All switch-mode power supplies make radio-frequency noise regardless of whether the output of the supply is a sine wave, a square wave or dc or whether the input is dc a sine wave or whatever. Power inverters used in boats and recreational vehicles are subsets of the general class of switch-mode power supplies. The quantifiable amount of RFI generated varies with design, implementation, and installation
The designs which minimize the generation and propogation of noise cost more than virtually any product which you will see available (at least now and the near future) and affordable for use in the marine
market, especially the high power units which are bi-directional in that they operate either as a charger
or as an inverter. Just as a comparison, we "enjoy" being able to purchase
an inverter/charger for less than a dollar per watt (in the USA) yet if you purchase
a "relatively simple" dc/dc converter which complies with international standards for conducted and emitted RFI (radio frequency interference) you might pay 50 dollars per Watt.
Conducted RFI is interference which is generated through the conductors (either input or output) and anything connected to such wires must have internal rejection of such interference to escape its effects. Most marine
and RV installations suffering audio system and radio
system interference suffer from conducted RFI first and emitted RFI second. Because there are many boats with installations suffering essentially no problems with their audio, radio, tv, and computer systems (having your specific model of inverter/charger) the first "thing" to look for is just how your unit is wired in that might effect conducted interference.
Emitted interference is the most difficult to solve and the "easiest" method is merely to provide as much geometric separation between the source and the reciever which means separating the wires and units as far as possible to take advantage of the phenomenon that the susceptibility decreases as the reciprical of the distance squared. The easiest method of preventing the generation of interference lies with the design and physical packaging....not something that you can likely change. Shielding of either the source or receptor is always bandied about within the marine community yet the truth is it is a waste of time because of the difficulty in being able to do a proper job (which very few people are capable of doing either in theory or in practice) especially when considering the level of power that has to be dealt with in the case of inverter/chargers. Really effective shielding uses continuous enclosure of the source with material that shields both the electric
and magnetic fields....that means using ferrous materials and magnetic materials...who has that capability? (that means that aluminum
foil and copper braids, etc. just will not satisfy the total requirement).
Twisting together the input power positive and negative cables introduces coupling between the two that takes advantage of self-canceling fields (called differential mode cancellation) caused by the "gulping" of power that the inverter takes from the battery in the inverter mode or the battery takes from the charger
in the charger mode. Adding capacitance between the two cables filters the voltage fluctuations as well (if needed).
By making sure that the inverter/charger power cables do not share cable runs with your susceptible loads eliminates "ground and power loops" (power loops are the mirror-image of ground loops) you will minimize the voltage fluctions induced into your suseptible receivers caused by the inverter/charger. That is why I advise single
continuous leads from the inverter/charger all the way to the battery terminals. Loads should be only attachted to the battery terminals on TOP of the power (inverter/charger or other heavy drain potential sources of interference like refer compressors). No other loads or sources should share wiring
with the inverter/charger all the way to the battery terminals. The ONLY exception is a short cable from a battery monitor
shunt to the battery negative. All wires must be attached to the load side of the shunt in order to allow the monitor
to "see" all sources and loads to/from the battery.
How are you doing so far?