Hi guys
I am building a beach buggy conversion, using an induction motor - with a catch though - I've programmed my own motor controller.
The car is a simple Beetle chassis with fibreglass body.
I am going to use a 15kW 220V 3ph induction motor (alluminium - 96kg - would be ideal but I've got a cast iron one at hand, weighing 130kg, double the weight of a 1600cc beetle motor).
I will be linking up 26 lead-acid batteries in series for a bus voltage of 310V or so. (For 220V motor - 220x\sqrt{3} with SVPWM)
I am going to mount the motor directly to the gearbox, where I can easily use 2nd or 3rd gear to obtain my reduction ratio of about 8:1, and can lock it and chuck out the extra gears/selectors to free up weight.
My first issue:
Is this weight too much, especially if it is going to be hanging behind the rear axle. I am not sure that batteries in the front where the fuel tank was ie. by the steering column, will balance the weight out sufficiently. How will this handle turning, having such a large moment of inertia for yaw (<---correct word?) rotation.
Otherwise I can try turn the gearbox around to have the motor midmounted - but this will require lots of modification, but may be worth it.
Onto software/inverter issues...
I have played around with a sample Inf HybridPack 1 but my setup I am using is more 'rustic'.
I am using IGBTs (SKM100GB 123D)from Semikron with a SKHI61 driver, setup with laminated busbars and DC bus capacitors for cancelling out the inductance of the long battery cables.
I am using a DSP (F2812 from TI, 32bit fixed point) for my program. I have written the software using simple field orientated control, where I specify setpoints for the torque and flux. It gets a speed measurement from a rotary encoder and uses the current measurement to predict the rotor flux position.
It is working nicely in testing, but I have to manually set the flux - has anyone come up with clever ways of setting the flux depending on the torque required. It seems inefficient to just set it to nominal flux when you start the car and then just adjust the torque setpoint.
That done I just need to worry about adding a transducer so that the torque setpoint is controlled by the pedal (should I use a potentiometer? A resolver seems more ideal given the repeated use but quite pricey)
Battery weighting issue:
Will it be better for performance to have batteries mounted on the sides of the car/buggy, increasing the cars "lateral/rolling" moment of inertia, or mounted near the centre axis in the front bonnet.
Battery charging issue:
I will be linking up 26 lead-acid batteries in series for a bus voltage of 310V or so. As I am going to program regen braking into my algorithm, are there any concerns for charge balancing of these?
(I am thinking there will only be problems when the batteries are full and i need to float charge them, as this is when some can be overcharged. Maybe I should only allow limited regen when batteries are quite low for now)
I am building a beach buggy conversion, using an induction motor - with a catch though - I've programmed my own motor controller.
The car is a simple Beetle chassis with fibreglass body.
I am going to use a 15kW 220V 3ph induction motor (alluminium - 96kg - would be ideal but I've got a cast iron one at hand, weighing 130kg, double the weight of a 1600cc beetle motor).
I will be linking up 26 lead-acid batteries in series for a bus voltage of 310V or so. (For 220V motor - 220x\sqrt{3} with SVPWM)
I am going to mount the motor directly to the gearbox, where I can easily use 2nd or 3rd gear to obtain my reduction ratio of about 8:1, and can lock it and chuck out the extra gears/selectors to free up weight.
My first issue:
Is this weight too much, especially if it is going to be hanging behind the rear axle. I am not sure that batteries in the front where the fuel tank was ie. by the steering column, will balance the weight out sufficiently. How will this handle turning, having such a large moment of inertia for yaw (<---correct word?) rotation.
Otherwise I can try turn the gearbox around to have the motor midmounted - but this will require lots of modification, but may be worth it.
Onto software/inverter issues...
I have played around with a sample Inf HybridPack 1 but my setup I am using is more 'rustic'.
I am using IGBTs (SKM100GB 123D)from Semikron with a SKHI61 driver, setup with laminated busbars and DC bus capacitors for cancelling out the inductance of the long battery cables.
I am using a DSP (F2812 from TI, 32bit fixed point) for my program. I have written the software using simple field orientated control, where I specify setpoints for the torque and flux. It gets a speed measurement from a rotary encoder and uses the current measurement to predict the rotor flux position.
It is working nicely in testing, but I have to manually set the flux - has anyone come up with clever ways of setting the flux depending on the torque required. It seems inefficient to just set it to nominal flux when you start the car and then just adjust the torque setpoint.
That done I just need to worry about adding a transducer so that the torque setpoint is controlled by the pedal (should I use a potentiometer? A resolver seems more ideal given the repeated use but quite pricey)
Battery weighting issue:
Will it be better for performance to have batteries mounted on the sides of the car/buggy, increasing the cars "lateral/rolling" moment of inertia, or mounted near the centre axis in the front bonnet.
Battery charging issue:
I will be linking up 26 lead-acid batteries in series for a bus voltage of 310V or so. As I am going to program regen braking into my algorithm, are there any concerns for charge balancing of these?
(I am thinking there will only be problems when the batteries are full and i need to float charge them, as this is when some can be overcharged. Maybe I should only allow limited regen when batteries are quite low for now)
