Balance Mic and Speaker for Camera Input with Ground Loop Isolation

Question

Novice looking for some help.

I use a Cardo Bluetooth headset to communicate with friends and record the audio onto an action camera while out riding motorcycles. By trial and error I was able to come up with this circuit which has worked reasonably well for years.

I'm taking the my audio from the Cardo communicator microphone and combining it with the audio from my friends over the communicator (Helmet Speaker). A voltage divider (again by trial and error) gives me more or less balanced audio for my action camera with stereo Mic Input.

There is NO CONTINUITY between the Mic and Helmet Speaker. Without the isolation transformer the Communicator shuts off. The circuit needs to be very small (to tuck into a helmet) and be passive since I can't provide additional power. I suspect my current circuit gives me 20mins less use on the communicator but this is just based on observations with other units.

My action camera (Sony FDR X3000) takes a stereo Mic input. It seems to do some audio balancing internally but if the Helmet Speaker volume is very high or the Helmet Microphone is very high I occasionally get problems (one of the channels gets crushed), but under normal conditions it works well (example below - no gain adjustments on my PC).

This works most of the time but is there a better / simpler way to do what I'm trying to do? Unfortunately I don't have any fancy gear like an oscilloscope.

Answer 1

I don't think you will improve much on what you have done here. The transformer is needed as the mic and LS common connections on the cardio are not connected, as is the resistive divider to level match.

It seems from the discussion at this link the Sony cam has some auto gain control (AGC) on the external mic in, which by the sound of it sets the signal rather high. You can't do anything about this, it is built into the Sony. The AGC will take the average of left and right mic input signals and adjust gain - that is why you see the engine noise from the helmet mic dip when someone talks over the communicator. You could play with the values of the two attenuator resistors, if the amount of dip is too much you could (for example) change the 100R to 47R (approximately 6dB less level) which will reduce the amount of AGC gain reduction. But quite likely this is not really a problem and you most likely already tuned these two resistor values. I think it is good to record mic and cardio to separate mic channels as this allows you to alter them independently in post (some video editor like Davinci Resolve would allow this).

But given that the interface needs to be passive and must isolate the mic from the speaker, I don't think you can simplify this any further. It's a pretty good job actually, given the restrictions you are under.