Is this a reasonable circuit for locked antiphase PWM?

Question

I'm building a motor controller for a robot that will have a very limited number of IO pins available. I need to drive two motors from only 3 pins, and Locked Antiphase PWM, where the motor is alternately driven forward and backwards by the H-Bridge, with duty cycle determining the net current, looks very attractive. The fact that it 'just works' when the robot is on a slope (rather than accelerating down it when given a small forward momentum) is an added bonus.

Does this circuit, utilizing the SN754410 H-Bridge and a couple of transistors used as inverters, look reasonable? Am I doing anything obviously wrong here?

Also, a couple of auxiliary questions:

• In 'regenerative braking' mode, the motor may drive current back out the H-Bridge. If I'm powering the whole circuit off a (non-rechargeable) battery, is this liable to damage the battery? If so, can I protect against it with diodes?
• How can I determine what frequency of PWM I'll need to avoid significant idle current draw with Locked-Antiphase and a given motor, other than experimentally?

Answer 1

Rushing ...

Antiphase PWM per motor looks OK.

Disabling IC using EN lines will often create large motor spike - as long as internal IC catch diode rating is OK that's OK. Otherwise add external diodes to supplies.

---

What battery chemistry. What size?

AA Alkaline will often tolerate significant regeneration.

Unless battery is small physically or you want exceptional shelf life then NimH will be better than Alkaline. For very long shelf life LSD NimH are getting very good.

At AA size, energy density of NimH now equals or exceeds Alkaline and high load dishcharge characteristics are much superior so there is little reason for Alkaline.
At small cell size Alkaline MAY make sense.

---

I have not tried this
Answer is "seems like this to me".
Caveat emptor :-):

Motor inductance and resistance will set t = L/R time constant. But, very easy to put small resistor from motor to ground, apply voltage and watch voltage in resistor and thus current rise on scope. PWM period which is small compared to total motor rise time will (probably) be good.