not enough torque

[Nontinium]

Good morning , I'm writing this post to request help , the main problem is two things the first and the main thing being the 1000W 3500 RPM 3 pole pair 48V motor not giving it rated torque of 2.5 Nm and only gives 0.8 N.m , this was confirmed as the motor works fine with a premade controller of the motor , further more the motor gets stuck and the torque reference goes to 19.99A however this does not happen with the premade controller , the motor works fine with the stm controller when no load is applied, the motor has been already profiled using the smt32 motor profiler, and motor barely gave any errors on the stm32 software however almost minimal torque, also as we are using a data acquistion system that is extremly accurate it measure 2A DC from the battery of 48V however the stm32 motor control software's torque reference showed 8.6 A , and the hall sensors are slight off , so at 370 RPM in the stm software its actually 500 and at 740 is 1000 and at 1110 its 1500 and soo on.

the reason we are trying to use the stm32 motor control software is to increase the efficiency 

the motor is MY1020 BLDC, is hard to find its datasheet , our is silver color if it helps, take 3 pole pairs with grain of salt even thou we made sure it was 3.

[Nontinium]

also one more thing few months early on a different motor we encounter the exact same issue of the torque issue not the hall sensor issue, however the torque in the six step algo is significantly higher than foc 

[STuser2]

FOC will have more torque compared to six step.

[Nontinium]

Thankyou for replying, Indeed it will have more torque however that's the issue sadly I'm having trouble figuring out why does it not, it's actually severely weaker, I have been on this issue for 4 months

[Nontinium]

We tried several different motors , and made sure that the parameters are the same , the exact behavior of all these motors are this it has a weak torque initial like non , like it won't move and hence the torque reference goes to 20A then I am forced to give the motor a push them the motor will that moving, also the torque build up slowly to a 1/4 of the actual value , so the PI gains can be adjusted to make the torque I crease faster how the MX torque is still 1/4 of the actual value and worse than six step

[GMA]

Hello @Nontinium,

This expected-to-measured speed ratio indicates that the parameters are not set correctly.
The Motor Profiler tool provides a feature to "Detect pole pairs" parameter and to perform Hall sensor profiling.
Refer to the Motor Profiler application note available through Workbench tool > About > Documentations > Documentation > User manual.

[Nontinium · Accepted solution]

Thankyou kind Sir, actually I solved this issue few days after opening this post just at my deadline, the issue was the current amplification part, we used EVALSTDRIVE101 to inspire our personal motor controller design however the current sensing part was changed from an amplification of 6.49 to 4.11 In addition I did not understand how to use the J5, J6 and J7 jumper which switched between raw current values and amplified current values, after realizing this issue a well documented and test proven BLDC motor was used and the torque was perfect the motor had a wattage draw of 311W at 48V and gave 1 Nm of torque perfectly, fixing both the overall torque issue and the startup torque issue, now only the PI gains are left for tunning.

if feasible, please update the software of the FOC to display the current drawn from the battery as i used the six step current measurement to debug the FOC current issue as the six step had a better current measurement indicators than FOC. 

Thankyou for responding kind sir.