Components and Tools
ComBase (MiRF2RS232)
The ComBase serves as main communication device between the PC and my robots. It's connected via RS232 to the serial com port of the PC and operates in the 2.4Ghz band. The core component is a nRF24L01 from Nordic Semiconductor. This little transceiver IC is pretty easy to use an reaches data rates up to 2Mbit/s. Driving it with the breakout module from SparkFun and a 2.4GHz duck antenna you can reach up to 80m range.
Design Comments
nRF24L01
Since the nRF24L01 supports SPI it's pretty easy to use it with an AVR controller. Also SparkFun's silkscreen labels the pin out of the breakout module self-explanatory. So connecting SCK, MOSI and MISO to the corresponding pins of the AVR is obvious. For CE and CSN I've chosen appropriate I/O pins as stated in the following table and IRQ is connected to the external interrupt 0 of the Atmel controller.
AVR ATMEGA8/168 |
nRF24L01 |
Description |
PB5 (SCK) |
SCK |
SPI clock signal |
PB4 (MISO) |
MISO |
SPI master in, slave out |
PB3 (MOSI) |
MOSI |
SPI master out, slave in |
PB1 |
CSN |
SPI Chip Select |
PB0 |
CE |
Chip Enable Activates RX or TX mode |
PD2 (INT0) |
IRQ |
Interrupt pin. Active low |
RS232
To connect the AVR controller to an RS232 interface it's necessary to convert the signal levels from 5V TTL/CMOS to EIA-232 levels and vice versa. This can easily archived with an MAX232 IC.
ISP
For easier updating of the firmware the ISP pins are routed to an header compatible with the Atmel ISP10PIN.
I/O header and LEDs
PC0 and PC1 are connected to LEDs for debugging and status indication. Also PC0 to PC5 are connected to a header so this board can be used as A/D capture device. The LEDs are connected with an jumper so they can be disconnected if the PC0 and PC1 is needed otherwise.
Power Supply
The design of the 5V power supply for the AVR and 3.3V for the nRF24L01 is pretty much the same as used in the KraftModul.
Schematic Diagram
Like all my circuits I've planed the schematic diagram and board layout with Eagle.
Finished Board
The vigilant reader may recognize a small difference between the routing diagram and the processed board especially under the AVR. This is because of some mistakes in the first layout which I've already fixed in the diagram.
Comments for further processing
The crystal is placed really bad and can't be used reliable in this configuration. It was a first test using an external crystal instead of the build-in 8Mhz resonator on an own circuit and I didn't realize that it is such sensible against noise from other components. Since the 8Mhz is fast enough this doesn't matter for the ComBase.
Since the breakout module of the nRF24L01 already has an on-board 3.3V LDO regulator it's not necessary to have one on the ComBase board.
Software
Documents
nRF24L01 Datasheet from Nordic Semiconductor
Breakout Board from SparkFun