Design questions - intercom interface

Thread Starter


Joined Mar 7, 2017
Screenshot 2017-03-10 22.12.24.jpg
I am relatively new to electronics and new member here. I'm designing an interface to enable a passenger in a private aircraft to listen and speak on the aircraft intercom system using her Bose noise-cancelling QC20 headset, while simultaneously listening to her iPad through the same headset. The iPad audio must not leak into the aircraft intercom system. The interface will also provide power to charge the QC-20 battery. The schematic is attached.

I am open to advice and have some specific questions.
  1. I'm confused about the numerous "grounds" involved. The aircraft interface provides 28v DC relative to "ship's" ground, and the aircraft intercom audio uses, I believe, an independent ground (I have not checked for a voltage between the two). The iPad has its own indeterminate ground level - it's usually on battery power but conceivably could be connected to a charger plugged into a 110v AC source from the aircraft's "inverter" power supply (not sure of the details on this). I've attempted to deal with these as follows:
    • Isolate the aircraft 32v DC power ground. It sees the L1 choke and DC-DC voltage converter input (Recom RS-6 2405D) but nothing else.
    • Connect the aircraft audio ground and the power converter's ground.
    • Put a high-pass filter on the intercom audio in lines (COMM L&R) to block any DC offset.
    • Put an optional high-pass filter on the connection between the iPad ground and the audio ground/power converter ground described above. Using JP4, this filter can be bypassed and the iPad ground connected to the interface's ground, or the iPad ground can be left floating.
  2. The 100nF/10K high pass filters have a -3dB freq of about 160Hz. But if what I'm trying to do is block DC, is there any reason not to go much lower (like I tried to do with the circuit off JP4)?
To clarify some oddities with the schematic, the crossover wiring between the two op amps is to make PCB layout simpler. The "Vol" jumper is to accommodate a volume control if it's needed, but I anticipate it will just be jumpered 1-2, 3-4, 5-6.

Many thanks,


Joined Aug 1, 2013
Nice work so far, especially for someone relatively new. You've covered all of the basics well.

Add both ceramic and aluminum electrolytic capacitors between the common mode choke and the input to the ps. PS's like (spelled: require) a low source impedance.

Don't know about small plane GND wiring. I would assume that signal GND and chassis GND are at approximately the same potential, but kept separated for noise isolation.

On the left side of the drawing, mic *in* and iPad *in* signals are in different directions.

Consider swapping U2B and U3A so that each signal channel is contained in one chip. Better crosstalk isolation, and one component failure doesn't wipe out both icom channels.

An iPad charger is fully isolated, so there is no change in ground paths when running on the charger or off.

Unless there is a low frequency noise component coming from the engines, I don't see any reason not to decrease the input highpass corner freq.

The primary function of a schematic is to show the electrical, not geographic, relationships among the components. I recommend uncrossing the crossover wiring; there is nothing wrong with having the 1-2-3 opamp on the top of one stack and the bottom of another.


Thread Starter


Joined Mar 7, 2017
>>Nice work...<< Thanks. All credit goes to The Art of Electronics and Google.

>>mic *in* and iPad *in* signals are in different directions<< yes, the mic signal goes from the Bose headset/mic to the intercom (right to top left) and all the other signals run left to right.

>>Consider swapping U2B and U3A...<< Great idea!

>>uncrossing the crossover wiring<< Ha! When I originally selected the LM833 component, Eagle laid it out that way and I assumed the A and B sides were locked in that configuration. But they're not. Looks much better.

Ok, I'm going to order the PCB and see how it goes. Many thanks.