I've seen photos of one apart, Eric (it's been stripped and analysed to death over on another EV forum), it's a BLDC, pretty standard except for the Lynch-motor-like axial gap coil and magnet configuration, not surprising, because John Fiorenza (the guy behind Motenergy/Mars) was part of the Briggs and Stratton team that built Lynch configuration DC motors under licence a few years ago.
As such, it works with any normal trapezoidal BLDC controller, as plenty of folk have found out.
There's nothing magical about it, it's just a standard arrangement of three phase windings and neodymium magnets, with three Hall sensors set into the stator, so it will run on just the same sort of controller as any other BLDC motor.
The terminology is a red herring, as PMAC is an uncommon term for a BLDC motor that some people seem to prefer to use. The topology of the motors is the same, it's just the terminology that's different.
I agree, sinusoidal control is better, but as I mentioned, such controllers are expensive and rare. The Sevcon Millipaks are definitely trapezoidal (albeit with some ability to modify the PWM duty cycle as the current ramps up on each phase transition, like many other controllers). I've had a couple apart (they aren't great to look at inside!) and they just use the same sort of circuitry as you'd find in a Kelly or other budget BLDC controller, if anything they are pretty cheap and nasty looking, with small format FETs that have dire thermal performance. I'm amazed more don't fail in service, to be honest, as they are no better than the budget Kellys in design. The newer Sevcons offer a "flux vector" feature that modifies the current waveform on the fly, but again this isn't a true sinusoidal controller. Even my simple BLDC controller (logic based) allows pulse-by-pulse current control, which effectively modifies the current waveform to remove over-current spikes that can occur when a low inductance motor is run slowly at high load.
The great thing about sinusoidal control is that you have no need of Hall sensors, as the controller can accurately sense and control the current vector in each winding to keep the controller and motor synced. This is pretty much how three phase AC motor controllers operate and is an excellent control strategy.
An acquaintance (who'd more adept than me at designing and building complex BLDC motor controllers!) has been working on a superb true sinusoidal vector controller for a while now. It is as near bullet-proof as you can get, as one of the really big advantages of going for this topology for a controller is that it automatically provides near-instantaneous phase current control, with the great advantage that this makes it short circuit proof. It would be expensive to put into production, though.
The problem with the vast majority of BLDC controllers out there is that many only use battery current sensing and try and derive phase current, an inherently risky strategy given the high rate of change of phase current with sudden load change, particularly for a low inductance motor running at low rpm. Some use a crude phase current monitor to activate a shutdown, some use external logic to detect fast current spikes and do the same, but neither of these techniques is very flexible if you still want to get the most out of the motor.
Jeremy
--- In electricboats@yahoogroups.com, "Eric" <ewdysar@...> wrote:
>
> Hi Jeremy,
>
> I believe that the ME0913 is actually a PMAC motor, not a BLDC motor. There is a difference between the motor types (see my recent post about the differences). I agree that a BLDC (trapezoidal wave-form) controller will run a PMAC motor and vice versa, but both types of motors will run better with a correctly matched controller.
>
> I also think that sinusoidal wave-form controllers are more common and reasonably priced than you make them sound. Even the old Sevcon Millipak controller could "operate in Trapezoidal or Sinusoidal wave-form switching modes" (Sevcon, 2008). Since the newer Gen4 controllers are designed for both PMAC or ACIM motors, I believe that they only operate in sinusoidal mode, though there may be a configuration setting to force one into a trapezoidal wave-form. I know that the Gen4 does allow for field weakening to extend the speed range of a PMAC motor beyond the BLCD limitation of max RPM when the motor back-EMF reaches the drive bus voltage. Another way to look at it is that BLDC controllers are based on a speed control loop and PMAC controllers are based on a torque control loop.
>
> On the other hand, Kelly's brushless controllers appear to be exclusively BLDC (trapezoidal wave-form) controllers.
>
> Fair winds,
> Eric
> Marina del Rey, CA
>
No comments:
Post a Comment