marcoski
15-11-2003, 08:31 PM
Just thought I would write a story on how the novak brushless system works.
Interestingly enough some of the control theories that I thought were in use are in fact not after I gave the unit a thourough going over with the oscilloscope as part of the research for this piece.
Its construction.
The motor comprises of a magnetic rotor surrounded by three stationary field coils. This is completely opposite to brushed where the coils rotate with the armature and the magnets are stationary. The rotor itself is a neodymium magnet arranged to have six poles around the outer edge. The field coils are wound on shaped soft iron pole pieces securely mounted to the motors can. On the rear of the motor are three hall probes. These are special electronic sensors that pick up magnetic fields. They are fixed in such a way that they can see the 6 magnets of the rotor, and there is one for each field coil. They are also set in relation to each field coil to set the motor timing.
Commutation.
This is quite complex! The principle of the motor is that if you switch on each field coil one after another, you will create a magnetic field that will rotate around the can from coil 1 to coil 2 to coil 3 and back to coil 1 again, ad infinitum. The rotor, being a magnet, will follow this field round and rotate. However, that method is not very efficient, so the coils are driven both ways, first to attract a north pole on the rotor, then reversed to attract a south one. But it gets even worse than that!! As its a three wire system, the interactions between the three drives becomes quite complex, but suffice to say it generates a very efficient rotating field. The timing of this complex drive is determined by the signals from the hall probes, automatically selecting the next step in the firing sequence as each probe sees a new segment of the rotor.
So half of the speed controller is devoted to this automatic sequenced commutation system
The other half is a conventional speedo as we know it. It just varys the drive voltage to the autocommutation system and motor just like a brushed motor - the more volts you give it, the faster it will turn.
The maximum speed is not limited by setting a ceiling on the commutation frequency, as I initially thought, rather, just like a brushed motor, equilibrium is reached between back emf and coil drive. So it is true to say that higher voltage cells will give a slightly high maximum RPM.
Acceleration is only limited by the field coils ability to accept current, the more current that can be delivered to the coils, the faster it will accelerate. So again, higher voltage cells should allow for faster acceleration.
Braking - this works by shorting the coils out.The dynamo effect of the magnetic rotor will cause the rotors energy to be lost as heat in the shorted windings, and the motor will soon stop.
In unlimited mode, all the above is true, the motor will get everything the cells can deliver. But there is the stock motor mode as well.
Stock mode.
In stock mode there is a ceiling on the commutation frequency, the motor simply will not go faster than 24000 rpm. This limit is crystal controlled, it is very accurate and repeatable.
There is also a current limiter to limit the acceleration, the maximum field coil current is electronically controlled to limit how fast the motor accelerates. This is subject to a wider production spread, but then so are stock motors! I would say given the method used, it would be fairly accurate across a field of novaks.
In stock mode the quality of the cells is less important, as even poor cells will allow the electronic governors to work well within their limits.
The motor supplied has the following charactoristics:
Power 196 watts
Kv 5800rpm/volt
Kt 0.45 Inch-Ounce/Amp
The speedo is capable of driving a 225 watt motor, and has locked rotor detection to prevent damage by trying to turn a stalled motor (car against the barrier, say)
Interestingly enough some of the control theories that I thought were in use are in fact not after I gave the unit a thourough going over with the oscilloscope as part of the research for this piece.
Its construction.
The motor comprises of a magnetic rotor surrounded by three stationary field coils. This is completely opposite to brushed where the coils rotate with the armature and the magnets are stationary. The rotor itself is a neodymium magnet arranged to have six poles around the outer edge. The field coils are wound on shaped soft iron pole pieces securely mounted to the motors can. On the rear of the motor are three hall probes. These are special electronic sensors that pick up magnetic fields. They are fixed in such a way that they can see the 6 magnets of the rotor, and there is one for each field coil. They are also set in relation to each field coil to set the motor timing.
Commutation.
This is quite complex! The principle of the motor is that if you switch on each field coil one after another, you will create a magnetic field that will rotate around the can from coil 1 to coil 2 to coil 3 and back to coil 1 again, ad infinitum. The rotor, being a magnet, will follow this field round and rotate. However, that method is not very efficient, so the coils are driven both ways, first to attract a north pole on the rotor, then reversed to attract a south one. But it gets even worse than that!! As its a three wire system, the interactions between the three drives becomes quite complex, but suffice to say it generates a very efficient rotating field. The timing of this complex drive is determined by the signals from the hall probes, automatically selecting the next step in the firing sequence as each probe sees a new segment of the rotor.
So half of the speed controller is devoted to this automatic sequenced commutation system
The other half is a conventional speedo as we know it. It just varys the drive voltage to the autocommutation system and motor just like a brushed motor - the more volts you give it, the faster it will turn.
The maximum speed is not limited by setting a ceiling on the commutation frequency, as I initially thought, rather, just like a brushed motor, equilibrium is reached between back emf and coil drive. So it is true to say that higher voltage cells will give a slightly high maximum RPM.
Acceleration is only limited by the field coils ability to accept current, the more current that can be delivered to the coils, the faster it will accelerate. So again, higher voltage cells should allow for faster acceleration.
Braking - this works by shorting the coils out.The dynamo effect of the magnetic rotor will cause the rotors energy to be lost as heat in the shorted windings, and the motor will soon stop.
In unlimited mode, all the above is true, the motor will get everything the cells can deliver. But there is the stock motor mode as well.
Stock mode.
In stock mode there is a ceiling on the commutation frequency, the motor simply will not go faster than 24000 rpm. This limit is crystal controlled, it is very accurate and repeatable.
There is also a current limiter to limit the acceleration, the maximum field coil current is electronically controlled to limit how fast the motor accelerates. This is subject to a wider production spread, but then so are stock motors! I would say given the method used, it would be fairly accurate across a field of novaks.
In stock mode the quality of the cells is less important, as even poor cells will allow the electronic governors to work well within their limits.
The motor supplied has the following charactoristics:
Power 196 watts
Kv 5800rpm/volt
Kt 0.45 Inch-Ounce/Amp
The speedo is capable of driving a 225 watt motor, and has locked rotor detection to prevent damage by trying to turn a stalled motor (car against the barrier, say)