Power Line Carrier Circuit - 12VDC

Thread Starter

geekoftheweek

Joined Oct 6, 2013
885
I am wanting to build a power line carrier circuit. I found a simple circuit I am hoping to build from, but have a few questions. I have looked at several modules online, but have not yet found one that seems to fit what I need. I don't have access to the exact specification at the moment that I have to follow due to it costs a bit of money and I'm trying to sort out if it's going to be worth doing it or not.

plc-bare.png

The nominal voltage on the left of the transformer will be 12V, but can be as low as 8 and high as 16. It will be normally powered from an automotive battery on a test bench. I believe the actual voltage swing I need to create is 2V below and 2V above the incoming voltage, but will know more once I can access the specification.

Now the right side of the transformer if I use a Vcc of 2.5V would I then see .5V to 4.5V corresponding to the 4V total swing of the incoming power assuming a 1:1 transformer? The next question is as the circuit is drawn I am thinking powering the transistor would then induce a change in the incoming power. What kind of change would it induce? I am guessing it would either raise or lower the incoming voltage one way or the other depending on how the transformer is wired. Would it then swing the other direction due to some sort of field collapse, flyback, or whatever? How do I go about selecting the transformer?

This circuit was found in the application information of an LM2893 https://rocelec.widen.net/view/pdf/cbmtgxstta/NATLS16609-1.pdf?t.download=true&u=5oefqw.

Thanks for taking a look
 

Reloadron

Joined Jan 15, 2015
6,987
This is the full typical application circuit:
LM1893 Carrier Current Transceiver.png

Now what exactly in detail is your question? The chip uses FSK (Frequency Shift Keying) the left side of the transformer is your data out to mains supply voltage. The right side of the transformer is your AC mains with the data capacitively coupled.

Ron
 

fileandfit

Joined May 20, 2022
8
Only have questions, I'm afraid. How is an automotive battery (DC) going to work in the transformer primary? What sort of information do you intend to transmit via the mains cabling?
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
885
Only have questions, I'm afraid. How is an automotive battery (DC) going to work in the transformer primary? What sort of information do you intend to transmit via the mains cabling?
I found a different schematic in an out of production IC that was built for this purpose and lost it. I do remember it had a transformer as well as a few inductors in it. I'll have to see if I can find it again.

I was hoping this could at least be a starting point.
 

LesJones

Joined Jan 8, 2017
3,816
I think you have the wrong idea about how the coupling circuit in your post #1 works.
1. The left hand side or the transformer is the mains side.
2. The transformer is NOT a normal 50/60 hz transformer. It is designed to work at the carrier frequency which for the device in your link is about 125 Khz. It couples this frequency signal into the mains supply. This transformer will couple almost nothing of the mains frequency to the right hand side.

Les.
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
885
I think you have the wrong idea about how the coupling circuit in your post #1 works.
1. The left hand side or the transformer is the mains side.
2. The transformer is NOT a normal 50/60 hz transformer. It is designed to work at the carrier frequency which for the device in your link is about 125 Khz. It couples this frequency signal into the mains supply. This transformer will couple almost nothing of the mains frequency to the right hand side.

Les.
Thanks Les. Instead of mains power it will be 12VDC. I'll work on getting some more details figured out.
 
Last edited:

Reloadron

Joined Jan 15, 2015
6,987
I think you have the wrong idea about how the coupling circuit in your post #1 works.
1. The left hand side or the transformer is the mains side.
2. The transformer is NOT a normal 50/60 hz transformer. It is designed to work at the carrier frequency which for the device in your link is about 125 Khz. It couples this frequency signal into the mains supply. This transformer will couple almost nothing of the mains frequency to the right hand side.

Les.
Les, looking at the drawing it looks to me that the left side of the transformer is signal and the right side is where it connects to mains. Further illustrated on page 4-172 of the linked data sheet. The chip itself is bi directional and uses FSK. It's just a bi-line chip to interface with the power line.

Thanks Les. Instead of mains power it will be 12VDC. I'll work on getting some more details figured out.
So you plan to modulate 12 VDC with data? This would likely go much better with a full description of your project. The chip in question here is designed to modulate an AC carrier. The FDK example below.
FSK.png

Again, what is the project in detail.

Ron
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
885
I borrowed a device I plan to communicate with this morning, connected it up to my lawnmower battery (it was closest and easiest), and put my DSO to work.

abs.jpg

I may have been wrong in my initial research (still haven't paid for the specification yet). What it looks like to me is a possibly 16 bit chirp followed by 8 bits of data. I also seem to have been way wrong about the total voltage swing... looks more along the lines of 1V total. I have never calibrated the DSO so the voltage readings are off.

I thought it would be more fun to build my own, but maybe I should take another look into already available units and skip the headaches that will arise out of the project.

I still am curious how I would go about building one so in the interest of learning I will from time to time work on my own design for kicks.
 
Last edited:

Thread Starter

geekoftheweek

Joined Oct 6, 2013
885
@Reloadron I was really just using the schematic for ideas. I'm not actually interested in the chip itself as I plan to use a PIC of some variety in the end project.

I am far out of my normal comfort zone on this and thank all for your input!!
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
885
Thanks @nsaspook I was actually wanting to find that page again. I disregarded it earlier since I may possibly have up to five devices connected at one time and won't be able to physically add and remove L2 each time. This is part of a diagnostic tool redesign to add more features.

Edit

I just realized I was looking at L2 wrong. What I was really referring to was "Multi-drop configurations are possible as long as each node includes an inductor to maintain the high impedance of the data-carrying portion of the power line. The inductors can be small surface-mount devices, as long as their current ratings, with a generous margin, are sufficient for the load(s)."
 
Last edited:

nsaspook

Joined Aug 27, 2009
10,427
Thanks @nsaspook I was actually wanting to find that page again. I disregarded it earlier since I may possibly have up to five devices connected at one time and won't be able to physically add and remove L2 each time. This is part of a diagnostic tool redesign to add more features.
With typical DC sources and loads the power link impedance will be very low resulting in low carrier signal levels so you must have some way to isolate the wire transmission line reactive impedance from the DC power resistive/reactive components in a practical circuit. It's possible to design RF circuits for PLC on DC lines without signal isolation but they are usually more complex and use E-line type signal coupling.

https://en.wikipedia.org/wiki/Single-wire_transmission_line#E-Line
 
Last edited:

ronsimpson

Joined Oct 7, 2019
2,404
Last time I sent data over the power line (DC):
On the power side you have 12V and often a capacitor C1 which will short out high frequency data.
On the load side you often have a load with a capacitor C2 which both want to short out the data.
L1 and L2 make the power line "open" at high frequencies. The DC load passes through the inductors just fine.
We were able to run the power over a long distance to many houses with out much loss.
Pick L1 and L2 so they will not saturate at the max DC load current.
Pick L1,2 so the impedance is 100 ohms (or something) at the data carrier frequency.
1653509164514.png
This way the battery and the load only see DC and the line sees the AC data on top of the load current.
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
885
Thanks again all. I am for sure in over my head, but have spent some time searching out looking up various things mentioned.

The average and most common distance I will see is around 30 ft. The max (if connecting five devices) would be around 120 feet.

The devices call for a 30A continuous circuit, but my observations show a typical draw of less than 1A. I will have the power cut out at 20A in the end and signal a short circuit.

Will get some other details around after work in the morning.
 

Reloadron

Joined Jan 15, 2015
6,987
Thanks again all. I am for sure in over my head, but have spent some time searching out looking up various things mentioned.

The average and most common distance I will see is around 30 ft. The max (if connecting five devices) would be around 120 feet.

The devices call for a 30A continuous circuit, but my observations show a typical draw of less than 1A. I will have the power cut out at 20A in the end and signal a short circuit.

Will get some other details around after work in the morning.
ESP32 or any data acquisition and use Bluetooth. No wires. Send data easily the distances you mentioned. This if all you really need to do is send data from point A to point B. Bluetooth modules are plentiful and inexpensive. Long as you have a wireless network.

Ron
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
885
A little more background...

power.jpg

The plan is to use a MIC5021 gate driver with over current protection to drive a NFET with a MAX4316 following all that to monitor current. My current setup uses a relay instead of a FET to toggle power to the circuit, The shunt accross the MAX4316 still needs to be determined. I think I used .001 in my current design, but found it's not very accurate due to the normally low current draw. Are there any effects of this setup to consider?

Is there a good page off hand anyone knows of to begin to understand the math involved with choosing inductors and capacitors for this? I know I have seen various pages now and again,and even done some experimenting with capacitors, but I am far from being comfortable at the moment. Since the signal will vary from 100kHz to 400kHz would it be safe to assume something in the middle will get the job done? I will look things up later on and see what I can come up with.

I will experiment with https://www.electronicdesign.com/te...cuit-communicates-over-lowvoltage-power-lines mainly to get my feet wet and see what I can actually make happen. Since the PIC will always be powered even when the final circuit is not the voltage regulator won't be part of the circuit. The above drawing will represent the incoming power supply.

@Reloadron My current rig uses an ESP8266 with a custom WiFi socket to communicate with an Android app I put together. The ESP also communicates with a couple PICs via I2C to run a display, monitor switches, turn on and off relays, and measure voltage and current. I am going to eliminate the display (it was a neat feature, but in all reality a pointless one), and probably upgrade to the ESP32. I think I will also upgrade the PICs to 24F something in the new version to handle 16 bit values easier and faster. I have had an ESP32 in the toy box for a while now, but still have yet to get it out and play with it. Maybe should look in to what all it offers that the 8266 doesn't.

@ronsimpson would Cv2 be the point to connect to for communication then?

In all reality transmission is not a requirement. I could just use a couple voltage dividers and feed it in to a PIC's comparators to skim off the data and call it good, but where is the fun in that?!?
 
Top