2
\$\begingroup\$

I found some DC power line comms circuits but they seem too complex for my needs. I have 6 outside LED RGB lamps that are linked with a single 2 core cable running around the edge of my garden, it is about 20m (60 yds) from the 12V DC power supply inside garage to last lamp in the chain. Each lamp is rated at 10w so max 60w from PSU means 5A max current. The lamps are controlled (on/off, colour, fading, etc.) by a small IR remote that only has a range of about 2m (6 ft), the IR goes through the lamp lens to an internal receiver. I want to be able to control all of the lamps together from a single remote. I have tried building a larger IR remote using 24 IR diodes and a 9V battery and this works to a degree, but still does not have enough range.

As the lamp housing is large enough to hold a further small circuit I thought that maybe I could add a small IR receiver inside the lamp and send the required signals over the 2 core DC power cable. I would put an RF receiver inside the garage and drive this from an RF remote (much better range), the RF receiver would need to send the signals to the lamps over the 12v power cable which each receiver would resend as IR within the lamp, so IR range is no longer an issue. Ideally the receiver would not consist of more than a few passive components and probably at least one transistor and an IR diode. Perhaps an FSK carrier at 100KHz on the DC line, carrier present = 1 (IR on) and no carrier = 0 (IR off).

I normally work with micros so my logic skills are fine but my RF and discrete skills are not very good. I want a simple solution, assuming one exists, otherwise I might just have to go back to building the mother of all IR remotes.

Update - I like the DCC solution suggested and on checking I see that the LED lamps are in fact rated as 12V DC/AC and so must already include a rectifier within them. I will monitor the IR signal to find out what the duration of a 'message' is how long each bit is. I have some IR LEDs rated as 1.5V 20mA so if I put one of these in series with a 560R resistor across the power lines inside the lamp then it should turn on when the power lines have the correct polarity. Can I assume that the LED will be safe when the polarity is reversed or does its high reverse resistance mean the entire 12V will be across it and damage it?

\$\endgroup\$
  • \$\begingroup\$ What chips did you find and why were them too complex? Building a receiver from a single supply with discrete components is probably feasible, but I'm not sure it's worth it. \$\endgroup\$ – Vladimir Cravero Oct 9 '15 at 16:39
  • \$\begingroup\$ What data rate is the IR link - 100kHz may be to slow to contain the modulation accuracy needed. Also you will need to add an RF blocker to the power side of the LEDs to prevent them getting upset by the presense of 100kHz or maybe attenuating it too much. You might also find that as the IR receiver is altering the LED brightness you get a lot of line noise and this will kill off the idea of a simple "relay" link. It can be made to work despite my observations. \$\endgroup\$ – Andy aka Oct 9 '15 at 16:54
  • \$\begingroup\$ Use the 12v to power a rf receiver. Don't bother with signaling over the power wire. \$\endgroup\$ – Passerby Oct 9 '15 at 17:45
  • \$\begingroup\$ Model railways use a system called DCC to control trains over 12V lines. \$\endgroup\$ – pjc50 Oct 9 '15 at 17:57
  • \$\begingroup\$ You could probably keep using the existing remote if you could focus the IR beam better. Try pointing the transmitter at a mirror and viewing the IR LED through your camera phone - most will pick up IR. Then rotate the transmitter and get some feeling for the viewing angle and luminosity. If you have any lenses or concave mirrors in your collection of junk you can try putting the LED in the focal point and get a parallel IR beam and see if that will travel the distance. \$\endgroup\$ – Transistor Oct 9 '15 at 18:34
2
\$\begingroup\$

One of the comments mentioned DCC, but since you want to keep it simple (i.e. understandable for you specifically), you might just have a look at DCC's physical design and invent your own from there.

Basically, all it's doing is providing constant power at some DC voltage and using the polarity of that voltage to mean something. So the circuit looks something like this:

schematic

simulate this circuit – Schematic created using CircuitLab

The transmitter's buffer and inverter need to be strong enough to power all the receivers, so I'd suggest making at least the final drive stage out of discrete transistors that are designed for that job. Search google images for "cmos inverter" to get a basic idea. You'll need one P-channel MOSFET to pull up and one N-channel MOSFET to pull down. With 2 lines to drive, that's 4 transistors total. When choosing FETs, pay special attention to switching times (shorter is better) and Rds_on (smaller is better). When your prototype works, keep a finger on the transistors. If they get too hot to hang onto, they need a heatsink.

The receiver's power supply is identical to a typical AC-in DC-out supply because it needs to accept either polarity. The signal is recovered by actually measuring the incoming polarity.

The diodes will drop some voltage, so if you really care about the final output, you'll need to use that much more at the transmitter end.

\$\endgroup\$
  • \$\begingroup\$ This looks like a promising solution and now that I have identified that the lamps will take 12V AC it could be the simple answer I am looking for. \$\endgroup\$ – TerryF Oct 11 '15 at 23:49
  • \$\begingroup\$ If they're actually designed for a 12V AC sinewave, then it's actually 12Vrms that they expect, or about 17Vpeak. And they may only care about the peak voltage. (carefully analyze my receiver circuit to see why) So if that's the case, then you'll need a 17V supply for your transmitter to match the intended operating point of the lamps. Or if they work just fine on 12VDC, then use that. \$\endgroup\$ – AaronD Oct 12 '15 at 3:36
2
\$\begingroup\$

60W from 12V is 5A, 5A from 12V is 2.4 Ohms, so any small signal you put on the wire is going to be slurped up by that low impedance, and the lamps themselves are probably emitting switching noise onto the power line too, you need to either find the frequencies emitted by the lamps and not try to signal on those bands, or make your signal strong enough to overwhelm them.

The simple solution is to run another wire and connect infrared LEDs to it, one for each lamp. it doesn't need to be a fat wire, doorbell wire would be fine, it's only 20 metres.

\$\endgroup\$
  • \$\begingroup\$ I am trying to avoid adding a 3rd wire as this is outside and is all waterproof, I have a 2 core cable connected through rubber glands at each lamp point and so I would need to rewire all of this with 3 core cable. \$\endgroup\$ – TerryF Oct 11 '15 at 23:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.