Phantom Power "floating back" to a second input, same device

[Herbert]

When filming, I use a Zoom F3 for sound recording. Synchronising the ‘fail-safe’ 32-bit files with the video footage from the cameras is PITA. And I come from a film editing background. I remember once having to synchronise 15 seconds of a speech recorded on 35 mm film with a 15-minute audio recording – as an editor, you simply learn to lip-read. Anyway, one way to avoid having to synchronise is to record directly with the camera. Fortunately, you can control the output of the ZOOM F3 – in stereo mode, you can even control both channels independently. As I record 90 per cent of my takes using a mono shotgun microphone (Sennheiser MKH-416) but work in stereo, the same signal is routed through the second channel, attenuated by around 12 to 18 dB, to provide a margin in case of clipping. As this cannot be configured internally on the Zoom F3 and there is a limiter on the output, I have to use a Y-cable for both inputs to which the MKH-416 is connected. At first, I thought I’d solder a circuit into the Y-cable to protect the second microphone input from the phantom power of the first microphone input. But then I decided to take a chance and tested the Y-cable on the F3 without a protection circuit. The first input supplied the voltage, whilst the second was switched off. And lo and behold, the F3 didn’t go up in smoke, there’s no distortion, and I can control both channels individually. But before I start doing this all the time: Why does the phantom power seem to have no effect on the second input? And what would happen if I accidentally switched on the phantom power for the second input as well? Best regards, Herbert.

 

[Devnull]

It shouldn't effect the second input. It's a 48V DC offset so the differential input subtracts it out leaving the audio signal. In theory turning on the phantom power on the second input shouldn't do anything because it's 48V coming from the same source.

 

[DonH56]

Phantom power is a DC source (nominally 48 V but specs also cover 12 V and 24 V, with some devices as low as 9 V) that is applied to both sides (+/-) of the output through a pair of fairly high-value resistors (usually 5k ~ 10k). Most mics and professional devices, and many consumer devices, include input coupling capacitors before the small-signal circuits that will block DC. Beware components that do not, since they may not say, though "DC coupled" listed as an advantage means no coupling capacitor is present. For consumer products, there is still danger, as the input capacitor may not be rated for 48 V (though most probably are), and ESD circuits (often diode clamps to ground) may be damaged by constant high'ish-voltage DC. Applying 48 V (or whatever) DC can damage unisolated inputs, so unless I know the device can handle it, I will not apply phantom power to it.

Shorting two phantom power sources does not usually cause damage since the resistors isolate the sources, but it not recommended by me (and probably most). Any offset between sources will cause power to be dissipated in the resistors, usually inconsequential even with fairly wide differences in voltage, but can also add common-mode noise and potentially a ground loop.

In my (painful) experience, "I tried it and nothing went up in smoke" is not the best experimental basis for wide-ranging conclusions.

 

[Herbert]

Thank you very much for your answers! As I mentioned earlier, these are the two balanced inputs of ONE device, each of which can be switched to 48V.
But doesn't that also mean, conversely, that they are already protected from the voltage / decoupled from it, right?
By the way, I used to have a portable phantom power supply that used ceramic capacitors to decouple the 48V. Since it was built using SMD components, relatively small capacitor-values were used. As a result, it acted like a high-pass filter. Also the Zoom F3 also sounds quite thin to my ears.
However, I haven’t yet tested a sweep tone on the inputs, which can also be switched to line level.

 

[LTig]

Thank you very much for your answers! As I mentioned earlier, these are the two balanced inputs of ONE device, each of which can be switched to 48V.
But doesn't that also mean, conversely, that they are already protected from the voltage / decoupled from it, right?
By the way, I used to have a portable phantom power supply that used ceramic capacitors to decouple the 48V. Since it was built using SMD components, relatively small capacitor-values were used. As a result, it acted like a high-pass filter. Also the Zoom F3 also sounds quite thin to my ears.
However, I haven’t yet tested a sweep tone on the inputs, which can also be switched to line level.

See the schematic of a (hopefully typical) microphone input:

C1 and C2 decouple the input from the phantom power voltage. Connect two of those inputs in parallel halves the input impedance, regardless whether phantom power is switched on for the second input or not.

 

[Herbert]

So, they are decoupled. The Zoom F3 has 3kohm input impedance, halved to 1.5k with my setup.
For Mono, I use a MKH-416 (15ohm) or ME-80(140ohm). I assume, with both mics a total impedance of 1.5k is ok.
Interestingly, the ME-80 runs on battery, but can also be powered with 48V. Sennheiser was so kind to confirm, though this shotgun consumer mic is vintage, but sounds surprisingly good when picking up speech in noisy environments… I tested the F3 with a sweep, starting at 20hz, directly from a source, response is flat.