* * * THE BRUSHLESS BMC MOTOR * * *

[safe]

Complete Analysis

My goal here is to gather every possible piece of information published for the BMC motor. I'm looking for things like the maximum rpms, no load current, inductance, internal resistence, etc. I'd like to also assess the purchasing options. (where to buy it) In the end I will do my very best to create a spreadsheet that accurately models all the parameters of this motor so that (like with the Unite Motors already modeled) it's possible to run simulations of what they can do.

36V
Rated output wattage 738
RPM @ rated wattage 3116
Efficiency (%) @ rated output 86.8
HP @ rated output 0.989
HP @ 30 in/lbs 1.345
Shaft, for Currie bike drive "D" shaft
Rotation viewed from shaft end CW
Input Voltage 36.0
Current @ rated output, Amps 23.6
Current limit, Amps 35
Motor weight (lbs) ~5
Motor dimensions H=3-1/4" D=5"
(black) Controller weight (lbs) ~2.75
(black) Controller dimensions L=7-3/32" W=3-1/4" H=1-7/8"
(silver) Controller weight (lbs) ~2.50
(silver) Controller dimensions L=7-3/8" W=3-7/8" H=1-5/16"

@ 48V
Rated output wattage 883
RPM @ rated wattage 3730
Efficiency (%) @ rated output 82.5
HP @ rated output 1.184
HP @ 30 in/lbs 1.545
Shaft, for Currie bike drive "D" shaft
Rotation viewed from shaft end CW
Input Voltage 48.0
Current @ rated output, Amps 22.2
Current limit, Amps 35
Motor weight (lbs) ~5
Motor dimensions H=3-1/4" D=5"
(black) Controller weight (lbs) ~2.75
(black) Controller dimensions L=7-3/32" W=3-1/4" H=1-7/8"
(silver) Controller weight (lbs) ~2.50
(silver) Controller dimensions L=7-3/8" W=3-7/8" H=1-5/16"

http://www.evdeals.com/Motors.htm

http://powerpackmotors.com/Powerpack-enhanced-shipping_included.html

 

[xyster]

Shaft, for Currie bike drive "D" shaft

Practical question for us Currie owners: what does this "D" shaft reference mean? And how is this BMC motor supposed to mesh with the Currie gearbox that's built onto the stock, brushed motor? At >3000 rpm the BMC seems to require more than just a single step reduction.

 

[Drunkskunk]

"D" shaft means the shaft has a flat side, so that it's cross section would resemble a capitol " D "

 

[TylerDurden]

The D refers to the shaft having a single flat where the sprocket is attached.

The stock Unite motor has a 79:11 reduction before the 9:xx chain-drive.

To use the BMC, a stacked freewheel setup like RandyBike might be easiest.

 

[xyster]

Thanks for the explan. So this motor is only a bolt-on replacement for Currie's USPD kit, but not Currie's ebikes. Much of the marketing I've seen is false or misleading.

http://superkids.stores.yahoo.net/600wasutocue.html
600 Watt BMC Brushless Super Torque Bike Motor is now on sale!! - One Week Only - Get yours while you still can! This is the most powerful 24v Motor for any Currie Electric Bike, USPD Pro Drive Kit, or any Left Side Drive Currie Scooter. Smoke the competition with your 600 Watt BMC Super Torque motor!!

 

[fechter]

The BMC motor is commonly available with two shaft options. One is the double D that fits the older Currie USPD planetary reduction drive. The other one is a smooth 12mm shaft that fits a freewheeling #25 chain sprocket for scooter applications.

The efficiency numbers listed include the controller losses.

I've pushed mine to 60v, over 6000 rpm, and up to 100 amps on takeoff.
With forced air cooling, it can handle around 50 amps for extended periods of time.

 

[Toorbough ULL-Zeveigh]

Is there anyone that wouldn't prefer double D's?


(I refrain from including reference to the shaft option.)

 

[TylerDurden]

Is there anyone that wouldn't prefer double D's?

I'll just say I've never been DDissapointed.

:wink:

 

[xyster]

Is there anyone that wouldn't prefer double D's?

I'll just say I've never been DDissapointed.

:wink:

I guess dirty minds think alike ....

 

[safe]

"Rated Output" tells me what they want the motor to be "sold as", but it doesn't help much with things like the no load current amps or the maximum rpm. You really need these numbers to do the modeling of the actual motors powerband. It's possible to estimate this stuff, but I hate to get into that. Seems to me that the information about something as basic as the no load current and max rpms should be posted somewhere.

 

[safe]

email I sent to:

PowerPackMotors.com

I'd like to do some modeling of the behavior of your motor in a
spreadsheet in order to figure out things like gearing selection, top speed,
etc...

So while I've seen the "Rated Load" values posted I'd like to know other
basic parameters as well.

What is the motors:

1. Maximum Rpm at a given voltage. (24V, 36V or 48v)

2. No Load Current at a given voltage. (24V, 36V or 48v)

Also, when it comes to how the controller limits the current does it place
an actual cap on the current that enters the motor or does it monitor the
battery side current? That's a big difference because when you limit the
battery side current the motor side can be much higher, especially at low
rpms. This produces a lot of heat (and extra torque) that for a geared bike
I don't desire. Most brushed motor controllers are battery side limited...
the motor side limiting is pretty rare I think. (but it's what I would prefer)

Response:

Free running rpm is about 125 rpm/volt. Current free running at 48v is
about 3 amps. I don't have info for current at 24 and 36v.

Current limit on controllers in general is to the current pulled from
the batteries. In the case of my motor/controller the controller limits flow
from the batteries to the controller to about the rated limit- 35 amps. The
phase current in the motor can be a good deal higher. In testing when I
have fused the phases 40 amp fuses would eventually blow. 50 amp fuses
hold. Of course since the current in the phases is AC it's hard to say just
what the peak current it without measuring it, which I have never done.

Tim O'Brien
Powerpackmotors.com

 

[safe]

My impression so far is that you really can't apply brushed motor logic to a brushless motor. When I've tried to plug in the values nothing seems to line up right like the brushed motors have done in the past.

Does one have to essentially throw away all the formulas for volts, amps, power(in), power(out) and efficiency and "start over" when you switch to brushless?

Are these two animals that different?

 

[fechter]

No.
Virtually all the same formulas apply to brushless motors.

I have measured the motor current on mine, you just need an AC ammeter instead of a DC one.

What doesn't line up?

 

[safe]

What doesn't line up?

I'm trying to make sense of the PowerPack Dyno doc file.

Here's the contents of the file as a screen capture. (see below)

What's weird is that if as Tim O'Brien says the voltage/rpm relationship is 125 rpm/volt then at 48 volts the max rpms should be 6000. But when you look at the way the numbers are presented it's "seems" to be in terms of the battery side current and the torque. He's basically picked three torque figures and then found the corresponding data points that they fit into.

There might be some method to his madness however...

If you were in the business of wanting to sell a motor that stays under the radar of the laws about 750 watts this is a good way to present the data so as to hide the fact that it looks like this motor should pump out some serious horsepower at 48 Volts. I've been trying to imagine a way to extrapolate the true values for things like the internal resistance from this data... but it hasn't been easy.

Do you see how strange this data is?

(it's probably just raw data coming off of a dyno... but it would have been easier if he oriented his data around rpms rather than torque)

 

[fechter]

As far as I know, that's real dyno data and it's about as accurate as it gets.

There is no data for a no-load condition. 6k rpm at 48v sounds too high. I estimate it's closer to 5k rpm.

Like any motor, it has a voltage constant and a torque constant. It should be possible to derive approximations for these from the dyno data.

Unfortunately, the dyno data includes all actual losses, like controller, windage, friction, and core losses, so it's hard to calculate how much is attibutable to each one and get to the actual motor constants.

 

[Beagle123]

Thanks for analyzing this motor. I'd love to hear your results and suggestions.

I contacted Powerpack Motors about this motor, and he said some interesting things:

> The information on the controller should be at the website. It's a 35 amp,
> 24-48v controller. The current is electronically limited to 35 amps.
> Continuous operating wattage is about 750 and peak output is about 1400
> watts.
>
> The 1000 watt motor at Superkids is the same as my 750 watt motor-
they
> just choose to advertise it as 1000 watts. Technically they aren't exactly
> wrong, since the system can in fact put out a good deal more than 1000
> watts- just not continuously. I always stick to continuous ratings, since
> anything else can be confusing.
>
> I am currently working on a second controller option that would be
higher
> current than the regular controller. The motor can handle quite a bit more
> than the 35 amps my controllers allow it, though it will get very hot
quite
> quickly if the current is allow to get too high- especially on hills,
where
> it might pull a constant 60 amps or more if the controller would allow it.
> I'll likely limit the high current controller to around 50-55 amps for use
> with my standard motor. I may have the upgrade controllers available in as
> soon as 1 month.
>
>
>
>
> Tim O'Brien
> Powerpackmotors.com


By the way, the shaft is 12mm, which is just a hair under 1/2". It's got
about 1" of usable shaft length.

[/quote]

 

[safe]

Beagle123, you and I are both on the same basic design track it would seem. The BMC seems the obvious step up after the MY1020Z3. If Tim O'Brien comes out with a higher current controller and can somehow allow it to be "Motor Side Current Limited" then I'm in paradise.

Brushless Motor

Higher Than Sustainable (for a standard) Controller Limit

Altered Motor Side Current Limiting to compensate and lower average heat

...add that to a geared bike foundation and you have amazing results for what still passes as a legal electric bike. (gearing itself becomes the only isuue of contention)

With Fechters encouragement that the brushed motor should produce "normal numbers" I'm going to have to invent a new spreadsheet tool to somehow extrapolate from that dyno data what the real motor parameters are.

This might take some time...

 

[fechter]

I've been working with Tim on the upgraded controllers. I'm glad to hear he has some in the pipeline. He got a pile of FETs and sent them over to the factory where these things are made and had them substitute them for the 4710's in the assembly line. The rest of the mods he has to do by hand when they arrive here. Sure beats the heck out of having to do the mods yourself.

I don't know how many he's getting made, but I bet they sell out fast once he gets them in. I have no idea what the pricing will be either, but considering the amount of labor involved, if they're $100 more than the standard model, it would be a deal.

Motor side current limiting could be implemented with an add-on limiter circuit, but it would be much more complicated than doing battery side limiting.

One possibility with a brushless motor is to install a thermistor in the windings (since they don't move). The output from the thermistor could reduce the current limit as the motor heats up. In the winter, you could go like hell and never overheat the motor. On a hot day, you would be more limited. This way you could always get the maximum safe amount of power from the motor.

In short bursts, the motor can take way more than the continuous rating. Adding forced-air cooling increases the continuous rating (about doubles it).

Modeling the motor's performance really boils down to determining a few key parameters. The kV, kT are the main ones, along with the winding resistance and no-load current. I might be able to measure the winding resistance and no-load current.

Just from experience, I know that at 48v (which sags under load) the winding resistance limits the current to around 90 amps. Of course this includes the resistance of the leadwires from the motor to the controller and resistance of the battery side wiring, which can be significant depending on how long they are.

 

[safe]

I'm very interested in everything in the pipeline... but let me get back to the data for a second. The first chart shows the data as given by the PowerPack Dyno Data Doc that they have posted. It's plotted so that the data points sit at locations based on the 125 rpm/volt number that was given, so the max rpm is 6000 rpms. (I'm just starting with what has been said) The second chart shows how one might begin to draw a curve that would include that data. Obviously some scaling took place, but it's all very straight forward and would never seriously screw up the shapes, just the scale of the data. Notice how well the torque and the current match. This is a good thing and makes us think that the data points are okay. The efficiency numbers also look good and fit the generic profile of efficiency. Then we look at power out... ugh! Something is seriously messed up with that data...

 

[fechter]

If you look at the power output curve of a typical motor, the peak is at about 1/2 the no-load rpm.

I drew in an approximate curve, but as you can see, I had to 'bend' it a bit to make it fit. I suspect this bending is a result of measurement inconsistencies.

 

[safe]

In something like an unlimited current motor setup the powerband looks like this... (see chart) ...which fits what you are saying if you scale it correctly.

 

[safe]

But for a motor that is limited by a standard battery current limited controller (35 Amps in this case) you end up with a chart more like this one. (see chart) So one wonders if somehow they ran the dyno without a current limit on the controller? (this might be the explanation that has befuddled me :? )

The scaling on these charts is very different...

 

[safe]

125 rpm / volts way off?

It seems that this is simply wrong. The number that seems closer is more like 80 rpm / volt as far as I can tell. If you radically shorten the max rpms then you get on the "other side" of the power curve and it does go down there. So something is wrong... but I'm not sure which part is wrong.

 

[safe]

Data Reconstruction Theory

How does one attempt to plot data that seems to make no sense at all?

First you need to create a tool that allows you to see what is going on. With the sheer number of parameters that can bend a power curve if you are fishing in the dark for how things might look you will never find anything. So what I did was to create a chart that plots the data points and allows me to dynamically see what is going on. You very quickly learn how to "bend" the curves to do what you want them to do and in time you arrive at values for the parameters that seem to match.

The first parameters and charts are for the 36 Volt data and the second are for the 48 Volt data.

The "Voltage Constant" doesn't remain constant and goes from 102 rpms per volt at 36 Volts to 95 rpm per volt at 48 Volts. Also the apparent resistance gets WORSE at higher voltage and the amps that are pulled seem to go down. This seems to suggest that there might have been some sort of battery sag problem during the dyno test. The efficiency numbers seem to be "reverse engineered" to factor back in an extra 10% to account for mechanical and controller losses.

So the "bottom line" is that I still don't have a solid idea of what the core parameters are for this motor. The 36 Volt data is probably closest to what is correct and the 48 Volt data demonstrates problems of some kind. One hopes that these problems aren't apparent all the time and are an example of something seriously going wrong during the test, but not true in real life.

 

[xyster]

First you need to create a tool that allows you to see what is going on.

I do?

you are fishing in the dark for how things might look you will never find anything.

I am?

You very quickly learn how to "bend" the curves to do what you want them to do and in time you arrive at values for the parameters that seem to match.

I did? Cool...that was easy! 8)