I'm making a small wall avoiding robot (very fun project :-))
I'm using the L293NE Quadruple Half Driver to control the direction of the motors, the one at the front is a stepper for left/right, and the one at the back is to drive it forwards and backwards. I've wired it up like this (except with both sides of the chip being used, and no switch): http://itp.nyu.edu/physcomp/uploads/dcmotor_bb2.png
The problem is that the driver chip is getting incredibly hot after a little bit of usage - I haven't got a heat sink nor any space on the protoboard for one unless I buy one and place it on top of the chip itself rather than using the pins labelled heatsink.
Is there another chip that I could use in place that wouldn't need its own heatsink (maybe the SN754410) or could would limiting the current going through the driver be the way forward?
I'm not sure on the specs of the motor - I salvaged a remote control car so don't have the specs for them, although they were being driven by 2x 3.5v AA batteries previously.
There's a bit of a smell coming from the chip, and it's very (very) hot to the touch - don't know if this means it's outside of its operating temperature range as you say though. Having said that - the motor does seem to be losing power as the chip heats.
Then it is vital you find out - measure it with a meter.
If it is too hot to keep your finger on then it is too hot.
This is caused by too much current. These parts use transistors inside so look for parts that use FETs or use FETs by themselves. But until you know the current you can't make any sort of design decisions.
How would measuring the current tell me the operating specs of the motor?
Agree on the temperature of the HBridge - it probably is excess current causing it to heat. You can buy a heatsink for it which might be a route to investigate.
Are you suggesting using a FET instead of the HBridge? I've seen a lot of tutorials and a Uni lab page that suggests these chips are more than capable of driving small motors...?
Your motor can also be in the point of its lifetime, that it is going to hang, cut, stall, what ever the term would be in english. It's slowly become stucked, current rises, smell turns to smoke in three places, guess where...
If you test the motor spinning free by hand, does it turn easily, without resisting your fingers?
The L293E has a saturation voltage at 1A in excess of 1V, so each half bridge is generating >1W at its max rated current, and in fact the device is rated to 5W but only if the ground pins are kept at/below 80C. The circuit board needs a large area of copper to dissipate the heat at full current (this little board doesn't look big enough), and/or it needs forced-air cooling. The chip will run hot at max current - but if its running too hot it will destroy itself. The simplest solution might be to direct a small cooling fan at the board.
You should also be aware that the absolute maximum current is 2A "non-repetitive" - this means if it ever sees 2A it is starting to destroy itself... Measure the stall current of your motor, and if its approaching 2A or over its not suitable for this driver chip.