Operating two carrier current transmitters on the same power lines is not simple, because the 2 transmitters would be driving RF into each other's outputs creating a potentially harmful (to the equipment) mismatch.
The now defunct LPB Inc. offered a solution in a combining unit that had to be custom ordered because it needed to be specifically built around the operating frequencies of the transmitters, but these devices have never been seen on Ebay and are extinct.
But today, here at KDX, with two carrier current transmitters and 3 coupling units we wonder what would happen if...
Transmitter # 1 were coupled to 1/2 of the 2-phase electric power;
Transmitter # 2 were coupled to the other 1/2 of the power lines.
This would put the 2 systems 180-degrees out-of-phase with each other, a decent amount of isolation so we estimate.
Another approach might be to couple one of the transmitters in the conventional way to the 220 VAC power lines and the other one coupled by neutral injection to a ground rod and the neutral buss of the power panel.
In fact, using both methods might allow coupling 3 transmitters on 3 different frequencies (?)
I wonder if I'm very smart, but I do not know.
A few comments about this:
"This would put the 2 systems 180-degrees out-of-phase with each other, a decent amount of isolation so we estimate."
This is true for the 60 Hz power line voltage and current but can not be true for RF signals of different frequencies so no isolation due to this is expected.
The technique used to isolate two transmitters sharing the same antenna, or in this case power line system, which prevents problems of one transmitter output feeding into the other is to use suitably tuned trap or bandpass filters on the transmitter outputs.
Another situation arises when the two signals coexist on a system with devices of unknown RF characteristics connected. Any non-linearity in the system, such as is produced by rectifiers or active devices, will result in mixing of the signals producing multiple sum and difference frequencies including modulation. This can produce strong potentially interfering signals on unintended frequencies. Other than being aware of this possibility there is not much to be done to prevent this short of removing all electronic devices from the power circuits. Such mixing products happen even with RF signals coming from other distant transmitters but the signal strengths involved are usually very weak and the mixing is not noticed. Two multi-watt transmitters in proximity and using the power lines most likely will produce noticeable mixes.
Keeping the Discussion Open
In addition to his other technical comments on this carrier current (CC) subject, Radio8Z said: "Two multi-watt transmitters in proximity and using the power lines most likely will produce noticeable mixes."
Neither disagreeing nor doubting the overall opinion posted previously, I am still visualizing the situation that would exist if the first of my ideas were put to the test for combining two AM transmitters into a 2-phase AC power system.
Each transmitter includes its own high-pass filter to roll off frequencies produced by the transmitter above the operating frequency; each transmitter would have its own coupling unit, which matches the 50-ohm RF outputs to a selectable range of impedances from 1-ohm up to close to 100-ohms, the tuning units containing SWR (Standing Wave Ratio) tuning controls which optimize the relationship between the transmitter and the line(s), and the Tuning Units act as a barrier to back-fed line-noise caused by the many circumstances mentioned by Mr. 8Z.
The two frequencies planned by member spareparts are 980 and 1080 kHz, certainly very close to one-another, and the "noticable mixes" predicted by Mr. 8Z would result from these frequencies interacting with each other.
The cross-bleeding of the two RF signals would happen largely by radiation/inductance across the separately phased AC lines, which would attenuate the criss-crossed signals by some amount, thus naturally reducing whatever mixing took place.
Adding a steep notch-filter to each RF output to reduce back-fed signals from opposing transmitters according to 15.209 measured at the distances stated in 15.221 would polish off the installation.
The weakness to the idea, if it were implemented and passed all other measures, is that radios plugged into AC outlets throughout the building would receive a stronger signal from the transmitter feeding that leg of the power circuit and a weaker signal from the other one, but the difference might be small enough to enable a suitable result.
Discussing technical issues as we are doing here is not meant to advise against using multiple transmitters rather to give some guidance for possible causes in case things do not operate as expected.
As Carl explained, the output networks and couplers used with each transmitter work in favor of isolating one output from another. Having each transmitter connected to its own phase of the standard 120/240 volt scheme used in household wiring might also help in keeping substantial RF power from flowing from one transmitter output to the other due to the pole transformer winding being between the two and is worth trying. 220 volt appliances such as clothes dryers, ovens, and central air conditioners will bridge the two "phases" together when operating so one thing to look for would be changes in each transmitter's operation when these appliances cycle.
It is most likely that reception won't be greatly affected by which phase the receiver is powered because the radio responds to radiated power more than RF conducted via the receiver power cord, though the power cord will certainly bring the signal close to the radio antenna.
Throughout a household both 120 volt phase wires run in close proximity and it is easy for an RF signal to couple from one to the other by capacitance. It is also likely that within a room both phases are present in the various outlets and there will usually be small enough distance from the RF hot phase to a receiver for reception.
My experience with a college CC system was gained with installations in dorms which used 208 volt three phase distribution. In such systems the 120 volts delivered is derived from one phase to neutral which, in such a system, is 120 volts. The transmitter output is connected between the neutral and one of the three phase 120 volt 'hot' lines yet the signal is heard regardless of which hot line provided power to a particular room. This was likely due to all three phases being available in the branch panels and the hallway lighting being shared among the three hot lines. Each room was not very far from each of the three phases so the RF radiated from one phase was received. In addition, the three hot lines running together to the panels likely had capacitive coupling so some RF was present on all three lines. In summary, the idea is to use the power wiring to radiate a signal close enough to a receiver sufficient for reception. (I suggested using the dorm wide PA system speaker wiring as the transmit antenna instead of the power line but the plant engineer wouldn't go along with this.)
The mixing mentioned will happen even with transmitters not connected to the power lines because the RF appears on the power lines which act as antennas. Whether this is a problem depends on the strength of the signals on the lines which is certainly greater if both transmitters are connected to the lines. In this case, it is worth looking for mix products at frequencies which can cause problems. Keep in mind that mixing can and usually does occur in the receiver itself.
The frequencies which result from mixing caused by non-linearities can be calculated. Let the two transmitter frequencies be f1 and f2. The mix frequencies will be mf1 +/- nf2 where m and n are integers. Fortunately, the strongest products will be for small integers so all possible m and n need not be used.
For the frequencies mentioned f1 = 980 kHz and f2 = 1080 kHz and some of the mix product frequencies would be:
980 kHz + 1080 kHz = 2060 kHz
1080 kHz - 980 kHz = 100 kHz
2x1080 kHz + 980 kHz = 3140 kHz
2x1080 kHz - 980 kHz = 1180 kHz
and so on. Note that the last one appears in the AM broadcast band. Whether or not these are problematic depends on the strength of each mix product There are also possible "higher order" products such as the 100 kHz signal can mix with the 980 kHz signal giving 880 kHz but these are likely very weak in strength.
Mixing is not simple but if you know it can happen it could explain observed signals where they don't belong. Carl, when you get the transmitters running it might be interesting to use your spectrum analyzer to check the spectrum since you might observe some mix products.
Straggling Thoughts
It is very unlikely I will try to send two simultaneous CC transmitters on the same powerline, the concept came to mind in consideration of the other thread where sparepart is planning to have 2 part 15 transmitters on frequencies not ideally suited for antenna radiation.
Regarding the 3-phase setup, the Coupling Units have 4 available output connections to the power panel:
1. Neutral;
2. Phase 1;
3. Phase 2;
4. Phase 3.
Any of them can be connected for any arrangement.
Uh, except that I did mention having the equipment on hand to think about some kind of combined operation, but at the end of the day, which it is right now, I'm not inclining toward doing it.
Appreciate as always the first class information typical of Part15.org.