Power-down sequencing with a "hard" power switch (stepper driver)

Question

I'm designing a stepper motor controller and would like to have some advice on sequencing the shutdown process with a "hard" power switch. The Toshiba TB6560 driver that I'm using requires a specific power-up and power-down sequence to avoid damaging the chip (\$V_{DD}\$ \$\rightarrow\$ \$V_{MA}\$/\$V_{MB}\$ \$\rightarrow\$ \$\textit{Control Inputs}\$ on startup and the reverse on shutdown, similar to a FILO buffer). I'm trying to find a good solution for powering the sequencing circuitry, as well as the logic chips it supports, after the main power is removed.

The sequencer is built with discrete components and uses RC timing with op-amps to delay the output lines at specific intervals as shown in the plot below. When the Enable line goes high, the outputs are turned on sequentially; the on/off behavior is staggered such that the first line to go high is the last one to go low. The three outputs from the sequencer are used to drive two MOSFET switches for \$V_{DD}\$ and \$V_{MA}\$/\$V_{MB}\$, as well as a 4066 quad switch for the various \$\textit{Control Inputs}\$.

The schematic for the power supply portion of my stepper motor controller is shown below (click for full-size view). I've marked up the schematic to correspond to the following numbered list:

1. C1 provides energy storage to operate all of the logic circuitry when the power is cut. I calculated the current draw to be 26 mA across the entire controller, excluding the stepper motor itself, so this supplies enough power for about 350 ms of operation after the main switch SW1 is turned off. Based on this timing, the capacitor will discharge to 15 V, which is the minimum input voltage required for the LM7812 regulator U1.

2. R1 is a 10 ohm power resistor for C1 to limit the inrush current to about 2.4 A, which is within the limits of the power supply PS1. From zero volts, C1 will charge in about 100 ms.

3. R2 is a 100 kOhm bleeder resistor to ensure C1 is discharged when the unit is turned off.

4. D1 is a low-forward-voltage Shottky diode to protect the output of PS1 when C1 takes over during shutdown. The forward voltage drop should be 0.5 V or less in this application, so my "+24V" rail will actually be around 23.5 V, which is fine for the stepper motor and downstream regulators.

5. The coils for relays K1 and K2 are operated via SW1 on the AC side. K1 toggles the Enable line on the sequencer X1, and K2 disconnects the stepper motor coils at shutdown to limit the current requirements (i.e., without this cutoff, if the device was switched off when the motor was drawing high current, C1 would drain quickly and kill the circuit before X1 had a chance to shut everything down properly).

6. X1 is shown as a block diagram element for the sake of brevity. I don't think its underlying schematic is important to answer my questions, but I can post more details if necessary.

Here are my questions:

• Is my implementation of C1 an appropriate way to ensure the circuit has power after the main switch SW1 is flipped open?

• If so, is the configuration of R1, R2, and D1 acceptable to limit the inrush current and protect PS1? I looked into using an NTC thermistor in lieu of R1, but I was concerned that the thermistor might not recover fast enough to avoid overloading the supply if SW1 was quickly cycled.

• Assuming the proper relays are used, is it acceptable to have an AC input (SW1) operate DC circuits via K1 and K2? I've seen this configuration in some industrial equipment, but I wasn't sure how common it was.

• Are flyback diodes necessary when K2 switches open the motor coils of M1, or are the snubbers on the H-bridge in the stepper driver U3 sufficient to limit the back EMF?

I'm also open to any other suggestions or best practices for a circuit like this.

Answer 1

I have simple recommendation to make the energy stored in your C1 capacitor more useful. Stop using 7812 and 7805 for U1 and U2. Use a switcher instead that can provide the needed voltages over a broader range. Even better if those could be buck/boost style combo mode switchers.

Alternately you could look at a better PS1 solution that supplies all three rails or multiple PS1's where PS1A provides the 24V and PS1B provides the 12V and 5V or an additional PS1C that provides the 5V.