4" toward that end.

Ding. A small light came on. I always keep some NIB magnets on hand for experiments and had some just the right size for what I wanted to try.

This is what I did. I epoxied 4 pairs of magnets aligned with the poles of the ceramic magnets and at a 90 degree spacing to agree with the 4 pole winding of the motor. Each pair of magnets is oriented with the pair having the same polarity out. Each adjacent pair is opposite polarity to agree with the polarity of the ceramic arcs. I diamond turned the commutator to allow the armature to be shifted as much as possible to that end.

http://vts.bc.ca/pics4/motormod.jpg

I have been testing the bike and keeping track of current draw vs speed and hill climbing ability. The current is displayed by a regular digital meter for now and is measured as the voltage drop on a set length of cable. In this case I see about a 20 millivolt drop on 1 foot of cable at 30 amps as calibrated by a lab ampmeter.

When I reinstalled the armature it was immediately obvious that the new magnets had a significant effect as the armature was pulled strongly to the new position.

In testing after this mod the motor seems to be producing at least 20 to 30% more torque and heating much less since I don't need to run it as much. That also translates to longer battery life.

Now I have it in mind to buy a handful of NIB magnets from one of the advertisers here that will be the right size to completely replace the current ceramic magnets. By using some 2" x 1/2" x 1/4 inch NIBs I can approximate the ceramic magnet size and pole configuration pretty closely with about $20 worth of magnets.

http://vts.bc.ca/pics4/nib2.jpg

I am curious, has anybody here tried this?

Swarf&Sparks
07-04-2008, 10:00 AM
If you're going to that extent, wouldn't it be a candidate for brushless operation?
Yeah, complicates the driver, but.......?

Evan
07-04-2008, 10:03 AM
Brushless motors in the kilowatt range are expensive, certainly more than $20.

lazlo
07-04-2008, 10:10 AM
Brushless motors in the kilowatt range are expensive, certainly more than $20.

Ebay :)

Seriously, you can find brushless motors pretty cheap on Ebay, but like Lin says, the controllers are a PITA. If you see a mated pair of controller + brushless motor, it goes for a pretty substantial premium.

I just paid $250 for the 1.5 KWatt AC Brushless motor that I've installed on my Bridgeport head, but that's an industrial-grade servo with the encoder and cables, and the encoder cables (double shielded with Milspec DIN connectors) are expensive as Hell.

You can probably pick up a 1 - 2 KW brushless motor (not servo) with the controller for around $100.

Evan
07-04-2008, 10:14 AM
I don't do ebay.

Swarf&Sparks
07-04-2008, 10:16 AM
Err, I meant, given the mods you're already up to, couldn't you homebrew a brushless?

A.K. Boomer
07-04-2008, 10:19 AM
Sounds like a win win Ev, But iv got to admit im with the other guys about finding something that would perfectly suit your needs and be ultra efficient --- im also wondering what effects it might have if you actually moved your NIB magnets in closer to the stator -- like the original motor magnets, it should make even more of a diff. as the fields will be amplified no?

Rustybolt
07-04-2008, 10:25 AM
Just out of curiosity, how far do you get from a single charge?

Swarf&Sparks
07-04-2008, 10:25 AM
Hmm, thoughts here....
If you end to end that motor with a small universal motor and strap both commutators.......
remove the brushes.... from both

Or have I mis-understood the concept?
I know the driver circuitry is not trivial.

lazlo
07-04-2008, 10:26 AM
Err, I meant, given the mods you're already up to, couldn't you homebrew a brushless?

That reminds me Lin: the model airplane guys use brushless motors a lot because they have, by far, the highest power density, and you can wind the motor so that it has immense torque at low RPM.

There was an article in Nuts & Volts about these guys who make home-brew brushless motors, from 20W up to 2 KW:

http://www.aerodesign.de/peter/2001/LRK350/Bilder/LRK350-350-20-Stator_Magnetenglocke.jpg

http://www.aerodesign.de/peter/2001/LRK350/Bilder_eng.html

They have a Yahoo group -- neat bunch of guys:

http://groups.yahoo.com/group/lrk-torquemax/?yguid=92583912

Evan
07-04-2008, 10:28 AM
I have no idea what the endurance is yet. I'm still sorting out the mechanical bugs (pretty much sorted now) and testing hill climbing ability. I decided to modify the drive system by making it a semi automatic 2 speed transmission. I am really pleased with that and it shift gears by the press of a button. I'll be posting a complete update when I have completed some more road trials.


Err, I meant, given the mods you're already up to, couldn't you homebrew a brushless?

Yep.

Swarf&Sparks
07-04-2008, 10:31 AM
Yup, that's the sorta deal I meant Rob.
Knew I'd seen it somewhere.

Swarf&Sparks
07-04-2008, 10:34 AM
You using your "2 hub" idea Evan, or some sorta solenoid derailleur?

Evan
07-04-2008, 10:35 AM
That looks interesting, thanks Robert.

Lin, I'll give complete details in the next update. It really needs pictures to explain how it works.

A.K. Boomer
07-04-2008, 10:44 AM
My little RC plane (peter-out) never would have gotten off the ground without its brushless, Now that you guys are talking about them can I ask if I have the right concept on how they work, it has three wires and its DC, does it use one of the wires as a feedback sensor? it has to have a motor controller so does this motor controller control the signal timing instead of how the brushes automatically do? and does it do it with this feedback wire? Thanks.

lazlo
07-04-2008, 10:45 AM
It funny Evan, they're always on the look-out for a rotor they can use for their home-brew motors. One of the popular donors is a $6 Sunbeam mixer from Walmart, that apparently has a really nice rotor :)

There's also a nice, cheap brushless motor chip from Allegro (IIRC) that does the hard part of the control -- you just have to add your own FETs. There are a bunch of home-brew controllers described there too.

Swarf&Sparks
07-04-2008, 10:45 AM
Look forward to it Evan.
I'm still in the air, I could strap on the hub motor/wheel at 36V, or go your route with the 12V motors I have.

Apart from that the hub motor/wheel is much smaller than the front wheel it will replace, so require much f@ckin about with brakes and trail angle.

Weston Bye
07-04-2008, 11:37 AM
Evan,
Although the NIB magnets are commonly in use in motor applications, bear in mind that they loose (recoverable) magnetism as temperature increases. You are doing well to pay attention to cooling.

Samarium Cobalt magnets have somewhat less magnetic strength, but perform better at high temperatures - indeed there comes a temperature point where the SmCo performs better then a similar sized Neo. However, the SmCo magnets are much more costly.

Evan
07-04-2008, 01:39 PM
Cooling will be a major issue no matter what sort of motor I use. This is hill country and in summer it gets pretty warm too. The last week has been in the 90's F every day, a good way to stress test the motor cooling. So far it has become very warm but not overheated even after two consecutive trips up our hill which is approx 220 foot altitude gain in half a mile. In actual use it will have many chances to cool on the down hills.

The main thing I have to do today is run into town and buy a bike helmet. It's the law. I'm not wearing my motorcycle helmet, too hot.

Swarf&Sparks
07-04-2008, 01:43 PM
Yeah, there are pushbike helmet laws here too.
The bast@rds can arrest me if they've nothing better to do.
I won't wear one.

Errol Groff
07-04-2008, 02:04 PM
Have you seen the web pages of this fellows electric bike conversion project? He is using a motor usually used in R/C helicopters.

http://www.wisil.recumbents.com/wisil/shumaker/default.htm

Errol Groff

Evan
07-04-2008, 02:21 PM
I saw it. His motor, controller and batteries ran over $1000.00. I have about $200 in this project and half of that is the batteries.

Swarf&Sparks
07-04-2008, 02:36 PM
Hell!
Apart from initial cost of the bike (a few years ago) I've got nothin in this and hope to keep it that way!
:D
(OK, a few bucks in MOSfets)

ScottM
07-04-2008, 02:44 PM
Evan,

the cheap version.
http://www.goldmine-elec-products.com/prodinfo.asp?number=G16431
http://www.goldmine-elec-products.com/prodinfo.asp?number=G16432


and the not so cheap - these seem a bit too curved.

http://www.rare-earth-magnets.com/SearchResult-CategoryID-29.html

I remember seeing some large curved magnets but I'm not able to find them now - sorry looked everyplace I could think of.


- Scott

Evan
07-04-2008, 03:00 PM
The cheap ones I already have, they came with the motor. If I could find motor arcs the right size in a NIB I would be interested but I havent found any. They need to be 72mm OD. They have to be just the right size or they won't fit. Otherwise I can piece the field magnets together from long individual NIBS. That's easy to do and probably less costly too.

Swarf&Sparks
07-04-2008, 03:10 PM
or how 'bout just machining soft iron pole pieces and let the magnets lie where they're flang?

lazlo
07-04-2008, 03:32 PM
If I could find motor arcs the right size in a NIB I would be interested but I havent found any.

How hard is Samarium Cobalt? Can you machine it?

Be a bitch to get it off the chuck :)

Swarf&Sparks
07-04-2008, 03:34 PM
But, OTH, NIBs in a plastic bag are great for CI swarf :D

Evan
07-04-2008, 10:50 PM
As far as I have looked it appears that not only are samarium cobalt more expensive but they must be harder to work with as they are only available in a very limited range of sizes and shapes.

The Neodymium magnets are available in a wide range of grades and strengths and the maximum operating temperatures also cover a wide range. The standard low grade NIB has a maximum operating temp of only 176F. If heated above that it loses some of it's strength permanently and if heated to the curie temp at about 590 it is permanently demagnetized. At the other end of the range the NEH type have a max operating temp of about 400F which makes them entirely suitable for motors. Within each of the six grade types there are sub types that have different energy products.

lazlo
07-04-2008, 11:32 PM
But, OTH, NIBs in a plastic bag are great for CI swarf :D

Yeah, but you'd be picking magnetized swarf out of your machine for the rest of your life ;)

pntrbl
07-05-2008, 02:16 AM
I dunno if it'll be any help to you Evan, or make any difference at all, but I've been working with a 5.5 Briggs starter motor recently and that little sucker pulls 50 amps!

I'm using a 100 amp panel meter that has a 50mv shunt and I'm reasonably certain it's accurate. Initially I didn't have a flywheel so I did some no-load testing and found it drew 20 amps. That led me to select a 40 amp Radio Shack relay and you can guess what happened when I got a load on it. :D It made about 3 starts and the relay contacts welded up!

At first I blamed the cheapo Radio Shack part but then got to wondering about current under load. Big difference. The panel meter says 50 amps and since a 40 amp relay didn't survive I'm believing it.

I don't suppose the actual accuracy of your current measurement is as important as the relative differences you see on the road, but I'll bet a 15HP starter motor pulls more than 5.5 does.

Electronics is fun stuff. You can burn up a lotta $hit for $20 ....

SP

Evan
07-05-2008, 02:31 AM
This one pulls about 30 amps cruising on the level. Coming up the steep part of the hill which is maybe 12 or 13% grade it sucks up about 80 to 90. I have it geared in low range at 44 to 1 so it isn't running close to stall. PM motors draw maximum current at zero rpm. The batteries I have are rated to put out 244 amps each, maximum. Each battery has it's own set of cables to the starter solenoid and to ground. On the motor side of the solenoid I ran two parallel cables to the motor. The cable is 2 gauge copper stranded from an old set of booster cables. I also have a safety disconnect that operates like a card lock. It is easily reached when riding and all it takes is a yank on the credit card size piece of circuit board to remove it from the card slot. That disconnects both negative terminals from the frame and disables everything. The card slot also serves as an amp meter diagnostic port.

darryl
07-05-2008, 07:11 AM
Evan, I've done several mods to replace the ceramic magnets with neos. Most were very small motors, one being from a shaver. In all the standard type motors everything went up except the motor rpm. Heat, current draw at idle, cogging effect, torque and power output all went up. The best 'conversion' was the shaver motor, because it had an unusually large commutator and could handle the extra current draw under load. I was amazed at how much extra power output capability this motor had- but it did get hot quickly.

Torque capability will go up with stronger magnets, but so will current draw and thus the amount of heat to be dissipated. You can regain the lost rpm by rewiring the armature with less turns, and this also redudes resistance, but the brush and commutator area needs to be upgraded in order to make a magnet conversion really feasible. You can take advantage of the increased power output capability by raising the voltage to the motor without rewiring the armature or increasing the size of the brushes and comm, but you soon run into arcing problems on the comm.

I'd have to say that in general you can upgrade a typical motor to about twice the strength of the magnets before you run into problems, as long as you provide enough cooling and can live with the increased voltage requirement to get the same range of rpms.

I did the magnet swap in one pancake motor, and while the rpms dropped, the increase in torque was amazing. The idle current did not increase substantially in this motor, so I believe that's a better type to to the mod on because you don't lose light load efficiency like you do on an iron armature type motor.

Evan
07-05-2008, 09:47 AM
Thanks Darryl, I knew you had done some experiments. You didn't quite say it but did you mean that the "idle" (unloaded) rpm drops with the stronger magnets? If so that explains the increased current draw as the back EMF is rpm dependent. It is probably due to hysteresis in the armature and the time lag required to switch magnetization of the iron. That would also explain why an air core motor doesn't show the same effect.

So far in my single similar experiment the motor is running cooler. That though is because I seriously upgraded the cooling when I installed the magnets by enlarging the vent holes in the brush assembly to increase air flow from the turbine fan I have on one end. Also, I don't need to run the motor as long to get from A to B and especially when hill climbing. Since it takes less time because it is going faster that translates to less heat buildup. I am really surprised at the difference just adding those magnets has made. In fact it has shown up a slight weakness in another part of the transmission that I have to improve today.

I also upgraded the bike by adding a different front fork that has a motorcycle style oil damped front suspension. I stopped at the local mountain biker store and traded one of my protest license plates for a couple of old bikes from their junk pile. I got a really cool score in the form of a couple of handlebar shifters that work like twist grip throttles so that is perfect for the e-bike as that is what one will become.

BTW, if anybody wants a copy of my protest bike plate here is a link to the full resolution image.
217kb
http://www3.telus.net/metalshopborealis/taxthis2hi.jpg

It now looks like this and fits on 8.5 x 11" paper. I printed it out on a peel and stick transparency and stuck it to a piece of white formica laminate. It almost looks real. :D

http://vts.bc.ca/pics4/protestplate.jpg

A.K. Boomer
07-05-2008, 11:14 AM
Thats great, dont think anybody will confuse the point your trying to get across,
If motorcycle insurance was as expensive here as it is there I would do something similar, but for now I think I'll just keep the idea of the darwin fish mounting and sodomizing the jesus fish (it can do that -- its got legs:p )

darryl
07-06-2008, 05:34 AM
The back emf is rpm dependent, and also dependent on the strength of the magnetic field. If you increase the strength of the magnets, the back emf rises thus preventing the motor from reaching the same rpm as it would have with the weaker magnets. The workaround is to reduce the back emf by reducing the number of turns of wire on the armature. A happy coincidence is that the resistance of the armature goes down. Another effect of all this is that the range of rpm over which the motor goes from no load to maximum torque becomes less.

With higher magnetic field strength penetrating the armature, the eddy current losses increase. I don't know how much loss percentage wise is due to eddy currents or hysterisis, but the obvious solution is to keep the magnetic field in the iron from changing, as is done in a pancake motor or in a tubular can type that has an iron-less armature.

Either of these motor types has a relatively large air gap so the printed or cup shaped armature can rotate without scraping the magnets or the iron. An interesting (maybe only to me) thing that I've seen is where the armature is the magnet and the outer part is the field, like how an alternator is built, but the field doesn't have slots for wire. Instead, a series of copper coils is laid inside the field iron, glued to it, so the only air gap is between one side of the copper coils and the armature. Activating the coils in sequence causes a magnetic field to move around in the field iron, but there is no cogging. I have not investigated the eddy current losses in such a design but I know they will still exist.

Another thing I've done to improve a motors output is to combine two motors into one. Some motors have a relatively short magnet and armature lamination stack. When I built my accessory drill press for the lathe, I combined two motors such that the armature stack became about 50% longer, and I also doubled the length of the magnets by using all the magnets from both motors glued into a single can. I read somewhere about certain 'nerds' being happy about enmeshing themselves in yards and yards of magnet wire, and I suppose that description fits me to some extent- anyway, I obviously have to rewind the armature if I'm going to that extent. Usually I use some form of mental gymnastics to determine what gauge of wire to use and how many turns to put on. In most cases I get pretty close to design rpm per volt, and in most cases the size of the comm and brushes is the limiting factor in the motors performance.

I suppose I'm rambling on a bit now, but I'll mention one other idea I had, but haven't tried. This could have good potential especially with the stronger magnets. Normally a motor has a can, endbells, etc. The gap between the magnets and the armature is fixed. What if you could vary that gap mechanically- you could have a large gap which would reduce the magnetic strength in the armature, allowing the motor to turn faster with a given voltage and with less losses under lighter loading. Then when high torque is needed as when starting out or hill climbing at a slower speed, the gap is reduced and the motor gives more torque while running slower. Essentially you get a fairly typically constructed motor that is efficient at high and low speeds and high and low outputs.

topct
07-06-2008, 08:39 AM
The electric trucks we used at Kaiser used a mechanism on the brush holders that would rotate them to change the armature to field timing. It did it automatically and the amount it would move was dependent on the load.

Evan
07-06-2008, 09:11 AM
I will definitely be doing some more motor experimentation. The main problem is getting magnet wire. We used to have a motor rewind shop but they are out of business and I don't know any other place to get any locally. I can order it but that means buying 12 lb rolls to get any sort of decent price.

darryl
07-07-2008, 04:40 AM
My bicycle will be getting a four motor drive system when I get around to it. I have four 80watt pancake motors which I will arrange in pairs facing each other with a drum connecting the shafts. The two pairs of coupled motors will be spaced about 10 inches apart and will have a flat belt running on the drums. The spacing between the motor pairs will adjust to set the belt tension.

This assembly will straddle the rear tire and pivot down onto it so the belt wraps partway around the tire and drives it by friction. A little playing with the geometry and belt tension will optimize this drive system. If I was to get really ambitious, I would make a tapered cone to fit each motor shaft, then arrange the motors the same way, but allow for a side to side spacing adjustment. This would give a way to have a range of infinitely variable speed on the belt, and it would also cause the belt to conform somewhat to the shape of the tire besides wrapping partway around the circumference. I have no idea if the belt would track- it might just want to run off one side or the other, I suppose I could arrange some small motors and turned plastic cones to test the concept.

By the way, I have used some of the pvc I scavenged to make ordinary pulleys for v-belts, and they wear surprisingly well. I would have no problem using that material to make my drums or cones. I'll probably stick with using drums (rollers) for the flat belt, so that means I'll have to glue at least three thicknesses of my pvc material to make a drum 3 inches wide.

Anyway, by using the four motors I can get 'gearing' by arranging the motor wires in series, series/parallel, or parallel. Because these are 12v motors, all parallel would give high speed/high torque/high current draw, while series/parallel would give lower speed, and all series would give lowest speed. At the same time, all motors in series would give the strongest regenerative braking effect, while series/parallel would give lesser braking, but still effective.

If I went to a 24 volt battery pack, I could use the all series and the series/parallel connections to drive with, but only the all series mode would give braking (unless I was coasting down a steep hill at breakneck speed). I wouldn't be able to use all parallel with a 24 volt battery, as that means each 12v motor would be seeing a 24 volt input, and the current draw would go up by 4 times (so would the power, and the heat, and there blows the motors). Maybe that would be my overdrive- a one use quick get-away blast to outrun the cops- :)

Well, that's my idea for my own bicycle- chances are better now that I'll do this with gas prices the way they are-

Of course, like everyone else using batteries I'll have to charge them somehow. I'm thinking of finding some darkness-powered solar cells so I can charge at night :). Or maybe a microbial reduction system to generate electricity- run it on potato peelings and grass clippings- or even electric eels :)

Hmm- back to a serious note, it doesn't offhand seem unfeasible to generate electricity using bio-decay- has anyone heard of this being done?

Evan
07-07-2008, 05:51 AM
Funny, one of the items I experimented with was a pair of pancake motors mounted facing each other with a HTD pulley mounted like a drum on the two shafts. These particular motors didn't have enough torque to warrant further exploration of that idea.

320 watts is probably sufficient for level cruising but will only be an assist device on any sort of hill. Mine is running as high as a kilowatt on a steep hill which it now climbs with ease. Since I increased the ventilation along with this magnet mod the motor is only becoming slightly warm to touch even after such heavy loads.

An interesting characteristic of the batteries is that they increase capacity over the first hundred discharge cycles or so if properly charged. As long as the VRLA batteries are not improperly charged the can last for a couple of thousand discharge cycles. They also have very low self discharge rates if kept cool and can retain up to 80% of full charge after two years of sitting unused. They can withstand hundreds of total flat dead discharge cycles as long as they are not left in that condition for any length of time.

I took apart a 20 dollar ceiling fan yesterday and it looks like it has the potential to make a powerful type of hub motor. This is going to be a very interesting source of plenty of time consuming experiments this coming winter. I'm going to need an instrumented test stand to quantify motor performance under known replicable conditions.

darryl
07-09-2008, 03:10 AM
Two things- yes, I agree 320 watts is only enough for level cruising, and I don't need or intend anything else so that's ok for my 'cycle. Evan, you have a pretty good test for your electric, what with the hill climb and the distance. From what you're saying, it seems like you have a pretty good handle on making your system work. Good for you.

Second thing- vrla batteries, what's the scoop on those?

Evan
07-09-2008, 06:05 AM
VRLA=Valve Regulated Lead Acid. The electrolyte is adsorbed in fiberglass mats which are the plate separators. It isn't a gel cell as the electrolyte isn't gelled. The cell is fully sealed with a pressure valve to allow the escape of hydrogen if the cell is overcharged. That of course is undesirable. VRLAs have by far the best resistance to full discharge damage and are rated for complete discharge to dead operation without damage as long as the battery is recharged promptly. They are also very reasonably priced. The ones I bought are $50 each for 12v @ 18 amp hour capacity. While they have a Motomaster label on them you can peel it off. They are made by Exide.

NEWS FLASH:

I am changing motors. I was in Surplus Herbie's (broken website http://www.surplusherbys.com/index.php) a couple of days ago and spotted some nondescript 12v can motors on sale. They normally wanted $40 for these and they have a bin full, several hundred. That was too high for me to take a chance testing them. The motor is about 2.5" OD and about 4" long. It has a 1.25" long flatted 5/16 shaft and is made by Unimotor. I found a bin with 5 of these marked down to 10 bucks because they rotate "reversed direction". ???? They are reversible DC motors, it says so on the label. I am not one to argue with a dummy (inside joke) so I bought 2 after a quick test. I have found that a very good indicator of DC motor torque is how much resistance it provides to manually rotating the shaft with the leads shorted.

I picked up 2 to play with and mounted just one in place of the modified starter motor. Hot damn Sam! On 12 volts it was drivable on the flats. It never even became warm as it draws less than 2 amps at idle! The motor is very low rpm, less than 2000 unloaded and under level cruising at about 10 kph still only draws about 6 amps. Then I wired it for 24 vdc. ZOOM. It's almost as good as the starter motor was before the magnet mod.

Even with zero cooling system it only became slightly warm after heavy loading. The motor weighs less than half the starter motor and is much more efficient. I went back yesterday and bought the remaining 3. I am now remaking the drive module to accommodate up to three of these motors. 2 will be used for normal operation and the third will be a boost motor for overdrive operation and hill climbing. I am remachining some of the parts to handle the anticipated increase in torque and am possibly going to switch the belt drive to a chain drive around the wheel perimeter.

I haven't taken one of these apart yet to see what they are but the magnets are strong enough that the field "leaks" out of the can.

I'll post an update as soon as I have the new configuration running. It looks like these motors might be able to at least double the range and increase power. They seem to be highly efficient

darryl
07-10-2008, 05:12 AM
Thanks for the info on the batteries, Evan. Sounds like the right ones to use. And nice score on the motors. It's always nice to find a motor that runs efficiently and has a good power to weight ratio. Makes me wonder how some seem so poor while others are so good.

I'm sure you know this, but your magnetic flux 'leakage' will increase as you run more current through the motor. It's a simple matter to add some sheet steel around the cases to contain this leakage. I've seen it on cordless drill motors, for instance, and on some dremel type tool motors, etc. The motor will gain some power, but of course there is the added weight to factor in to the overall efficiency equation. Maybe the extra sheet metal could be used as part of the motor module framework.

Evan
07-10-2008, 06:12 AM
I took one of the new motors apart to see what is going on. The answer is simple. They have huge ceramic magnets compared to the armature size and windings. In fact, the armature is only about 3/4 filled. I could easily go up a gauge on the wire and still fit it all in the slots if I rewound the motor. That's a LOT of work though and I don't have the wire to do it. I once rewound the field windings on a 2000 watt Generac alternator designed to mount under the hood to provide 120vac power. I still have it and it still works but it took a lot more time and considerably more magnet wire than I had at first estimated.

darryl
07-11-2008, 06:22 AM
We used to have several shops here where I could purchase magnet wire, now we're down to one. The largest size wire I can buy from them is 16 ga. I just paid $25 for a part roll, probably 5 lbs or so. I think that was probably a pretty good deal, considering costs these days. I needed 14 ga, so I wound a pair of 16 ga wires in parallel instead. I did this for an armature once, and though tedious, it's easier to get the wire on tight using the smaller diameter wire.

Speaking of magnet wire, I'm glad I saved those focussing coils from some scrap black and white tvs years ago. Actually, I think one coil of wire I saved was from an old speaker magnet, before permanent magnets were used. Strangely enough, I have a roll of wire which I believe is bare, but wound with thread of some kind as insulation. Odd stuff.

Evan
07-11-2008, 07:43 AM
I completed the new drive system yesterday. It's very different from the original although it uses some elements the same. It's an all chain drive semi automatic transmission with 3 motors, high and low gear and overdrive. Shifting is all electric and is integrated in the throttle position control.

A very preliminary test yesterday shows that is is powerful. A quick run up the steepest part of our hill was very interesting. It hauls me up that hill as if it were on level ground and if I click in the overdrive it accelerates even though the hill is about 12% or so. No idea what the current consumption is yet but I am running it on 24 volts for higher efficiency. The motors barely even warmed up. The transmission is very quiet. I'm using #25 chain which is standard for electric scooters and bicycle drives.

davidh
07-11-2008, 10:13 AM
pictures, pictures, pictures ? ? ?

Evan
07-11-2008, 10:29 AM
Later today maybe. We have to go to town this am to take the "new" dog in for some minor surgery. I'll bet he won't think it's minor though. :eek: