Sending data on AC supply lines

[Rouen]

I 'am looking for a circuit, to send data to a remote site and the only transmission path is the AC supply lines.
The circuit needs to use FM superimposed onto AC.
Thank you.

[rshayes]

How remote is your location and what type of power lines?<p>Usually, carrier current systems use frequencies in the 100 to 200 KHz range. The transmission line characteristice are very nonuniform in this area, and some frequencies work much better than others. Frequencies this high do not pass through distribution transformers very well. A few milliwatts will work within a house; some wireless intercoms worked this way. The X-10 system uses pulses instead of a carrier. It is also usually considered to be limited to the secondary side of a distribution transformer.<p>During World War II, some use was made of carrier current communication by radio amateurs. The 1944 and 1945 editions of "The Radio Amateur's Handbook", published by the American Radio Relay Legue (ARRL), had a brief summaries (8 pages) of carrier current communication. Ranges were short, about 5 to 10 miles using 25 watt transmitters. Wartime issues of QST magazine might have more information. These might be available on microfilm or CDs in a technical library.<p>Power companies use carrier current communication, but they can couple to the high voltage line without going through a transformer. This would increase the range a great deal. Some of the professional journals of the 40's and 50's probably have papers on these systems. Again, these might be found in the engineering library of a large university.<p>At one point (late 50's) an emergency warning system was proposed in the United States. This used pulses of 180 HZ signal superimposed on the 60 HZ power. This was put on the power lines at major distribution points. This frequency was probably low enough to pass through power transformers efficiently. I believe that is was generated by some form of magnetic amplifier tripler. A carrier of a few hundred hertz might be usable with a slow data rate.

[ljbeng]

Look at Phillips Semiconductor TDA5051A. It works up to 2400 baud. I have been experimenting with them. Just beware of mistakes in the datasheet. Get ahold of an application engineer.

[toejam]

I have a pair of radio shack intercoms that i think uses that system to work inside a house. I did get a neighbor through it once and it had to get by two pole pigs to get to me.

[haklesup]

The X-10 protocols are explained on links from this site http://www.geocities.com/IDO_Bartana/ Check out the Technical articles link. <p>The X-10 protocol works for digital data sent at rates no faster than 120 bps since it is encoded at the zero crossings. If you want to transmit audio, you will need a different technology.

[MrAl]

Hello there,<p>Take a look at National's application note
number 146 (AN-146).
It's an entire FM carrier system transmitter-
receiver with PC board layouts and all <p>Let us know about what you do if you can.<p>Take care,
Al

[d1camero]

for a product, check out http://www.phonex.com<p>d1

[amuron]

If you have lots of data http://www.intellon.com/ is a good source. Their parts are pretty easy to deal with, although the firmware has a significant learning curve. The other thing is you may be able to get some good interface ideas from their data sheets.<p>Ron

[Rouen]

Hi thanks to all who replied to my question reguarding sending data via power lines.
I now have lots to work on.
For futher interest I will be sending this data on a 220V/50Hz line.I will try to interface with an infrared receiver four floors up to be used for interactive satellite. The user will have a infrared keyboard, at this time it is operated via the satellite decoder's IR input then through the RS232 port to a telephone unit on the ground floor, the user now wants to operate from upstairs.

[rshayes]

That sounds like the wireless intercom type of equipment would work. The chances are good that the entire building is operating off of the same distribution transformer (unless it is a large building).<p>Coupling to the AC line requires some care. The coupling network in a wireless intercom is probably a well insulated transformer coupled to the two sides of the AC line with small capacitors. The transformer will have to be rated for high voltage between the primary and secondary. Depending on the local regulatory agencies, this might have to be double insulated. Since it is operating at a couple of hundred kilohertz, this transformer won't need a large core or a large number of turns. The coupling capacitors will also be special, probably with a high voltage rating and flame retardent materials. A wireless intercom will aready have this type of network inside. You might find a few articles on wireless intercom designs in the popular electronic magazines of the 1960's or 1970's.<p>The earliest modems used for remote terminals used frequency shift keying of two tones. One tone was either 1070 or 1270 Hz (originate). The other was 2025 or 2225 Hertz (answer). These were meant to go through voice grade telephone lines. These tones could be passed through a wireless intercom, since they are in the middle of the audio range. These modems usually used RS-232 for the data interface, and would handle about 300 baud, which should be adequate for a keyboard. The modems used for amateur radioteletype communication (RTTY) would also work, and there were many construction articles for these modems in amateur radio magazines up to about the 1980's.

[MrAl]

Hello again,<p>Just take a look at National's AN-146 app
note. You can extract all the theory you need
from the schematics they present in that note.
You can use that to build up your own or use
theirs, which is already designed.<p>Take care,
Al

[rshayes]

Comments on the National AP Note (AN-146)<p>The connections to the AC line would give any standards organization (Underwriters Laboratories, DIN, or the South African equivalent) an instant heart attack.<p>First, no switch, and especially, no fuse. In other words, no protection in the event of a fault and no fast way of disconnecting power.<p>Second, the coupling capacitor is only specified as a 200 VDC part. The peak voltage on a 110 VAC line is very close to this under normal operating conditions. There is no safety factor. Transients on the AC line can easily exceed 200 volts. This part should be a type approved for AC line use and probably rated at 600 V or higher, even for a 110 VAC line. Your 220 VAC line will require even higher ratings.<p>Third, the transformer specified, TOKO type RAN-10A6729EK, has terminals that are only 7 mm apart, and there is a case connection between the primary and secondary terminals. The total creepage path across the part is probably about 3 mm. I would not consider this adequate even for a 110 VAC line. The internal insulation is probably the enamel on the wire. Two layers of enamel insulation is probably not adequate.<p>Fourth, the voltage on the collector of the output transistor in the transmitter is described as 40 to 50 volts peak to peak. The 2N2222 transistor has a maximum collector voltage rating of 60 volts. There is no safety margin.<p>Fifth, the TOKO transformer has a 5:1 turns ratio on the primary. The peak to peak voltage across the primary and the resonating capacitor will be 200 to 300 volts peak to peak. The tuning capacitor should be rated accordingly, at least 200 V and probably higher. The stress on the winding insulation may also be excessive.<p>Sixth, the circuit board layouts shown place the AC line connections close to other circuits. The creepage paths are probably too short even for 110 VAC lines.<p>The circuit itself probably works as described and would provide the transmission channel you are looking for. It just doesn't appear to take into account the things that can happen when you connect directly to an AC power line. Some of the parts (the 2N2222 and TOKO coil) appear to be very highly stressed and the design should be changed to reduce these stresses.