Losing my flippen mind over the L298.

[Tophinater]

I really need some advice on a good motor driver for my vaccuum robot. I have been using the L298 chip the first time I got everything wired up and all set the chip just desided to crap out on me and take half of the other PCB with it. I have been using it with the kit that www.hobbyengineering.com sells. I liked it because its cheap and I can build it myself but just wondering if theres anything out there better(as in doesnt get as hot) and for a cheap price.

[Garak]

What diodes are you using for back emf protection? Are they all installed in the correct direction?

The L298 will shutdown if the motors are drawing to much current. It should start working once it cools down.

How much current are the motors drawing? What voltage are the the motors running at? What is your PWM frequency?

All motor controllers are going to get warm. There will always be some voltage drop accross the driver and how much heat it has to disipate depends on the current. The more expensive motor controlers use MOSFETS and have a lower voltage drop, the cheaper ones use regular power transistors and may drop as much as a volt or more. 1 Volt at the chips 4A limit will disipate 4 watts of heat.

[Comatose]

L298s will work okay if their current and thermal limits are respected. What drive motor or motors are you using? Its pretty easy to blow through a 2A per side rating, and cheap boards often won't be thermally or current limited.

For what its worth, we just recently launched a dual motor driver (if you're running two motors) its R/C pulse input only, but you can put it into a single pulse input mode which is extremely simple to interface to a microcontroller. Is 24V max and 10A per channel, thermally and overcurrent protected, so its quite hard to blow up, even if you hook up a motor thats too big. Its $65, which is pretty good for a medium-current dual driver, but may be more than you're looking to spend

http://www.dimensionengineering.com/Sabertooth2X10RC.htm


Beyond that we'll really need to know what motor or motors and battery voltages you're using to give a better recommendation.

[Tophinater]

All diodes and parts are installed correctly. I have used the motor controller before and it worked just fine. I was going to use some standard 280 sized gearboxed motors that draw approx. 750ma a each when locked. At 12v that controller isnt being taxed at all. The controller blew without a PWM signal even going to it. Also the V-reg for the logic as well as some of the PCB wiring went with it. I was planning to start with a 100hz signal 10-95% duty cycle at a 5v logic signal.

Comotose, that controller looks nice but its a lot more then I want to spend. Im thinking about giving my pololu serial controller a try but its only designed for 1amp per channel.

[Comatose]

I understand, if you're only drawing an amp, there's no reason to spring for a 10A controller. Sabertooth 2x10 is really best for speed 300 through cheap cordless drill size motors, power window motors, that sort of thing.

If you actually blew the trace wiring, thats some serious heat. I'd suspect a short circuit on one of the outputs.

[Garak]

The logic side going up in smoke would suggest that it may not be a short on the outputs that caused it. It sounds like you may have applied 12v to one of the logic level inputs but even then the impeadance on the input would be to high to do any damage. Maybe the voltage regulator failed and put 12 volts into the logic power input, that would likely fry everything. Another possiblity is that there was a peice of solder or something under the board shorting the pins.

I have 3 L298D boards that I use all the time and have no problems except with higher frequency PWM. I've used them to control motors as big as wiper motors(on slow input).

Anyway it sounds like a simple fuse would have saved your board. I'd put a 4A on the motor power input, my boards have a seperate logic input which I also fuse.

[Tophinater]

Im thinking that the Vreg did fail. Looking at the lines that burnt up on the PCB would suggest that. I had all the voltages and wiring hooked up correctly, I double checked them. Thats the only thing I can think of is the Vreg going bad. I noticed there are some other parts on the PCB that burnt up that I didnt see before. Im going to try and see if I can repair it and maybe the L298 chip will still work. If it does, Ill just pull out the Vreg and power my logic a different way.