RF module issue and solutions

[frank26080115]

Has anybody ever tried to use one of these?
http://www.robotshop.ca/home/products/robot-parts/communication-control/data-telemetry/on-shine-low-cost-tx-rx.html

anyways, i have a sort of homebrew serial o'scope, and i checked out the output of the digital pin when the transmitter was sending a binay 0

here it is (the program is just something I did in Processing, the o'scope is just an analog pin on my arduino)
http://i156.photobucket.com/albums/t31/frank26080115/robot_charlie/radiooutput.png

when i send a binary 1, it will go high for several milliseconds, it is perfectly capable of UART communication during that time, but after several milliseconds, the receiver goes back to idle

the problem is that, since the output transmits garbage out to your microcontroller's RX, it will keep tripping the serial interrupt and you can't do anything useful.

solutions?

how about a small 8 pin co-processor on both the transmitter and the receiver?
the transmitter co-processor will output 0x00 forever on it's TX pin to the transmitter module, this way it at least the receiver side can verify that the transmitter is there and ignore the 0x00.
when your main microcontroller's TX outputs into the transmitter co-processor's RX, the co-processor will send a 0xAA (0b10101010) to the transmitter module, then another byte which is what you actually wanted to send.
mean while, the receiver co-processor gets the 0xAA instead of the 0x00, and thus prepares for the next byte, which is then stored and sent to the receiver's main microcontroller.

thus, only valid data is ever received. you can also add features such as device ID, apply math functions to the data, I^2C or SPI interfaces, and also various indicator LEDs

another solution

the TE pin on the receiver module is a very noisy and also it is always around 2.5V, BUT it will go to 5V for a very short period of time every time when the transmitter outputs binay 1. You can use a comparator so that only voltages over 4v will cause it to trip an interrupt pin, and in the interrupt routine, you can then enable the serial port, then exit the interrupt, and then the actual data will then trigger another interrupt, in which the byte is stored into a buffer

[Steve Joblin]

Interesting approach... i purchased one of these low cost units from Renolds (http://www.rentron.com/rf_remote_control.htm)... it works quite well actually... I am powering the transmitter with 12 volts and I have a good quality antenna on both the transmitter and receiver.  It does sometimes produce errors, but that is to be expected as there is no error checking    

I guess if you want a good quality, reliable, error free short range communication link, you should look to bluetooth or zigbee.  Bluetooth has become quite affordable with products like this coming to market : http://www.sparkfun.com/commerce/product_info.php?products_id=8474

[frank26080115]

will using 12v on the transmitter really going to make a difference?

[Steve Joblin]

Yep!  the more voltage you supply (between 5 and 12 volts), the stronger the signal that the receiver will be able to detect.  Please note that the specification of 12 volts is for the unit I purchaesd... I'm not sure about the specs for the unit you have, but it is worth investigating!

[frank26080115]

my robot can't really have a 12v supply on board, the battery is 7.2v and i'm using servos, i really want full duplex communication and not just a user side transmitter.

[Steve Joblin]

hmmm. sounds like bluetooth may be what you need.

[frank26080115]

Actually, I have three wireless USB modules from Cypress, I haven't used them yet, they need a 2mm female pin header
You can sample them, and they usually almost always send (only ranges 50m and under, they don't give you the 200m version)

[frank26080115]

when powered by 12v, will the input work on 5v?

[Steve Joblin]

the two power sources are not directly related... the greater the voltage on the transmitter, the stronger the signal (up to the max voltage recommended).  On the quantrary, there is no difference in the quality of the reception if you change the voltage on the receiver... the quality of the receiver is soley based on the quality of the antenna, not the voltage.

[frank26080115]

yea but if i power my transmitter, can i safely interface that with a 5v MCU?

[Steve Joblin]

Yes... the power supply for the transmitter is to power the transmitter... the transmitter and the MCU are only connected so that the MCU can tell the transmitter what to trasmit.  They can be two completely different power supplies (just make sure they share a common ground!).

[damanfb]

I am working on a similar project as well.  I have read and checked on the oscilloscope that the receiver begins to pick up noise if it is not sent any data.  It's adjusting its gain looking for weak signals which ends up amplifying noise.  To alleviate this for the most part, you can create a packet of data to send.  First output a dummy byte, then your data.  You can also add a check sum for verification. 

For example,

I have my micro controller outputting a binary 1 for ten seconds while the receiving microcontroller is initializing.  Then I have the transmitting PIC send a byte of 238 followed by the four data bytes I am sending.  It is working out alright for now.  Still some other noise issues that need to be worked out though.