L293 H bridge leakage?

This is probably the wrong place for this but maybe someone has seen it before?
Should I get 4V (ish) on the motor with the motor supply disconnected?
I have (half) a L293D Quad 1/2 H bridge running a 12V DC motor. My Pro Mini is fed from 5V USB/FTDI supply and is also powering the L293D logic power (VCC1, 1&2ENable and 3&4ENnable Chip pins 1,9 and 16). All grounds are common. A separate 12V supply powers the motor via chip pin 8 - VCC2 . Arduino pins 9 and 10 feed the inputs (10 & 15) on the H bridge. The outputs (11&14) connect straight to the motor. I'm only using half of the chip (bridges 3&4)
With the motor supply disconnected I'm getting 4V on VCC2 - and the motor is trying to turn in accordance with the PWM from the Arduino. Is this normal?
things I've done or noticed so far:
H bridge gets quite hot.
The arduino cannot see the 12V when it is connected to pin 8
Thoroughly checked the wiring -all seems ok.
4v appears on pin 8 with that pin out of its IC socket.
The 4V on pin 8 will stay at 4V with a 470 ohm load to GND

Here's the datasheet

Your link goes to an L293. I've no idea about the difference compared with an L296.

Here's a link ti a L296 data sheet.
The pin 8 is ground but You measure voltage. Have You forgotten to use a common GND for the logic of the bridge and the controller?

sorry. It's a L293D NOT a 296.
296 is handwritten on the pack but 293 is stamped on the chip.
Well observed!
I've edited the post.

Okey. Second round. Reading section 6.3 Vcc2 must be between Vcc1 and 36 volts. Leaving it open, for tests, measurements is not within its range. That might explain the strange 4 volt.
How do You make sure that the current passing the H-bridge is within limits?
What is the motor data? If the motor current exceeds the capacity of the H-bridge heating, maybe smoke is evident.
Give us the motor data.That bridge, the L293D only handles up to 600 mA if I read the data sheet properly.

The strong 5V USB supply may boil the chip and turn the motor. Nobody forbids the Vcc1 also going to Vcc2 (pin 8 ), perhaps reduced by a diode forward voltage to 4V.

The motor is small 1" can motor salvaged from the scanner of a multifunction printer. It has an attached shaft encoder I'm using to determine the position of a direct driven 3mm screw drive slider. The load current is around 100mA the stall current is 250mA. The driven PWM won't exceed 50% . I suspect the motor is actually rated at 24 volts. It's part of a model railway animation. I noticed this phenomenon whilst testing the encoder reading part of the sketch as the motor was resisting me from manually turning the motor shaft despite the motor power being disconnected. This resisting of movement caused the motor to 'hunt' with the encoder changes. My sketch Ramps up the motor PWM to overcome the starting friction. I think it was this or the DC braking effect that was causing the heating if the chip. Not super hot, but enough to raise an eyebrow.

The dinosaur L293 chip burns up to 4V of the motor supply, depending on output current, and thus can become very hot easily. Nowadays you better use MOSFET bridges with near zero power waste even with high output currents.

Can you please post a copy of your circuit, in CAD or a picture of a hand drawn circuit in jpg, png?

Can you post a picture of your project so we can see your component layout?

A couple of pictures/diagrams are worth more than a thousand words/

Thanks.. Tom... :slight_smile:

2 pics attached

OPs Pics;

Circuit edited.

Tom... :slight_smile:

That is one serious piece of railroad!

I would be very suspicious of the ability of that shaft encoder to accurately determine the correct positions of the trolley.
This is a job for a stepper motor and some optical encoders.

Hi Paul.

I only included motor wiring on the diagram. There is also a motor shaft encoder on pins 2&3, a zero position microswitch on pin 4 and other stuff too. If you look look closely behind the motor in the photo you can see the shaft encoder. I'm not familiar with using optical encoder to determine stepper motor position.

This is only a small part of this animation. A master pro mini looks after the control commands for 25 motors and some sensors and sends and requests data over I2C to 5 slave pro minis which drive the motors and read sensors. I considered using a mega2560 but found it easier to do it is smaller bites :smiley: .

The whole thing has got out of hand, but it's keeping me safely isolated :o

I chose this arrangement because I needed to keep the direct drive screw diameter thin (3mm) to allow the moving under-track magnets to be as close to the steel wagon wheels as possible. This means about 250 turns for full travel which would ideally complete in under 5 seconds or 3000 RPM. I would need a pretty fast stepper motor to do this. I've noticed that the encoder sometimes misses a step at very slow rpm but I can automatically re-calibrate every 4th move and at a resolution of 20000 steps over 120mm travel, a loss of a few steps isn't going to make any significant difference. (0.006mm per step - I think)