FSK Modem for DC

Does anyone have any good resources regarding FSK PLC over DC? I am required to communicate over a 12VDC power line but I really have no clue how this stuff works. I have been looking around at some modems, although most of them seem to be for 110/240 VAC.

I figure I have two options:

1) Get an FSK chip and build the system from the ground up. 2) Buy a modem that fits the application.

Honestly, I would rather take option 2, but if that is not possible it seems there are plenty of chips out there to do what I need. If I can find a modem with any sort of Ethernet/RS232/Serial output, then I can easily interface it with the Arduino, but finding such a modem has been my problem.

If anyone has any experience with FSK, I'm all ears.

Thanks

*I've just been told that a UART would be a feasible application here, although I'm not too brushed up on the topic

A UART is usually for converting parallel data into some form of protocol driven serial (Universal Asynchronous Receiver / Transmitter). The Arduino has one built in. The real problem will be modulating the 12DC power line which might be easier said than done. What sort of power supply are we talking about? Farly high or very low impedance? Is the nature of the load one of fairly constant current or is there a lot of current noise (constantly varying in an unpredictable manner)? How far does this communication have to go?

BillO: A UART is usually for converting parallel data into some form of protocol driven serial (Universal Asynchronous Receiver / Transmitter). The Arduino has one built in. The real problem will be modulating the 12DC power line which might be easier said than done. What sort of power supply are we talking about? Farly high or very low impedance? Is the nature of the load one of fairly constant current or is there a lot of current noise (constantly varying in an unpredictable manner)? How far does this communication have to go?

I believe that the maximum length it should have to go is 300 feet. The current is fairly constant, and even if it varies it is small and predictable. The nature of the load is a 12V Motor and some flashing lights, total current requirements should be under 10 amps (depending on the load on the motor which we are trying to minimize). The power supply is an NRS-12/40 Battery Charger that charges an array of batteries that then supply the devices. Looking at the Data Sheet for the NRS-12/40, there was no output impedance given, only some voltage/current/power statistics (which you might be able to calculate it from but I'm not sure, else I'd just have to hook up a simple load to the unit and test the difference between no load and load I guess)

If you couple the audio from the modem to the DC line through a capacitor, and the same on the receive end you should get it to work. Keep to using 300 baud for best results.

But it may be hard to drive such a low impedance line with any decent amplitude. If the modems are high enough quality it might just work. Some old high quality modems that follow the Bell 212 standard might do the job with a little modification. There used to be a single chip version that, a 73K302 or something like that. Don't know about availability these days though.

The motor might be putting more signal on the line that you are - you might have to add some filtering to it to help.

I don't need simultaneous communication in both directions, but I do need communication in both directions. I only say this because I do not think there will be any overlapping instructions being sent from either end, therefore half-duplex should be fine. I say I do not "think" because the system is not fully developed and I have not thought through every single avenue of communication, but as far as I have come so far, I can't see there being a need for simultaneous communication.

I looked up the MX614 and the 73K302 and they both seem reasonable, although they state they are made for telephone or dial-up lines, is it still OK that I would be using them with a straight DC power line?

Please excuse my ignorance on this topic, but while I probably could figure out how to interface these chips with my arduino, how do I actually:

1) Put one of these things on either end of the power line? 2) Interface the little chip with the fat power line?

As far as 1) goes, it seems the 73K302 at least uses a simple TX and RX line, so I guess you would do the typical setup there (assuming those are the lines that transmit across the power line). But, when it comes to 2), how do I actually connect them up? I realize I'm pretty sure I'm putting this thing somewhere in parallel with the line, but it is the physical parts that stump me. I realize this may be a little vague, but all my knowledge is theory with no practice, as in I might be able to know everything about how the Bell 202 interface works as well as these microchips, but actually using them is a different story (when it comes to actually putting something together in the real world). Unfortunately I don't completely know everything about those chips yet but I will do some reading.

Thanks again,

cyph

1) Put one of these things on either end of the power line?

I said in reply #3 connect it up with capacitors. A capacitor will block DC and allow AC to go through it. You have an audio signal which is AC so you AC couple to it.

Grumpy_Mike:

1) Put one of these things on either end of the power line?

I said in reply #3 connect it up with capacitors. A capacitor will block DC and allow AC to go through it. You have an audio signal which is AC so you AC couple to it.

Right, so it is pretty much the modem -> capacitor -> junction box? The junction box just having a terminal where the power line and the modem are connected?

Yes don't forget the ground return as well Modem ground -> capacitor -> power line ground.

Grumpy_Mike: Yes don't forget the ground return as well Modem ground -> capacitor -> power line ground.

If I get a chip like the 73K302, would I need to have any sort of amplifier on there? I would assume so since it is a small chip, but then again I have no experience with this stuff. Just curious what is necessary to get down the line (maximum length 300 feet).

Until you measure the impedance of your line there is no way of answering that question. Size has litle to do with drive capability.