|
Here is a first draft of the set of schematics for the Primus micromouse.
These schematics are shown below as images. These are a bit big on the screen and the qualty is not great. You can download or view PDF versions by following the links shown at the bottom of this page.
Primus has three circuit boards. The main board holds the processor, a driver for the sensor emitters, a voltage regulator, the nokia LCD and some buttons and connectors. A picture of the schematic is here:
Nothing very clever here. Notice that there is a serial port connection. Including that meant I did not have enough pins for the LEDs and the speaker so one of the LEDs is driven in parallel with the speaker. All that means is that the LED will change brightness while you play sounds through the speaker. That speaker should be a piezo type to minimise current drain when the LED is turned on. If you can only get a conventional type, you will need to add a series resistor of about 270Ohm to limit the current. That will also limit the sound output by a great deal. If you do not limit the current flowing, the regulator will soon become very hot and you run the risk of burning yourself.
The sensors look like this:
Although the AC coupling components can be left out, they make the code a bit simpler. However, without them we can use the before and after readings to give some indication of sensor reliability. For example, a very high before reading would indicate possible saturation risks in a brightly lit environment.
The motor drivers are also pretty simple.
All the work is done by the SLA7062. These have been hardwired to step in sixteenths. There should be no problem with this and it will give us the smoothest control of the motors. Motor current is set by the potential divider R15/R16 which should give you about 0.3V for the reference voltage. There is no need for very high precision. Transistor Q1 is normally turned on by current flowing through R14 thus pulling Ref up to near 5V. Any voltage above 2V disables the motor driver circuits in the 7062 reducing the otherwise significant current drain of stationary stepper motors. The processor can turn off the transistor by raising the MOT_ENA line. the REF voltage is compared with the voltage on the SENSE pins. Using three 1 Ohm resistors in parallel gives an effective resistance of about 0.33 Ohms so that the currnt is limited to around 0.3/0.33 = 0.9Amps per phase - exactly what the motor specs call for. There is no point in using a larger current, the motors will simply get hotter with no noticeable improvement in performance. Indeed, they may saturate and performance will be less. the other reason for using three resistors in parallel is to make it easier to build. Peak dissipation in these resistors is I2R - about 300mW. It can be a bit hard to find suitable carbon resistors so we can share out the load over three common 125mW types instead. They dissipate only around 100mW each peak and the average is well below that so they barely get warm.
The smoothing capacitors are essential and must be rated well above the battery voltage as the motor driver uses switching techniques to control the motor coil current resulting in large inductive spikes back through the power lines which can reach surprisingly high voltages. The capacitors do a good job of calming these down
Here are links to the schematics in PDF form.
|
