Phantom power hum depends on mic?

[mhelin]

I have built a Neve clone and powering it from a 48V SMPS. For some reason on one mike (Violet Black Knight, transformerless output) there's very strong hum (50 Hz buzz) but on cheap t.bone SC450 (transformer output, Feilo?) none. Why is that?

 

[JohnRoberts]

common mode rejection can be degraded by imbalance or impedance mismatch between mic + and - outputs.

I do not know what a Violet Black Knight is, but if you can, check the impedance balance of the two outputs.

Caveat this is not a simple VOM measurement as you are interested in impedance at hum frequencies.

JR

 

[gyraf]

..measure voltages at XLR to find difference in voltage on transformerless?

Jakob E.

 

[mhelin]

Voltage was 41.6V on both pins (positive and negative) with the Violet mic though this was on my second Neve unit (similar though).  However,  the friend who borrowed the one with worse problems (and PSU) did have success using another transformerless SDC. Maybe it's something to do with Violet mic, got to check it or ask JZ if he has any ideas. 

 

[mhelin]

Shouldn't the voltage be the same with and without the load? Maybe I have there a too large filter resistor, got to check it.

 

[CJ]

might be a filter cap inside the quiet mic,  check phantom supply for ripple,

 

[Tubetec]

Maybe theres a bad interaction between the switchmode phantom supply and that particular mic .
I seen several incidents where certain switcher supplies just dont play nice with certain other equipment ,even though there shouldnt be much low frequencies in the outputs sometimes Ive seen cases where when two switchers interact badly and there ends up a ton of 50hz honk and rf buzz superimposed on the audio lines ,maybe try some extra CLC stages on the outputs of the phantom switcher .

 

[PRR]

> Shouldn't the voltage be the same with and without the load?

No!

If that were true, it would stay 48V into a shorted wire, pump infinite current, clocks would run backward (or something would smoke).

Look how Phantom works. The 48V has a 3.4K series resistor (often two 6.8K, one each leg).