Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

DC powerline Communication - Signal to DC coupling

Status
Not open for further replies.

Ianb007

Junior Member level 1
Joined
Feb 17, 2010
Messages
16
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,429
Hi. I have set up 2 microcontrollers that communicate with each other via USART - 1 line. I now need to get this signal onto a 24V DC bus and then get the signal back 2meters further down the DC bus. This means that the DC 24V powerline also carries the signal. Any advice/ideas/experiences?? My idea is to use transformers? Will capacitors work??
 

To transmit data or another signal over a power line, you need to inject it in it and pick up where you need.
The easiest way, if your data do not carry a DC component, is to use a capacitor coupling, from your data source to the power line, and back to data receiver.
To prevent the injected data from propagating in a not-desired direction, use an inductor in the power line which would function as a choke. The injected signal would travel along the "clean" power line, not where the inductor is. The same on receiving end.
As your DC power line ( and it holds for AC power lines, too) has two conductors, your inductor should be wound by two conductors "bifilar" way. If your DC power-line wires run close to each other, you could try to use only one capacitor to inject the data into one wire, and back. If you observe some loss, you can use two capacitors, to inject the data signal at once into both wires of your DC power line.
Make sure the capacitors are rated to more than DC voltage on your power line, and when making the inductor with two parallel-wound wires, make sure the wires are insulated and this insulation withstands the DC voltage on your power line.
The capacitance of the capacitors can be estimated from the slowest data rate; I would suggest to use 0.1 uF/ 240 V each. Inductor inductance should also be chosen for the slowest data rate; I would suggest 10 mH, e.g. 10-20 turns on a E-shaped ferrite core, 0.25 square cross-section.
Some experiments are needed as your DC power line should be located "above" ground which is used as the return line for your data.
 

    Ianb007

    Points: 2
    Helpful Answer Positive Rating
Thanks so far. Will it be sufficient for a 8Mhz data rate. 9600bps. Or should I rather decrease my frequency?
 

You can't AC-couple an UART signal without restoring the DC level at the receiver. Known powerline communication standards use modulation.

Will it be sufficient for a 8Mhz data rate. 9600bps.
The data rate can be either 8 MHz (8 MBPS) or 9600 BPS. If it's 9k6, what's the
meaning of 8 MHz?
 

As you intend to transmit data over 2 meter distance only, I think all other effort is not worth of it; I would simply use a small-size coaxial cable over those 2 meters.

There are data systems over power lines covering a wide area and long distances, up to several km. But all cause ugly interference and collect other interference, too.

Solve small problems without complex designs, please. You can certainly play with transformers, chokes and capacitors as I and others suggest, to compare with the simple piece of cable and see the difference.
 

Ok i have played around with the inductance and capacitor values. If a load is connected the signal changes and goes up together with the input signal but then when it reaches the 5V value immediately drops down again. I guess I will have to design and implement a monostable vibrator. I guess the inductors I design will have to be rated/designed for the current my max load will draw??
 

Yes, you are right, your inductors must pass the rated DC current while they block your data signal from getting to unwanted direction.
As you wrote that the data pulse goes up and then down, it looks like you are using either too low data rate, or, your coil inductance is not high enough. You can try to use a coil with ferrite core to increase inductance, or, increase data rate.

One of earlier commenters mentioned that your data signal might need to transmit the DC component. Then you cannot use your DC line as the two DC "signals" will be in conflict (you cannot separate one from the other).

If you still insist not to use a separate thin coaxial cable over two meters distance, you can add a fiber cable to your DC line. Using IR LED and a photodiode, you can transmit the data signal as is, including its DC component. Over several meters you should see almost no loss even if you use a plastic fiber.

There are other solutions but all more complex than an added thin coax cable.
 

As mentioned before passing the UART data directly on the power line is impractical since you need the DC component of the signal, that means you will need to change the DC level of the power line according to the data you want to transmit.
There are various of "of the shelf" UART transceivers for power line communication. They are simple to use and should also be cheap to implement. I suggest you will search for "dc powerline transceivers" this should give you some results.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top