Has anyone had any experiences with combined negative voltage and positive current feedback amplification?
Here's the basic design : active crossover, normal voltage feedback amplifier directly connected to driver. A current sense element on the return from the load sends a small amount of positive feedback into the amplifier.
The ideal driver has 0 resistance 0 inductance and 0 capacitance.
That would mean that if the ideal driver is not moving at exactly the right velocity, the back EMF would not match the drive voltage and the ideal driver would draw up to an infinite amount of current to match the driving voltage. perfect tracking.
Unfortunately real life drivers are not perfect and can be approximated to an ideal driver with a series resistor and a series inductor.
If we were to compensate for the volt drop across the series elements, we would once again have the behaviour of an ideal driver. This compensation can be done by having a small current sense resistor (0,1 ohm?) in series with the driver, just as it returns to ground. A small inductor can also be in placed in series if compensation for the driver inductance is also to be achieved. This signal is then added to the incoming signal (positive feedback).
Total compensation is not realistic as the driver's resistance would change with temperature and perfect compensation happens at the edge of oscillation. But partial compensation is realistic and can be a simple mod to an active speaker.
If we can add a percentage of the voltage across the driver's internal resistance to the output signal, we can effectively reduce the driver's resistance. Assuming we add 75%, we will be increasing the damping currents by 400%. Inductance compensation needs to keep the L R ratio of the driver and the current sensor.
I've never seen anything like it, I've never built anything like it, I think the theory holds, The active crossover and amp parts are on order. Any experience or input will be welcomed.
Murray.
Here's the basic design : active crossover, normal voltage feedback amplifier directly connected to driver. A current sense element on the return from the load sends a small amount of positive feedback into the amplifier.
The ideal driver has 0 resistance 0 inductance and 0 capacitance.
That would mean that if the ideal driver is not moving at exactly the right velocity, the back EMF would not match the drive voltage and the ideal driver would draw up to an infinite amount of current to match the driving voltage. perfect tracking.
Unfortunately real life drivers are not perfect and can be approximated to an ideal driver with a series resistor and a series inductor.
If we were to compensate for the volt drop across the series elements, we would once again have the behaviour of an ideal driver. This compensation can be done by having a small current sense resistor (0,1 ohm?) in series with the driver, just as it returns to ground. A small inductor can also be in placed in series if compensation for the driver inductance is also to be achieved. This signal is then added to the incoming signal (positive feedback).
Total compensation is not realistic as the driver's resistance would change with temperature and perfect compensation happens at the edge of oscillation. But partial compensation is realistic and can be a simple mod to an active speaker.
If we can add a percentage of the voltage across the driver's internal resistance to the output signal, we can effectively reduce the driver's resistance. Assuming we add 75%, we will be increasing the damping currents by 400%. Inductance compensation needs to keep the L R ratio of the driver and the current sensor.
I've never seen anything like it, I've never built anything like it, I think the theory holds, The active crossover and amp parts are on order. Any experience or input will be welcomed.
Murray.
