BMC 600 watt motor: problems controlling externally

[boostjuice]

Cyclone ...
DC 48V Brushless motor
1500 W

* Max speed 3223Rpm
* Rated speed 4111Rpm
* Rated output 1500W
* Max output 4329W
* Rated Torque 83.46Kg-cm
* Max Torque 98.36Kg-cm
* Rated Amp 29.52A, 75A
* Insulation class E class
* External Driver
* Control method PWM
*Ambient temperature -15C~40C *Efficiency: 93.23%
* weight 4.8 Kg

Cyclone have misinterpreted figures from the plots on the dyno sheet, whilst ommitting or 'making up' others. In particular the 93.23% max efficiency claim seems to correspond with the PF (Power Factor) curve. Rooky mistake? or does a higher number just sound better.... Don't even get me started on the "4329W" max power claim, the scale on the chart doesn't even go that high....

 

[jmygann]

do you need 9 motor gears ? would 3 do ?

 

[mrbill]

I'm in the process of procuring another of these "broken" motors (at a similar price) and am curious to try connecting the phases in Wye instead of Delta, unless someone has done this already and found it not to be a productive exercise. Will this result in different Hall placement? Is there a standard method for hooking this up. E.g. Which ends of the phases are attached in common? All "starts" or all "finishes" together, or does it matter?

I have acquired a second motor and have wired the phases in Wye with all "Finish" ends wired in common, and each "Start" end going to the controller. I decided to try using the stock Halls first, since they're already there and it would save me the hassle of wiring up my own Halls. Using the stock Halls, the maximum freespin RPM is about 55% and the power draw is about 64% that of the Delta wiring. So far so good.

But, when I mounted the motor on my bike and did some efficiency tests I couldn't get better than about 65% efficiency. Off the bike and spun on the bench the motor makes an occasional "knocking" noise at partial throttle as if the rotor is hitting something at an irregular interval, yet I can find no physical interference. Sigh. I suspect the timing is a bit off.

So, today I wired up my own three Hall sensors and placed them three teeth apart and rotated them about the stator to different positions, but I could not find one that worked. I started with the same position I discovered with the Delta-wired motor, at the center of the tooth immediately CW to the stock Halls, but on this motor that position was not usable.

I did a couple things differently this time. As I have not yet done destructive filing of the stator teeth, I positioned the Hall sensor on top of the stator tooth so that the top of the device (rather than the front with the printed marking) was closest to the magnet ring. Will this cause the Halls to switch unpredictably?

My next step is to file the stator teeth immediately CW to the original Hall sensor placement, where I found the ideal Hall placement for the Delta motor, but before I do this surgery I'd like to confirm a couple of things with the gurus on this board.

1) Ideal Hall sensor position for CW rotation will be the same for Delta or Wye configuration for this motor. (18 stator teeth, 16-pole magnet, 3 phases)
2) Hall sensors must be placed so that the label side of the device is facing and closest to the magnets.

Any other suggestions?

Thanks.

 

[fechter]

I did a couple things differently this time. As I have not yet done destructive filing of the stator teeth, I positioned the Hall sensor on top of the stator tooth so that the top of the device (rather than the front with the printed marking) was closest to the magnet ring. Will this cause the Halls to switch unpredictably?

Thanks.

This is possible. I think the face with the printing should be closest to the magnet ring regardless of the orientation. I think they are plenty sensitive so that it should work with the halls facing the edge of the rotor. They can even work on the outside of the rotor in some motors.

 

[mrbill]

I'm in the process of procuring another of these "broken" motors (at a similar price) and am curious to try connecting the phases in Wye instead of Delta, unless someone has done this already and found it not to be a productive exercise. Will this result in different Hall placement? Is there a standard method for hooking this up. E.g. Which ends of the phases are attached in common? All "starts" or all "finishes" together, or does it matter?

Wye connection should not affect hall placement. I don't know whether you want all the starts or all the finishes togther. It will reverse the motor direction if you get it backwards, but you can fix that by reconfiguring the hall/phase wiring. Other than that, it doesn't matter.
I think it should be possible to switch between delta and wye using the same hall positions.

I have completed my exercise of wiring up a second of these BMC motors using a Wye configuration and cannot find a configuration that gives acceptable performance.

In this most recent trial I filed away a small part of each of three stator teeth, immediately clockwise to the stock Hall positions and mounted the Hall sensors as I had in the first motor. Assuming both of these motors are physically equivalent, this is the same Hall position I found that gave optimal performance for the Delta motor.

At 26 volts, full-throttle, freespin, the motor draws about 30 watts, and the RPM is about 68% of the Delta motor. But, the maximum efficiency is about 66%, Not the 75-80% that I saw with the Delta motor. I tried moving the Halls slightly one direction or the other when using the same phase lead hookups, and found that the behavior was quite sensitive to Hall positioning. Moving the Halls even one device width (1/5-tooth width) changed the freespin wattage and RPM significantly. The position at the center of the tooth gave the lowest wattage.

Other behavior noted:

1) When spun very slowly so that I could see the rotor rotate in step-wise fashion, I noticed that the angle of each step was not constant. It appeared to be a long-long-short pattern.

2) When spun very very slowly, the rotor would stall every three steps. This is at minimum throttle, just enough to get some motion.

3) The motor would not spin smoothly but would knock every rotation. This may be a problem with one of the bearings as I could sometimes duplicate this knocking noise by rotating the rotor by hand. But, the bearings do not appear to be worn when inspected visually.

I suspect that the odd 18-tooth/16-pole configuration of this motor prevents it from operating efficiently in Wye mode. I would like to hear of someone has found a way to get this motor to run efficiently in Wye mode, but for now I conclude that it is not possible, at least with my unsophisticated controller. I will probably reconfigure this motor in Delta mode and keep it as a spare.

 

[mrbill]

I will probably reconfigure this motor in Delta mode and keep it as a spare.

I have rewired the second motor (M2) in Delta mode and have made some findings:

1) M2 spins faster than when wired in Wye mode, but not quite as fast as my first BMC motor (M1), referenced at the start of this thread.

2) Freespin wattage (at 26 volts) is about 70 watts compared to 50 watts with M1 after I adjusted Hall position for minimum freespin, full-throttle power.

3) Spinning is not entirely smooth and slightly noisy, as it was when I tested in Wye mode. It spins more noisily and roughly than M1.

4) At 25 volts efficiency peaks at 69% (~550 watts input) and drops to 65% (~1000 watts input) and 60% (~300 watts input). This is about 5-10% worse than the efficiency I measured with M1.

5) Hall sensor position for minimum freespin wattage was one Hall device width CCW to top-dead-center of tooth immediately clockwise to stock Hall sensor position. (M1's ideal Hall position was exactly top-dead-center of tooth immediately clockwise to stock Hall sensor position.) Why should the Hall position be any different if the motors are otherwise identical?

I checked the windings, and they are identical to that of M1 both in number of turns and direction (CW or CCW) around each stator tooth. I also checked that the magnet ring has 16 poles; it cogs 16 times when rotated by hand through one revolution.

The magnet on the inside of the rotor bell shows some evidence of scraping on the stator teeth along the lower edge where there is a faint wear groove on one side, but I cannot see any contact with the stator teeth when I rotate the bell by hand. The only thing I cannot test or examine is whether the magnets have retained their original strength. This motor (M2) was labeled a "T1" motor by BMC and had additional controller circuitry that I have since removed and discarded. (It was an extra daughter board mounted to the top of the main controller board.) I was told this was a "high torque" motor. Maybe it was abused in its first life.

My conclusion is that this motor is defective and that not all BMC motors with broken controllers have good motors inside.