Hardware Optimisation for BOOSTXL-DRV8301

"I am getting the same performance level from InstaSPIN as my other ESC. I followed the lab guides, and have done multiple tests but I don't notice any major differences in efficiency at all. [I used a Watt meter to calculate the input power, and I calculated the thrust generated from the rotor blades to work out the output power.] I read that many users found success in boosting overall efficiency by ~20% using InstaSPIN. "

Are you calculating efficiency only at maximum torque?
If so, you won't see any differences. In fact, a block commutation ESC will produce more torque than InstaSPIN-FOC (if using default 1.0 modulation) at full torque. This is simple phsyics. The block commutation switching can provide more average voltage to the phases hence more current hence more power.

Where FOC improves efficiency is all the times when you are NOT needing to produce maximum torque. FOC is minimum current for necessary torque. Block commutation is using more voltage than necessary at many of the other times, especially during accel/decal and load changes. So it is only over the full use case of a dynamic application that you will see efficiency gains. Just looking at power out / power in at maximum power usage is not showing you the full picture.


Changing the hardware will not really have any perceivable effect on this topic. However, I should comment that your voltage filter pole is incorrect (the User Guide is not accurate on this topic).

We do not need nor want the voltage filter pole > maximum frequency of the motor. For high speed motors this pole becomes too high and you start seeing more channel to channel error caused by the HW filter component drift. It is better to keep this pole < 500 Hz.

The only stipulation is that the USER_IQ_FULL_SCALE_FREQUENCY_Hz variable must always remain < 4 * this filter pole.

So if you need to set this variable to 1600 Hz you would want to keep this filter pole > 400 Hz (+ some buffer).

Changing the current and voltage scaling really isn't going to buy you much at this point, but you can choose to do this on your own HW when you design. I didn't check your math, but it looks like you understand the equations.

ChrisClearman said:

"I am getting the same performance level from InstaSPIN as my other ESC. I followed the lab guides, and have done multiple tests but I don't notice any major differences in efficiency at all. [I used a Watt meter to calculate the input power, and I calculated the thrust generated from the rotor blades to work out the output power.] I read that many users found success in boosting overall efficiency by ~20% using InstaSPIN. "

Are you calculating efficiency only at maximum torque?
If so, you won't see any differences. In fact, a block commutation ESC will produce more torque than InstaSPIN-FOC (if using default 1.0 modulation) at full torque. This is simple physics. The block commutation switching can provide more average voltage to the phases hence more current hence more power.

Where FOC improves efficiency is all the times when you are NOT needing to produce maximum torque. FOC is minimum current for necessary torque. Block commutation is using more voltage than necessary at many of the other times, especially during accel/decal and load changes. So it is only over the full use case of a dynamic application that you will see efficiency gains. Just looking at power out / power in at maximum power usage is not showing you the full picture.

I calculated the power output and power input at different RPM speed (1000 to 9000 RPM, which is the max speed), both ESC and InstaSPIN-FOC have the same power consumption across the speed range. Does that mean I haven't tuned my InstaSPIN-FOC well enough to reduce the current and voltage needed for necessary torque?

ChrisClearman said:

Changing the hardware will not really have any perceivable effect on this topic. However, I should comment that your voltage filter pole is incorrect (the User Guide is not accurate on this topic).

We do not need nor want the voltage filter pole > maximum frequency of the motor. For high speed motors this pole becomes too high and you start seeing more channel to channel error caused by the HW filter component drift. It is better to keep this pole < 500 Hz.

The only stipulation is that the USER_IQ_FULL_SCALE_FREQUENCY_Hz variable must always remain < 4 * this filter pole.

So if you need to set this variable to 1600 Hz you would want to keep this filter pole > 400 Hz (+ some buffer).

My USER_IQ_FULL_SCALE_FREQUENCY_Hz variable is 1050 Hz, so I guess I should keep the default voltage filter pole value 364.682 Hz. Thanks.

"I calculated the power output and power input at different RPM speed (1000 to 9000 RPM, which is the max speed), both ESC and InstaSPIN-FOC have the same power consumption across the speed range. Does that mean I haven't tuned my InstaSPIN-FOC well enough to reduce the current and voltage needed for necessary torque? "

you are making these calculations at steady state speed and load I assume? With a well tuned BLDC controller you can certainly get equivalent efficiency at a steady state. There could be things that could be slightly better tuned on the InstaSPIN-FOC solution, specifically insuring the motor parameters are accurate and the speed loop is properly tuned. But overall this doesn't surprise me.

It is when you start looking at the overall power use in a full application that you should start to see FOC make small gains (5-20% is what is reported).

There are also the other advantages of lower peak currents and better dynamic accel/decal...but depends if your application really benefits from that or not.

For a hobby style ESC it would be pretty hard to justify InstaSPIN-FOC for a very low cost "consumer" style application. For an industrial / military / prosumer application it could make sense.