How to Couple PL Carrier Across Two Phase Power

The new circuit is:

Looks to me that you have a capacitor in the wrong place (the one at the top of the pic).
Phase A doesn't come out except through the cap, which can't be right.
You can inject your signal via caps, but the phases still have to come out to be used for power.

The right most signal line seems ok, it is coupled by the cap into phase B.
Why doesn't the left most signal line just go through a cap and into phase A? So it's all nice and symetrical... maybe I'm missing something?
Del
(I reserve the right to be completely wrong on this as it's not my field and I think the whole concept of signal over power lines is V iffy at best)

Quite.

It's not my field either, but what I've learned by following links from Wal's transformer suggestion in the earlier thread by this name (namely the Etal P2820 transformer datasheet, the power opamp datasheets for the TLE2301 suggested by Etal, and TI's replacement part, the AFE031, and further Googling EN50065 CENELEC), is this:

The critical way to get your signal enforced on the AC line is to first use a low-impedance power opamp (capable of delivering 1 to 1.5A, with reasonably-fast slew rate and with 3-state capability), and add a coupling step-up transformer with a series capacitor. The capacitor's value should resonate with the transformer winding's leakage inductance, at the frequency of the signal modulation (e.g., 100kHz).

This is very important, because of the "zero-impedance" nature of series resonance. At series resonance the coupling impedance will be low: just the esr, or equivalent series resistances, of the transformer and the capacitor, plus the opamp's low Zout. It's this low esr that gives the power amplifier a chance to force its signal onto the AC line. Keep in mind the AC line may have all kinds of devices plugged in, some with filtering capacitors that have a low impedance at 100kHz.

Etal specifies the leakage inductance, or L-sub-ell, as we call it, of their transformers. Sometimes they show breaking the capacitor into two parts (one on each side of the transformer), and even adding some series resistance.

Studying Etal's transfomer specs, I see they have very low inter-winding capacitance, and a rather high leakage inductance, as a fraction of the magnetizing inductance. The latter is no doubt a concomitant requirement of the former. But thankfully, it does help lower the coupling capacitor's value (see formula), and thus the maximum-possible inrush current at power-up.

One other important thing: the resonating capacitor must be an X-rated type.

So far as I can tell, you'd have to use a separate coupler for each phase.

If your goal is to couple an existing signal on one phase to another, perhaps a single coupler with a 1:1 transformer could do the job (right), but I didn't see that possibility discussed.

Hmm, I wonder if the opposite AC-line phases should have out-of-phase signal connections?

In the single phase application the N was your reference.

You have no reference here for similar unbalanced transmission.

Can you show your single phase arrangement please?

This is raising some interesting Qs.
From my feline perspective I don't see why you need a reference as long as you are injecting and extracting the signal from the same two wires. The signal doesn't 'know' or 'care' about what's on the wires or their relative phase as I see it.
Does it even matter if the generator back at the utilty is actually running. Yes, I do realise the impedances will be different between a stationary generator and a running one but in the grand scheme of things the impedance is pretty low regardless.
Or am I oversimplifying (again )?
Del

as long as you are injecting and extracting the signal from the same two wires.

Exactly right Del.

You need a transmission line not an antenna. The transmission line can be two phase conductors to form a balanced line or one phase and decked neutral in an unbalanced line.

I'm struggling to model the concept for the OP here. "reference" may not be the best way to describe it. Jump in if you can...

Hi Wal, Win,

I posted a sketch below.

Wal, my single phase arrangement is Fig 1. It has been verified working ok.

Hi Win, Wal,

I came up with circuit in Fig 2 from Win's post. It make use of two PL coupling transformer in the single phase application but arranged in center tap configuration.

Is RA and RB needed for the center tap to derive the "neutral"? Can the circuit function without RA/RB? Does it make the coupling circuit more stable?

What is difference between the circuits in Fig 2 and Fig 3 (which uses one PL coupling transformer)? How will primary side PL carrier superimposed on midpoint neutral (60 degree phase shift wrt zero degree) appears like after it is transformed to the two secondary coils? Will the PL carrier be seen riding on the "neutral" sinewave with the 60 degree phase shift (with a 1:1 turns ratio for the PL coupling transformer)?

When receiving, will the PL carrier waveform on secondary coil be different for the circuit in Fig 2 and Fig 3? I mean the peak to peak voltage swing? My thinking is the peak to peak voltage swing for the circuit in Fig 2 is closer or more like that in the single phase coupling circuit in Fig 1. Fig 3 will have largest peak to peak voltage swing on which the PL carrier is riding on. Is this correct?

Many thanks!

CW

peak to peak of the carrier envelope compared with

Fig 1, 2 and 3:

None of the above.

For the 100kHz circuit the 50 or 60Hz arrangement is simply irrelevant. The only thing, it can be said, that the impedance presented is undefined, and variable, as 60 Hz loads are switched in and out. Additionally, the low voltage circuits need real good protection from spikes and switching transients.

Other than that, the 100kHz has nothing common with the power circuits. For brevity's sake I assume a10 - 100 Ohm variation on the lines at 100 kHz. Then it does not matter, if you drive it with current or voltage. The amplifiers will need to have AGC (automatic gain control), to keep line voltages level. Unless the level is controlled on the line, your design will go nowhere.

By the way, I suggest to rethink the proiect. You may or may not need a set voltage on the line. The goal is an assured reception on the far end.

For experimentation off the shelf amplifiers are expensive, but well working starting points. I prefer voltage feed, as quality HV capacitors are available off the shelf.

Hi Win,

Thanks for the help. Your proposed circuit work best and more closely similar to existing single phase coupling.

Leveles,

You are right! The primary winding inductance will give low impedance for 50Hz power frequency. This translates to very low voltage and frequency across the secondary winding which can be easily blocked by the 1uF capacitor at secondary output.

Once Again, Thanks All