The cheapo RX modules have an extra output that can be used as an analog RSS, but it's actually the output of the AGC. Laipac does make a better ASK RX module with good IF and AF stages, a very responsive AGC and bit-slicer, and consequently much better noise immunity. The RX is an ASK device and is a bit more sophisticated with an AGC circuit and bit slicer. Manchester coding could also be adapted to these modules. IIRC the Arduino library supports the ATtiny85 now. It's based on the Virtual Wire protocol by RFM, which uses a 4-to-6-bit encoding, a simple PLL for clock recovery, and 8x oversampling majority polling on the receiver. Have a look at Mike McCauley's VirtualWire library for the Arduino to get an idea of what's needed to get these modules to work reliably. They require a DC balanced bit-stream and a training preamble for the receiver. Those 433MHz modules (separate TX and RX) are not well suited for USART comms. Is there something more I can do to get decent range out of these modules? Ordering from DX is sometimes a crapshoot, so maybe this is another case where I've rolled snake eyes. I know good 433Mhz modules are capable of ~200m range - I used to have a 433Mhz gate opener powered by a 3v coin cell that would work from inside my hose to a receiver on a gate over 100m away.
I soldered 17cm of 24AWG solid core wire to each module for antennas. Once I go beyond ~4M, I get nothing but noise - even looking at the received data with a logic analyzer nothing is recognizable at any baud rate, or even just on/off cyles in the 50-500Hz range. With my modules I was able to get error-free short-range (2-3M) 9600bps transmission using a soft UART on a tiny84a, with 5v power on the transmitter. I found a helpful link that simplified my work:
#How to connect receptor rf 433mhz to avr codevision Pc
I have a 433Mhz wireless receiver/xmitter pair that I got from DX, and intended to use for sending data from an AVR to a PC with the receiver hooked up to a USB-TTL module.