Lots of noise with 434MHz receiver/transmitter

Wireless & IoT
1

Hi, I’m working on a project using the 434MHz receiver/transmitter pair (WRL-08946 and WRL-08949), and I’m having problems with noise. It’s a wireless SNES controller “mouse” for me to use from my couch with a computer hooked up to my TV. I’ve been fiddling with different antennas and stuff but I don’t really know what I’m doing. I have the receiver’s data output hooked up to my (really old :D) oscilloscope to get a vague idea of what’s going on.

Both ends are using a PIC18F2550, right now I am just sending a test signal from one of the data lines. The transmitter is running on 2 NiMH AAA’s, and clocked at 4MHz from a 4MHz resonator. The receiver is running on USB’s +5V and clocked at 48MHz from the same type of 4MHz resonator with the PLL.

When the controller’s about a foot or so from the receiver I get a decent signal (not great), but if I move a few feet away I get a lot of noise up to the point where there’s no discernible signal at all. I’ve tried different lengths of wires, made little helical antennas from the wire, and all different combinations and orientations, but like I said, I don’t really know where I’m going with this :D. The transmitter is pretty limited space-wise since the controller is only about 13 cm wide but I can do whatever with the receiver (though small would be nice :D)

The product page says something about 500ft in open space so I must be doing something wrong, although I expect the range to be reduced since I’m only running the transmitter at 2.6 or so volts.

Any ideas as to what I can do to fix this up? Any help is greatly appreciated, and thanks in advance! :smiley:

Wow, that post got pretty long pretty fast :smiley:

2

what length are you using for antennas? I’d start with a 1/4 wave antenna on each device - 6.5 inches. With those devices, I’ve gotten at least 50 ft indoors through 2 walls using just a straight 6.5 in piece of wire stuck in a solderless breadboard.

Perhaps a more important point - what are you sending? what is your encoding scheme? describe your input to the transmitter.

3

dabigjhall:
project using the 434MHz receiver/transmitter pair (WRL-08946 and WRL-08949), and I’m having problems with noise.

…

The transmitter is running on 2 NiMH AAA’s, and clocked at 4MHz from a 4MHz resonator. The receiver is running on USB’s +5V and clocked at 48MHz from the same type of 4MHz resonator with the PLL.

The radio link is not a substitute for a piece of cable.

The data delivered is subject to bandwidth limits and

impulse noise from other radio sources. For correct

operation the min and max data rates must be observed.

[[u]Tx Data sheet[/u]

200 min 1k nom 10k max bps.

It is expected that your ‘payload’ data will be combined

with a clock or use a self-clocking encoding scheme

such as [[u]Manchester Encoding[/u]

Next, the gain of the receiver will depend upon the

signal strength, and at very close range the receiver

will likely be saturated. A fair test is over a working distance

of several tens of metres.

[[u]Rx Data sheet[/u]

Note the data rates for the RX are narrower than

the Tx, and therefore rule the system.

At this point it would worth sending a plain 1k BPS 50%

duty OOK modulation into the Tx and viewing the

recovered Rx data on your analog 'scope. Code to do this

on the existing PIC is trivial - probably easier than

rigging a '555 timer analog circuit for testing.

Antennae design for such a simple RF link would not

make much difference to the performance. Single-ended

loose wires of 30cms would suffice (as noted by philber).](http://tinyurl.com/d6fk4x)](Manchester code - Wikipedia)](http://tinyurl.com/cp5w2e)

4

Thanks for all the info!

At this point I had just been sending a sync pattern (0x55) repeatedly at approximately 1953 bits/s (works out to that with the timer interrupt) and trying to detect it with the receiver - it wasn’t detecting it properly amid the noise so I hadn’t gone much further yet.

That’s good to know about your results with a 1/4 wave antenna, Philba, I had done the same thing at first but apparently I have problems elsewhere that I had blamed on the antenna. I will have to bend it over a little bit to get it into the SNES controller (about 13 cm wide, so a little shorter than 1/4 wave) but I think that will be fine even if it’s somewhat inefficient (after all, I don’t plan on having my couch 50 ft from my TV :D)

Manchester encoding seems like it should solve a lot of my problems, too, thanks for that and the other info bigglez. One thing in that dat. sheet that had worried me was the part where it said the input and output duty had to be between 40% and 60%, and it looks like Manchester coding will take care of that. I have also been looking at error correction stuff on Wikipedia and it looks like I can group my 12 bits of data into 3 groups of 4 bits each and use (8,4) Hamming code without much trouble. That would give me a little error tolerance which should help.

I think this should take care of my range and noise problems. I’ll get back to work on this in the morning and put this information to good use :smiley: Thanks again for all the info!

I hope this doesn’t double post, I kept getting “forbidden word” errors when I tried to preview. For some reason “dat. sheet” is censored when spelled out? Oh well.

5

My tests were done with manchester encoding. People often report good results with “serial I/O” directly connected to the TX pin, though.