1000W BMC testing

[oatnet]

For those of you who are using an xlyte controller, does this look like the pinouts you are using? I didn't have time today to set up a test rig and try the 36 combinations not to mention that I am skeptical about the functionality of the hall connector I have...

Here is the thread with pictures of the bike I am building around the BMC V3 1000w:
http://endless-sphere.com/forums/viewtopic.php?f=6&t=12461&p=187988

-JD

 

[dirtdad]

I just pulled this from the "Known to Work..." thread. The 1000W BMC wires up like the 400W and 600W. Crystalyte has also changed the colors of their wires over time and with different models, so there is that disclaimer, too.

Phase Wires
motor = controller
yellow = green
blue = yellow
green = blue

Hall Sensor Wires
motor = contoller
red = brown
black = black
blue = green
green = yellow
yellow = blue

 

[BVH]

Maybe a little off-topic...If the Hall sensor wires are not correctly matched to a controller, is it possible to have an "ok running" motor or it it more of a "all or nothing"/"it runs or it doesn't run at all" situation? I'm just riding my BMC V3 bike for the first time and it's my very first Ebike. I'm using a Kelly controller. Both the Phase cables and the Hall sensor wires are all matched color to color. I haven't opened it up yet but I can easily do 30 mph so I'm guessing that it's wired correctly. It's got what I consider to be a modest amount of gear noise when under power and I just want to be sure it's not some type of other motor electrical noise from not being hooked up 100% correct.

 

[skeetab5780]

BVH you may want to check out this post. sorry i cant be of much more help than this, but there is alot of cases where color for color on phase/hall sensor wires would be incorrect. Im probly not the person to tell you right or wrong. But this post will push you in the right direction.
http://endless-sphere.com/forums/viewtopic.php?f=2&t=6241

 

[dirtdad]

There is usually a combination that will give you forward movement but be rough and/or slow. As I recall, matching up colors between bmc/Kelly is one of those bad combinations that works anyway.

 

[oatnet]

Thanks Dirtdad! Yeah, I found that one when I searched E:S I was just looking for realworld experience. Actually that is the same pinouts I posted above.

I've gone through a dozen or so xlyte controllers over the past 2-3 years, and I haven't seen the wire colors change... However ebikes.ca recently changed their pinouts, and since I always lookup the pinouts diagram on their site, I wired the new pinouts on an old controller. I rewired per the original diagram and it works like a charm. Too bad I wasted all my free time figuring this out so that I didn't have time to actually ride.

Oh, and Ilia reports that the BMC flat connector is wired exactly backwards of the New ebikes.ca standard. Oy.

-JD

ORIGINAL RECIPE:

NEW EXTRA-CRISPY RECIPE

Ilia's BMC-to_kelly diagram, showing the reverse configuration of the new Ebikes.ca standard

 

[dirtdad]

Yes, and Going Electric is proposing another "standard". The problem with standards is they dont usually succeed if they are being forwarded by one company. Sometimes a "defacto standard" manages to win the day. This is why organizations like ANSI exist. I am not suggesting that we could ever get a standards body to propose something chinese manufacturers would ever pay attention to, however.

 

[BVH]

Thanks guys! I found Shinyballs posted his BMC V2 to Kelly in the thread linked. Mine is exactly the same on the V3 1000 Watt

Kelly..........BMC V3 Phase cables

Phase A - Yellow
Phase B - Blue
Phase C - Green

Kelly..........BMC V3 Hall wires

Red (Wire tag 7)------------Red
Blue (Wire tag 12)----------Blue
Green (Wire tag 11)--------Green
Yellow Wire tag 13)---------Yellow
Black (Wire tag 14)---------Black

 

[oatnet]

Thanks guys! I found Shinyballs posted his BMC V2 to Kelly in the thread linked. Mine is exactly the same on the V3 1000 Watt

Kelly..........BMC V3 Phase cables

Phase A - Yellow
Phase B - Blue
Phase C - Green

Kelly..........BMC V3 Hall wires

Red (Wire tag 7)------------Red
Blue (Wire tag 12)----------Blue
Green (Wire tag 11)--------Green
Yellow Wire tag 13)---------Yellow
Black (Wire tag 14)---------Black

Yep, that matches Ilia's diagram I put up 3 posts above. He also has posted his settings for the Kelly Controller:
http://ebikessf.com/bmc-to-kelly/index.html

-JD

 

[RPA]

Finding a controller is what this thread is all about. We are selling the motor because BMC says they will have a controller in August.

Any word on this new BMC controller yet?

 

[dirtdad]

BMC's controller release date keeps getting delayed. Kelly is also working on a controller for the motor. Kelly's controller release date also keeps getting delayed. Kelly recognize the heat problems caused in the controller by the low resistance in the motor and the high electrical rpm requirements and are updating the firmware of the controllers to accommodate it. I hope to have a sample of both controllers to test at some point but given the delays and complexity of the problem, I am not holding my breath anymore. However, I am ga-ga about the possibilities.

 

[RPA]

BMC's controller release date keeps getting delayed. Kelly is also working on a controller for the motor. Kelly's controller release date also keeps getting delayed. Kelly recognize the heat problems caused in the controller by the low resistance in the motor and the high electrical rpm requirements and are updating the firmware of the controllers to accommodate it. I hope to have a sample of both controllers to test at some point but given the delays and complexity of the problem, I am not holding my breath anymore. However, I am ga-ga about the possibilities.

Thanks for the update. I figured it would be delayed. I can't wait to hear the performance numbers with the new controllers but I'm not expecting much more than I already have. I usually max out in the high 30's low 40's at 48 volts and I can't see BMC releasing a controller that can handle more voltage than the motor is rated for. I'm very pleased with this motor so anything extra is great.

 

[oatnet]

I konw the puma had 32 poles, and a gear ratio of 5:1, which meant 160 rpm for each turn of the wheel. Since the Kelly can handle up to 60,000 rpm, then the most it should produce with that motor is 375 wheel rpm. Figuring my tires are 27" in diameter, hence 83.21" in circumferance, they would take 761.45 revolutions to cover a mile. So at 375rpm, it would take about 2 minutes to cover a mile, meaning a 30mph top speed on a standard Kelly. I see a BMC-specific controller on Kelly's website, I wonder how many RPM it can handle.

I keep thinking how this problem sounds very simiilar to the problems puma's/bmcs used to have. I know smart folks fixed this on the kunckles-Infineon thread with a faster clock speed, I don't know if current infineons are also overclocked.

Does anyone know if the number of poles and gear ratio in the 1000w BMC?

-JD

Edit:

Whups, just stumbled across this thread:
http://endless-sphere.com/forums/viewtopic.php?f=2&t=10699&p=164066&hili#p164066
The puma has 32 magnets = 16 poles, which means a 600000rpm Kelly controller should be good up to @60mph, and the new 70,000 rpm version @70mph, if the poles are the same as the Puma. Has anyone cracked the case on ther V3 yet and counted the magnets?

 

[RPA]

I was just browsing Kelly's site and found this: https://www.kellycontroller.com/shop/?mod=product&product_ID=742&cat_ID=41,39

 

[oatnet]

I was just browsing Kelly's site and found this: https://www.kellycontroller.com/shop/?mod=product&product_ID=742&cat_ID=41,39

Nice Find!
==========================
KBS72121, 72V for BMC V3 motor
Weight: 2.3 lbs.
Stock: 50
This KBS72121 controller is produced specially for BMC V3 motor.
Up to 70000 electric RPM with high speed firmware.
Can not do regenerative braking.
Controller does not have some secondary functions.
==========================

 

[Muad'dib]

I guess it is too new for detailed specs. It looks like the phase wires are some kind of micro gauge, don't know how that could work. The stock phase wires on the V3 are the weak link that holds back the motor from doing what it was advertised to do: they heat up and melt even when I capped the controller at 30 amps. That is with a Crystalyte controller, but my understanding was that it is a simple matter of needing bigger wires for carrying current.

 

[oatnet]

Huh. Here are few more that I think just popped up:

===================
KEB72330, 72V for BMC V2 or V3 Motor
Weight: 5.5 lbs.
This KEB72330 controller is produced specially for BMC V2 or V3 motor.
$259,
https://www.kellycontroller.com/shop/?mod=product&product_ID=688&cat_ID=41,22
===================
KEB72450, 72V for BMC V3 Motor
Weight: 5.5 lbs.
This KEB72450 controller is produced specially for BMC V3 motor.
$299
https://www.kellycontroller.com/shop/?mod=product&cat_id=41,22&product_id=683
===================

So I guess now we have a selection now. I am guessing on amp loads/RPM baased on the controller number :
$259, 5.5 lbs - KEB72330, 72V for BMC V2 or V3 Motor (50a continuous, 100a burst)
$279, 2.3 lbs - KBS72121, 72V for BMC V3 motor (48a continuous, 120a burst)
%299, 5.5 lbs - KEB72450, 72V for BMC V3 Motor (75a continuous, 150a burst)

It is hard to imagine putting 150a bursts through those tiny phase wires. If they were clipped to 6" long, just clear of the axle, do you think they could handle 150a for 60 seconds without melting the phase wires? I am skeptical.

I already have a few KEBs, one on my xtracycle and another that I was planning on testing this motor with. They are great, solid, reliable controllers, but their size makes them hard to hide on a stealthy ebike (unless you have an xtracycle... ) Llooking at the manual for the KBS, I see the mounting plate is 100mx78mm - incredibly tiny, perfect for a stealth bike. It is incredible that this tiny package can handle @3500 watts continuous.

I had problems finding the KBS page with the manual before (I must have been stuck in their 'store' site), so here is the link, as well as a copy of the KBS manual. While this page it clearly says "Brushless", the contrllers are listed on the bottom under 'Series' (aka without regen) and "Permanent Magnet" (with Regen) :lol:

http://www.kellycontroller.com/mot/KBS.html
View attachment KellyKBSUserManual.pdf

-JD

 

[BVH]

My KBS 72121 should arrive Tuesday or Wednesday. I'm currently using their 72450 but want something smaller. I was told the KBS72121 for V3 BMC is using different firmware than the standard KBS72121 and is tailored towards speed while the standard KBS72121 is tailored towards torque. I opted for the torque version.

 

[oatnet]

Hmm... Here is a (3rd) kelly website:

http://www.newkellycontroller.com/

 

[dirtdad]

I have two of the Kelly V3 controllers in stock and 3 V3 motors, one built in a Mavic XM729 rim. I tested them @48V and it worked fine. The issue is my battery pack is only good for 30A or so, which is not enough to extract any more from the V3 motor than the V2. If you have a nice high discharge pack, the Kelly controller, and the V3 motor you should have a very powerful set up.

 

[fechter]

BMC's controller release date keeps getting delayed. Kelly is also working on a controller for the motor. Kelly's controller release date also keeps getting delayed. Kelly recognize the heat problems caused in the controller by the low resistance in the motor and the high electrical rpm requirements and are updating the firmware of the controllers to accommodate it. I hope to have a sample of both controllers to test at some point but given the delays and complexity of the problem, I am not holding my breath anymore. However, I am ga-ga about the possibilities.

Did they give any more clues as to why other controllers don't seem to work with these motors?

 

[dirtdad]

BMC's controller release date keeps getting delayed. Kelly is also working on a controller for the motor. Kelly's controller release date also keeps getting delayed. Kelly recognize the heat problems caused in the controller by the low resistance in the motor and the high electrical rpm requirements and are updating the firmware of the controllers to accommodate it. I hope to have a sample of both controllers to test at some point but given the delays and complexity of the problem, I am not holding my breath anymore. However, I am ga-ga about the possibilities.

Did they give any more clues as to why other controllers don't seem to work with these motors?

I was in touch with them all during testing the motor/controller combination. Their stock controller was not able to do a high enough electrical RPM. They doubled the max PWM frequency and did several rounds of firmware upgrades to support it. They claim they were able to reduce phase wire heat due to their reprogramming. Without some higher discharge batts I cannot test it, but the controller works well and the motor pulls like a tractor even with 48V 30A available, the phase wire only just get warm.

That is all I could get out of them, but it does not satisfy my curiosity. The Infineon could already handle those high electrical RPMs, and so can the Crystalyte, as far as I can tell.

 

[steveo]

Hey everyone,

I though i might bump up this post again.

Has anyone ever thought of the reason why the bmc motor runs so hot is due to how the motor is wound?... the bmc stator is wound completly difference from almost all brushless motors out there....

I have someone that has tested this motor with a few different setups ... please check below..

-steveo

 

[voicecoils]

Hey everyone,
I though i might bump up this post again.
Has anyone ever thought of the reason why the bmc motor runs so hot is due to how the motor is wound?... the bmc stator is wound completly difference from almost all brushless motors out there....
I have someone that has tested this motor with a few different setups ... please check below..
-steveo

Very useful chart, thanks. Do you know if the tester was using a power supply that held the voltage at 36v, 48v etc exactly? Or is that a nominal voltage being used with a specific chemistry (eg lifepo4 = 3.2v, 48v lifepo4 is really 51.2v nominal). I'm just trying to get an accurate sense of the respective RPM/volt of BMC motors. Cheers!

 

[Muad'dib]

Hey everyone,

I though i might bump up this post again.

Has anyone ever thought of the reason why the bmc motor runs so hot is due to how the motor is wound?... the bmc stator is wound completly difference from almost all brushless motors out there....
-steveo

You can see for yourself how it is wound here: