0
\$\begingroup\$

I am attempting to use a DC-DC boost converter to step up a 7.2V battery to 24V to run a stepper motor, however, I am unsuccessful in doing so.

I have tested my stepper using a DC power supply set at 24V, and it runs beautifully. Yet the moment that I hook up the converter, my motor begins to skip steps left and right,even at very low speeds. I am watching the voltage output from my converter the whole time, and the needle never budges from 24V. I have the recommended electrolytic caps on both of them.

I did notice that the motor ran better (skipped less steps) when I put small ceramic decoupling caps across the motor driver. I suspect that interfering frequencies from the boost converter output and the chopper drive are causing my problems.

The components are an A4988 stepper driver on the breakout board from Pololu set at its max current of 2A and a LM2588 boost converter from TI, using a schematic designed by TI's Webbench software.

\$\endgroup\$
4
  • \$\begingroup\$ Do you have access to an oscilloscope? \$\endgroup\$
    – Tom Carpenter
    Commented Mar 10, 2016 at 1:09
  • \$\begingroup\$ @TomCarpenter No. \$\endgroup\$
    – dpdt
    Commented Mar 10, 2016 at 1:16
  • \$\begingroup\$ Is the boost converter + battery capable of an input current of ~8A? \$\endgroup\$
    – Tom Carpenter
    Commented Mar 10, 2016 at 2:22
  • \$\begingroup\$ Answer: no. Webench says current exceeds design limits. \$\endgroup\$
    – Tom Carpenter
    Commented Mar 10, 2016 at 2:25

2 Answers 2

Reset to default
1
\$\begingroup\$

The stepper motor will act as an alternator when spinning and generate a voltage opposite to the drive voltage proportional to the angular velocity. To counter this voltage known as back-electromotive force (back-EMF) and keep the torque, which is proportional to the current, high, the driver ramps up the voltage as the motor speed increases. To allow for this voltage headroom, the motor is designed to have a winding resistance lower than what would allow only the nominal current (e.g. 1A) to flow at rest when given 24 V.

The pololu driver is fairly autonomous and actually has current feedback to counter this: it can adjust the motor voltage (via pulse-width modulation) in order to keep the current within limits at any speed. The board will function with motor currents up to 2 A per phase, so 4 A total.

If you set the driver to its maximum current of 2 A per phase (as you stated) and the winding resistance is 12 ohms or less, it will draw 4 A at 24 V DC (96 W!) in the worst case. Powered from a 6.7 V (7.4 V - BJT forward voltage) source, the boost converter low side switch would have to conduct a peak current of 4*(24/6.7)*2 = 28.7 A. The low side transistor of the LM2588 boost converter IC has an absolute maximum current rating of just 5A, so the 24 V will most likely brown out immediately when the motor is driven.

You can set the current setpoint by adjusting the little trimmer pot on the PCB. You should set the current limit to 400 mA per phase or less, or you will operate the boost converter out of specification.

\$\endgroup\$
2
  • \$\begingroup\$ I think you forgot that the driver board will act as a step-down: my boards draw about 0.5A at a 2A current limit. Considering that I got it to work at a 2A current limit, this can't be right... \$\endgroup\$
    – dpdt
    Commented Mar 27, 2016 at 17:51
  • \$\begingroup\$ Indeed the driver will draw much less than 2 A from the 24 V supply to provide 2 A of winding current at low rpm, but when you increase the motor speed the driver will have to raise the effective motor voltage (by increasing the duty cycle) in order to keep the winding current at 2A per phase. Eventually it will overload the boost converter, unless you keep the RPM very low. \$\endgroup\$
    – jms
    Commented Mar 27, 2016 at 18:22
1
\$\begingroup\$

It was a noise coupling issue: I put a 0.033uF ceramic across the logic input to the board and it worked like a dream. The noisy converters must have been messing with the current sensing as I wired all the grounds together.

\$\endgroup\$
1
  • \$\begingroup\$ What do you mean by putting a ceramic across the logic input? I'm encountering the same problem right now. Did you solder the ceramic from vdd to gnd? Thanks in advance! \$\endgroup\$
    – Georg Milecki
    Commented Jan 10 at 12:31

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.