ECE272 - Motor Performance Upgrade - Spring 2006

Total Price: ~$16.00 + $hipping

During winter term, I had decided that my Tekbot was lacking speed. To solve this problem, I purchased tuned performance motors (Tamiya Co. 'Sprint-Dash') which spin at some unimaginable RPM. However, these new motors demanded a "recommended" 3 amps at 2.5-3.0 volts. The supplied 'Tekbot' motor control board has a claimed maximum capacity of ~800mA, so obvious problems would occur if it was used to power the new motors. I then proceded to build my own H-Bridge, using this one as a reference. At first I ordered and tried to use a LMD18200 IC H-Bridge, which would have been far simpler. But it seems this component will not turn on with lower voltages. This left me with the option of build an H-Bridge with individual transistors. Although not perfect as there is no "Enable" input, my new motor control board acts just like the Tekbot model. In addition, the board, as well as the motor, are currently running at ~7-8 volts, significantly heating the motors. I predict the motors will fail or melt the plastic gearbox after a short span of time unless I find a way to cool them. In the future I might reduce the voltage the motors operate at, but for now the insane speed is nice. Another problem, although not entirely unexpected, is that the batteries are draining rather quickly. Maybe a future project will fix this...

The design is based around the inexpensive and readily available TIP120/125 Darlington transistor, which can handle plenty of current, although they are somewhat inefficient. Because of this chip's packaging and inherently higher heat output, heatsinks are necessary. Two smaller transistors are responsible for selecting which path the current takes through the H-bridge. The Tekbot platform uses only one input to determine the motor direction, so the two smaller 'input transistors for each bridge needed to be driven with one input. This results in one signal which is split into two, one of which is passed directly through to a control transistor, and one which is inverted and then passed to the other control transistor. A 7404 logic chip, a hex inverter, can be used, however I used the 7486, a quad XOR chip, because I had one available. With one of the XOR gate's inputs tied to VCC, it acts just like an inverter. Perhaps I will update the design with an "enable" input, but for now it is good enough.

When testing, the motors and controller would work for about 30 seconds and then slow to a halt. After some testing, it was discovered that the fuse on the charger board was breaking the circuit after considerable heating. I believe the motorcontrol board is drawing more amps then the fuse is rated for. For the time being, I have eliminated the fuse as shown below, but this would seem to be a very bad idea. I hope to fix this soon.

Update: Although highly amusing, the intense speed comes at some significant disadvantages. First, the battery life is only a matter of minutes. Second, I predict poor long-term reliability because the motors are outside of their safe operating limit and become too hot to touch after lots of use. The only reason they still work is probably their high-quality tolerances. I plan to run the controller off of less voltage, maybe only 3 volts (only 1/2 of the batteries) in order to allow me to power the tekbot for longer intervals and to return to the recommended operating voltage for the motors.

Parts List:

The Final Product:

Notice the Significantly Larger Transistors and Their Related Heatsinks:

Testing the Logic Portion of the New Board:

The New, "Specially Tuned" Motors:

Mounted and Connected to the Tekbot:

The Temporary Fuse fix -NOT Recommended, potentially hazardous to your health:

Full Video Coming Soon...!

Click to See a Video of the Tuned Motor Test Operation! ~1.24MB

-Requires Windows Media Player