Everything sounds do-able EXCEPT the 9V battery, That simply WONT DRIVE A MOTOR unless it be a pager vibrator motor and that not for long. As to Sparkfun I really don't care for their prices or in a lot of cases the quality of the equipment and parts they sell BUT that is STRICTLY My opinion and that is the reason I don't buy anything from them, So I am not the one to make a call there. Pololu on the other hand has an Excellent Reputation and as far as I have seen excellent quality equipment or at least I very rarely hear anything wonky about them here in this forum. If you are contemplating motors and servo's you WILL need a separate power supply for the motors/servo's as they are too high current and FAR too noisy to share anything but a common ground for the control signal return path. Never share power as It is usually if not always more work to make both work well from a common power source than it is to provide separate power for them. IMO
I see no need for diodes if you plan to use this driver as designed.
You will definitely get better performance with a heat sink.
Your heat sink should be sized at a minimum to match the
size of the chip itself.
As for a battery, check the the Ah rating of your proposed battery
compared to what your motors will draw. It looks like they are
70mA no load, and 800mA stall current.
Thanks kd,
You are correct about the flyback diodes and tech support at Polulu agrees.
I'll definately put on the heatsink with the thermal tape that I ordered. I hope it fits.
For power to my motors I'll probably use 6 rechargeable or alkaline batteries. I might use a 7805 to regulate to 5 v, but I don't think I need it. I'll play with it and see what works best. Thanks again.
I would recommend a thermal "paste" (for notebook heat sinks and high power graphic card heat sinks). The tape will not make the best contact between the two planes (heatsink and device flange). If sparkfun does not have it, your local computershop will. Just look for the stuff with "silver" in it.
No, you will need a nut a bolt. It very, very slowly dries out (good thing). You would still need a nut and bolt to make the tape work. The important thing is to make good thermal contact with the most surface.
Radioshack has this online which I repaired an overheating Mac with:
Same stuff that gets an extra amp or two out of my full bridge drivers.
Two things... the thermal tape will work as well as anything else... as the plastic case wasn't designed for heatsinking and has a very high thermal resistance. There is another type of motor driver that uses a big IC with a metal tab on top, that device is capable of a great deal more current. I built a valve controller that 'Worked' a three wire solenoid "Pilot Valve" about 10 12 years ago. I used one of these parts An L298 and it was directly heatsinkable... A minute of searching found these Items (There were several of them,, all shields)...< http://compare.ebay.com/like/150765336052?var=lv<yp=AllFixedPriceItemTypes&var=sbar for $7.25 and the Spec sheet says it is good for 4 amps PEAK and 2.5 A @ 80% duty cycle. I scrapped the Ic in favor of a relay and an NPN transistor as the area in which my application was used was Agricultural and nearby lightening strikes took out the IC as frequently there could be 50 meters of wire between the solenoid and the controller. The issue was 'Transverse fields" or the way the energy from the strike literally electrified the earth around the device. I've measures several KV of induced voltage... lost a couple of small strip chart recorders before I totally realized the potentials involved. USE your 'browser and don't be afraid to look around. I've found that the people who make the biggest noise (Big Ad's All over the place) are the absolute WORST for prices. It should be an easy task to develop a 20 A reversible motor driver with a relay and a power Mosfet... Braking too, even to be able to use the stored kinetic energy to charge the battery for braking. It is a driver I designed many years ago for bi-polar or Latching Solenoids... I don't think I ever had one fail from anything beyond the occasional defective relay. It was a 20A Omron DPDT relay and a D44H11 transistor (Mosfets would have been my first choice but they weren't as reliable or as inexpensive as the relay/transistor combination... At that time power Mosfet's cost about 4 times what the D44H11 cost. Very reliable for my application, I made several thousand of them (not me directly but my crew did) and ALL worked without fail. Use the L298, NOT the L293... unless you are good enough to add power transistors or Mosfet's to handle the real power. Easy and Simple.
Okay, I understand now. This Toshiba chip is oddly robust for how ridiculously small it is. Just as Doc has suggested, your mileage may vary.
Since this chip has no thermal protection other than the package and 175C overtemp circuit, the "paste" or "tape" is irrelevant. All you want to do is protect the MCU from the "EMF" or current backlash (AC) from the motor(s). Doc may have some suggestions better than my 1N4001 diode on your PWM output, but I do believe you are on your own on such a small device (298s or 293s have much more built into the device). Maybe somebody can suggest where to put the diode to get your reverse back
a small Tranzorb of the appropriate voltage might work well for 'back emf' but I have another real 'old school idea that I will develop and share in the next few days might just work better... I don't have a good spice model for a motor, have to go with my experience first. Real tired now and not thinking as clearly as I might like. For the follower idea use the PIR device as a distance controller and the beacon for required movement...
I'm going to bed shortly. I do however need my hour with the 2 books I am reading. A Book on C and the C++ Primer. Electronics is easy and fun, C is difficult...
cool. I would like to hear it. Like I said this small robot is just a prototype for my bigger robot that I will build. For that one I don't think I need an h-bridge, but they will be much bigger motors. about 200w.