I MIGRATED THIS DISCUSSION FROM WIRELESS - THIS EXPLAINS THE FOCUS OF THE FIRST PAGE
THIS IS the BEST (imo) Telemetry module out there currently !!!
This is a radio module that's small, fast, inexpensive, powerful, US legal 900mhz (and others). It has adjustable power output, spread spectrum frequency hopping, encryption, and TONNNNSS of other features. Also, at only $70 a pop (from digikey.com), they are Very easy on the wallet! especially compared to the digi xtend radio that is near $200 per. ouch!
I have three of RFM's DNT900 radios, and am working on Arduino programming to set-up and control them. I am working with the friendly tech at RFM who supports their products, and we are discussing a few more details about the needs of the radio. I intend to post more as I get farther along, and create a code block that people can copy-paste into their program to utilize this fantastic radio.
It has been a month, and I have just recently set up the architecture and got communication between two radios. I have them set to 200kb/s RF over air, allowing for up to 1watt of transmission power.
Still to go - i need to set up radio access lists, add 128 bit encryption, and adjust baud rate.
oh, and clean up the code! messy
Replies are closed for this discussion.
As of this morning...
I now have authentication lists based on the units MAC addresses, and 128 bit AES encryption enabled.
an exclusive & secure connection that hops frequencies, and is spread spectrum - very nice indeed! that means it will be very tough to sniff and interfere with your wireless control. It will also be more resilient to the environment around it
At the lowest power level (1mW), I was able to transmit from the far south basement corner to the far upper north corner of the 2nd floor. Pretty good i think - especially considering the duct work, walls, and floors separating the radios. I'm also using a crappy, made-in-china, throw-away, 1/4 wave-lenth antenna on both units. The baud is currently set to 115200 b/s.
Next on the list...
- construction of a semi-directional antenna
- reading back stats from the radio such as signal delay, and adjusting the power level accordingly
does anyone know of an efficient way to parse hex words? im newbie
This is quite cool.
I like it ! I have never used one of those before - but that sounds like a simple integration & form factor
possibly multiple modes - so people can interface it with multiple systems?
personally I am partial to the CHR-6dm 9dof sensor with its powerful processor, and its great EKF built in. rock solid, and really great feedback
Lew, im glad to meet someone who's had experience with RFM before! Could you clarify what you meant when you said they were half duplex? Because it is my understanding that the transceiver acts as full duplex. The actual hardware itself only works at half so as not to blow out its own receiver (maybe thats what you were referring to?) - but it does it seamlessly so users can use it at full duplex.
I've never used the XBee's before.. they seem neat, but better suited for in-house or short-range applications. The DNT900 is serious business. Im waiting for a reply from RFM tech support - clarification on the power output settings. Once that is answered, i hope to begin to push this above 1mW, and be able to get some benchmarking on the real-world capabilities.
Today i built some general purpose string sentences that will communicate data between paired radios to deliver flight control, and return flight statistics in real time.
Currently its pushing 30 control sentences (3 unique sentences, 10 iterations per second) from base to remote, and the base is receiving 8 flight-statistic sentences (2 unique sentences, 4 iterations per second) back from the remote. It comes with sentence headers similar to NMEA gps style, and a checksum.
I also am working on the parsing of these sentences, using code based on the GPS parsing section of ArduIMU - compliments to who originally coded that part (jordi? williams?)
Half-duplex refers to the fact that these units cannot send and receive simultaneously. They must take turns switching between transmit and receive mode. Doing so causes what is known as turn-around latency. They have a minimum poll interval built in, to assure that said turn-around takes place even if you are in the middle of a lengthy transmission (in which case your transmission will suffer an interruption during polling). Since all this is buffered, the only thing you'll notice is latency.
Just to clarify, there is no such thing as software making a device full-duplex. This is strictly a function of hardware (RF) design, typically requiring some type of band-pass/band-reject filter, which at 900MHz takes the form of a tuned cavity, commonly known as a duplexer.
Like most 900MHz transceivers, the RFM900 employs PIN diodes to provide solid-state switching of the RF input and output stages. This is another clue that you're dealing with a half-duplex device. Given the excellent turn-around latency (assuming AES encryption is turned off) of this device, I wouldn't worry too much about it. As you said, from a software perspective, you can keep jamming bytes into the send buffer without having to worry about losing them.
They're all on their website - on the DNT900 product page theres others (specs, brochure, etc)
And Lew - thanks for the lesson! so even though the dnt900 can act as full duplex, its still referred to as half because of the fundamental operation of the transceiver radio. I think its incredible how small all these devices have gotten. my pops and brother and I used to go to the Ham swap's once or twice a year, and i swear that was more fun than christmas. Most of the stuff was pretty big and bulky back then -
Sentence parsing is currently half working. I have the correct values in *string form, but i'm not bright enough to have figured out why doing an int(*string) yields 49 instead of the 10 that it should be. serial.print(string) works fine, and yields the expected 10, but i need it as an int variable. ideas?