0
\$\begingroup\$

I am currently trying to set up a rig to test my 3d printed gears under different torques and motor speeds. The power input side is going to be the BLDC motor, driven by a hobby ESC (BLHeli_32), powered by a generic AC to 12V PSU. On the braking side of the gears I am reluctant to wiring up a 3 PWM-controlled resistors across the "braking motor" as it would be very wasteful of energy and produce heat, as I intended to run the motor at around 100W for many hours or days.

I am instead considering using a full wave three-phase bridge rectifier and connect the DC outputs back into the DC-in circuit (still with PWM MOSFETs to control torque). What starts confusing me is that, under the assumption of ideal, identical, lossless motors, the output voltage would be sqrt(3)*V_in (minus diode losses). Something feels confusing about it as the voltage would exceed supply voltage, and I can't figure how the system would reach equilibrium. Is this system possible? Would I need to use significantly lower kV motor on the output side? Would I need to put a diode or flyback diode to protect the PSU against backflow of electricity when the input motor throttles down?

I've schemed loosely what I think purely-electrically this circuit should look (bar the PWM on output), but may have made mistakes as I have very little electronics knowledge (aerospace engineer).

Hand-drawn schematic

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
2
  • \$\begingroup\$ This is a total mess, sorry I do not understand nothing about your intentions, but this diagram is 100% not going to work. \$\endgroup\$ May 28, 2020 at 10:10
  • \$\begingroup\$ Added hand-drawn schematic, sorry. I understand the confusion, but perceived two linked motors with different kV as effectively a transformer per phase \$\endgroup\$
    – GSammons
    May 28, 2020 at 10:32

2 Answers 2

0
\$\begingroup\$

1st note: The BL_Heli ESC won't last for long, since it is meant for RC planes, cars, etc.. they don't work at nominal power all the time, aka S1 duty.

2nd note: with this circuit you won't be able to control the torque. Only when generator voltage will become higher than battery voltage, the generator will produce a negative torque.

As said, this would reach the equilibrium when generator would produce higher voltage than supply voltage, and in a very short window it will freely run and in a very small rpm increase it will prevent further speed increase. Of course the two motors may not have the same characteristics to achieve this.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks. I will use a reasonably overspec'ed ESC (50A) ESC for the purpose to avoid overheating. I didn't draw in the PWM circuit to simplify the question, but with PWM + a much lower kV generator motor would I not gain torque control? \$\endgroup\$
    – GSammons
    May 28, 2020 at 14:35
0
\$\begingroup\$

The system described is called a "four square test rig" or "four square dynamometer." It is a scheme that has been around for a long time and may be known by other names. It can be implemented mechanically with a continuously variable transmission or with a regenerative dynamometer.

As pointed out by @Marko Buršič, with no adjustment provided on the power return bridge, operation is limited to a limited range of speed and load. For a wider range of adjustment, the power return bridge needs to be a regenerative PWM motor control rather than a simple rectifier. You will be able to find demonstrations of systems like this on YouTube.

\$\endgroup\$
1
  • \$\begingroup\$ I didn't know the name (thank you, will do research on it now) but was aware of the mechanical dyno, but was keen to use an electrical recirculation over a torsional one as it provides easy transient torque control and could potentially scale to the whole rear end of my RC 3d printed differential, with a generator per rear wheel \$\endgroup\$
    – GSammons
    May 28, 2020 at 14:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.