Just to clarify, when the load on one socket causes the voltage on a different socket to sag, that is not because of the 6.8 kOhm resistors on either socket. I never said or implied that in any way!
This was indeed the way I (mis)interpreted your original post - it was late in the evening and I did not read it carefully (and did not understand phantom power supply in the first place), before making a hasty measurement and post. My bad and I am sorry.
I am relatively new with recording and recorders, and the R-44 is my first and only recorder I have so far - bought mine in the first week it started general shipping in the US (s/n ZW80244) based on the specs only (without reading any reviews). Never made any measurement until the other day on the phantom power supply, and more measurements today – Here is a summary:
- The phantom supply indeed has a 6.8K for the pin #2 and a 6.8K for the pin #3 – pretty conventional and nothing unusual
- Have a pair of AKG 414, a Rode NT-4 (stereo mic with two XLR connectors), a Rode NT1A, a number of pots, and a half dozen of the Switchcraft XLR barrels (left over from a dozen I bought for making various attenuation pads) – they come in really handy.
- I used the mics with the pots in parallel to draw the exactly 5mA in each connector in the previous measurement
- In today’s measurements I shorted pins #1, #2, and #3 – measured current 14 mA. I did this to both input 1 and input 2, drawing 28 mA in total. The phantom voltage in this case dropped to 47.3V. This is pretty good compared to the spec (20 mA max current, 48 +/- 4 volts). Did not go further because I did not want to risk damaging the circuitry.
- Also measured the output noise floor today as below:
o Used a RF spectrum analyzer to measure the output in 20Hz – 20KHz only
o R-44: level setting centered, 16 bits/48 KHz (what I always use in recording), no filtering
o The EIN is calculated by assuming the gain is -20 dB - the “sensitivity” setting (i.e. for sens = -44 dB, the gain = -20 – (-44) = 24 dB). Not sure if assumption and/or measurements are correct.
Sensitivity Output noise, EIN dB
Setting
4 -123 -99
-2 -122 -104
-8 -121 -109
-14 -122 -116
-20 -122 -122
-26 -118 -124
-32 -112 -124
-38 -106 -124
-44 -100 -124
-50 -93 -123
-56 -87 -123
If the measurements are correct, then the actual noise performance is pretty good compared to the specs and the results by some others (will repeat the measurements using a proper audio analyzer at work next week). Not sure about variations from unit to unit though, in terms of both the phantom supply and the noise floor.