This is all correct (as usual), but to gain the added advantages
of the Sep-ex system you need to have an intelligent controller
that can independantly control the field and armature.
Independent control adds all of the features of a very expensive
series control with contactors, diode ect. for regen and field
weakening to increase speed.
All of this can be accomplished with an inexpensive H bridge field
circuit (inexpensive when compared with field weakening resistors,
diodes and contactors to accomplish the same effect on a series motor
that must carry armature current in the series field).
The control must be capable of determining the appropriate field current
for regen and field weakening.  Usually this requires a hall effect or
similar sensor to detect actual motor RPM.
The control must also accept accelerator input, rotor speed input
and brake input to determine how to adjust the field strength.
You can play with all of these inputs to achieve a very efficient
system with a much wider operating range than the series system.
This however requires an intelligent control.
The TEVan has the extra silicon to regen to zero speed, but the
low voltage (48V) control I did for a patended greensmower control
did not.  The greensmower required a constant 4 mph speed on varying
terrain, so it did not need down to zero regen.
I also designed a shunt system for a flatbed delivery truck that
could carry a maximum payload of 1,000 lbs.
A simple comparison of the Curtis series system in this vehicle
showed the following:
The Curtis system could not overcome a steep grade with even
half of the rated payload.
My shunt system could pull the grade with 1,000 lbs.
The Curtis series system would go 24mph max on a level grade with
no load.
My shunt system could do 30 mph.
This is a simple comparison to show the advantages of a shunt system
over a series on an EV with a fixed gear ratio.
The same could be accomplished with the series system but at a much
higher cost.
The shunt system is best when used with low battery voltages however.
If you have a high voltage pack (200Vdc +) the same could be achieved
competively with an AC motor/control.

Rod

Lee Hart wrote:
> Rod Hower wrote:
> 
>>>A sepex system designed for an application is cost competitive and
>>>superior to a series system. How do you make this work for an EV
>>>market? It's very difficult. You have to depend on the operator to
>>>program in the appropriate parameters for their application. This
>>>is assuming the motor is the same. If you have several different
>>>sepex motors it gets even more complicated. So the bottom line is
>>>series is a Wal Mart solution.
>>
> 
> I think we are confusing simplicity with motor type (series = simple,
> sepex = complex, AC = very complex). It's not that clear-cut. A sepex
> controller can be even simpler than a series motor controller.
> 
> A series motor has a natural torque-speed curve that looks a lot like an
> ICE. Also, they run if you just apply a fixed voltage to them; that's
> why even a contactor controller is practical. That's why series motors
> are so popular.
> 
> A PM or shunt motor allows an equally simple controller; all it needs to
> do is provide a DC voltage. You can vary the voltage with contactors or
> a PWM controller to vary the speed. You'd leave the field of a shunt
> motor at full voltage to simulate a PM motor.
> 
> But the torque-speed curve of the PM or shunt motor is considerably
> different than an ICE. It works, but will feel strange to the driver.
> The throttle acts like a cruise control setting; it determines the speed
> regardless of hills. You can add complexity to the controller to make it
> behave in a more familiar fashion; but that's a choice, not a
> requirement. Electric bikes (for example) don't bother.
> 
> Also, the PM or shunt motor with fixed field works best at high rpm and
> low torque; it is not as efficient at low speed and high torque (because
> it requires low voltage and high current). You are more likely to need a
> transmission with a PM or shunt motor.
> 
> A sepex motor is just a shunt motor with independent control of the
> field. You don't have to have two controllers (one for armature and one
> for field), though that gives you the greatest versatility. For example,
> I ran my old aircraft starter-generator (a sepex motor), with just a
> 2-step contactor controller on the armature, and a rheostat in the
> field. This is an extremely simple controller; even simpler than that
> for a series motor -- yet it gave me stepless throttle control, and full
> regen.
> 
> Thus, there is no *need* to tailor the controller to the motor, but you
> may *want* to do it as a matter of finesse. The motor makes the extra
> control capability an option, but it is not mandatory to use it.
> 
> Otmar wrote:
> 
>>I have developed a few sep-ex drive systems for low voltage
>>industrial use. My programmer and I made one on Zilla platform
>>running a sep-ex ADC 9" motor for his Honda Del Sol conversion. It
>>works pretty well, but there are drawbacks.
>>
>>As discussed before, the motor design could use interpoles, which we
>>don't have on ours.
> 
> 
> Lacking interpoles or compensating windings, you want to keep your field
> stronger than your armature. This automatically happens with a PM or
> shunt motor because they are always at full field. Arcing becomes a
> problem only if you try to run high armature current *and* weaken the
> field (low sepex field current) at the same time.
> 
> 
>>In order to do low speed regen in the controller, the high current
>>section that drives the armature needs twice the silicon of a series
>>drive. Then there is the increased complexity of the small H bridge
>>field controller.
> 
> 
> Were you using the extra silicon in the armature controller to do a
> boost converter? If you have sufficient field control, you can just
> connect the armature to the battery thru a big diode (and that diode is
> already there as part of the MOSFETs). This gives you regen at higher
> speeds, but not down to zero speed.
>  
> 
>>Overall I decided that in the volumes we are doing, for the hobbyist
>>market, a AC drive makes more sense. The incremental cost in silicon
>>for AC drive is not that big anymore, and the benefits are numerous.
> 
> 
> I agree if you are making a "high feature" drive (one that totally
> emulates an ICE, runs transmissionless, does regen to zero speed, etc).

Reply via email to