Implementation of a capacitive coupling circuit for power line communication (PLC)

Question

Introduction:

Hello everyone, I just wanted to see if there is anyone out there who is familiar with power line communication technology that can help me figure out what is wrong with my LTSpice simulation. In order to communicate over a power line communication channel one needs to implement a coupling circuit to interface the power line with the communication hardware. While there are several approaches, I have chosen to use capacitive coupling with a 1:1 ferrite core transformer for galvanic isolation.

Coupling Circuit:

The coupling circuit needs to do the following:

1. Reject the low-frequency high-power power signal (230V, 50Hz).
2. Pass the high-frequency low-power communication signal.

So far I have an implementation that is capable of meeting the aforementioned requirements but only for reception. From most of the literature I've seen, the coupling circuit is usually used for both transmission and reception. I am not sure if there is a fundamental flaw with my coupling circuit design or if there nuance with LTSpice that is not allowing me to simulate the circuit properly.

To test that my circuit was indeed operating for reception, I did the following:

1. Created a voltage source to represent the mains voltage (230V, 50Hz).
2. Created a variable AC source to represent the message signal and to plot the frequency response of the circuit as a whole. Ideally I would like to me able to communication over the power line at a frequency of around 100MHz.
3. Created a behavioral voltage source that sums the message signal and AC source (something I was not able to do with components and is the underlying issue and the base of my problem).

Simulations:

The first setup involves testing the coupling circuit for reception. The message signal and power signal is added using the behavioral source to emulate transmission. We would expect to see only message signal at the node marked "Receive" with the power signal being attenuated by the included high pass filter.

Setup 1

Performing by making the message signal a variable AC signal, we can plot the frequency response of the coupling circuit resulting in the following.

Response 1

From this response it is clear that the lower frequencies (especially the 50Hz power signal) are well attenuated with the communication signal of higher frequencies being passed. From this we can confirm that the circuit works for reception. Now for the part where the problems start. Transmission of the communication signal.

NB. Remember that what we are trying to achieve for transmission is the addition of the message signal and power signal on the "Live" line.

The next step involves creating another coupling circuit for the purpose of transmission over the power line and is identical to the circuit used for reception. The setup is given by the following circuit diagram.

Setup 2

From this setup the message signal is applied on the right hand side to propagate over the power line channel. The signal passes the transformer and is present on the left side of inductor "L3" but does not seem to pass any further. The "Live" line simply consists of the 230V 50Hz sin wave with no additive message signal. From literature this seems to be the common approach but I cannot get a working simulation which has stopped me from progressing. Any and all help would be much appreciated, thank you.

References

https://www.imeko.org/publications/tc4-2014/IMEKO-TC4-2014-269.pdf

https://www.researchgate.net/profile/Timo-Korhonen/publication/228872526_Design_of_broadband_coupling_circuits_for_power_line_communication/links/00b7d53b527b1b22e9000000/Design-of-broadband-coupling-circuits-for-power-line-communication.pdf

https://ukzn-dspace.ukzn.ac.za/bitstream/handle/10413/12513/Mosalaosi_Modisa_2014.pdf?sequence=1&isAllowed=y

https://www.st.com/resource/en/data_brief/steval-xplm01cpl.pdf

Answer 1

The signal passes the transformer and is present on the left side of inductor "L3" but does not seem to pass any further. The "Live" line simply consists of the 230V 50Hz sin wave with no additive message signal.

How can you hope to transmit a message voltage onto a line that is firmly fixed by the 230 volts, 50 Hz voltage source V1: -

In other words your message signal is "seeing" the equivalent of a short circuit (V1). Your fixed voltage source should be made to have an output impedance that is negligible at 50 Hz and significantly higher at your message frequency.

Answer 2

First of all, you can use the same line coupling circuit for both RX and TX. You don't have to separate them.

Depending on the band you are working on (e.g. PRIME) and the load you are trying to drive (can be as low as 2 Ohms - basically you can think of the load as the distribution transformer), you may need to place a series inductor:

^{simulate this circuit – Schematic created using CircuitLab}

Above is a simplified schematic of a coupling block. Magnetising inductance of 1 mH is a typical value for PLC coupling transformers (NB: None of the bands' operating frequency reach 1 MHz - typically 40 kHz to 150 kHz depending on the band):

In your case the load is the voltage source (possibly V1) which is theoretically 0 Ohms. So it's quite normal that you can't transmit any messages. You do need to model the voltage source properly and consider using the series inductance.

---

One of the companies I worked for in the past had spent years on this technology. Achieving a stable operation is really difficult on real world environments even if you use DSPs and AFEs (Analog front-end). Be ready to lose a lot of hair if this is going to be a real-world product.