52 Volt Phantom Power

[huub]

Kay.. I really have no electronical knowledge, but my practical experience tells me this:
With long cable runs ( 300 meters or more) something is lost; on a Qbox the fantom power lights become less bright, and some mics, that are more critical to phantom power, stop working, or become problematic.. Does this mean that with long cable runs some voltage is lost, or something else?

[fifthcircle]

FWIW, My now-departed colleague who designed the Vac Rac preamps designed them to output a 52 volt supply to make up for losses in long cable runs and to make sure the mics out there that are voltage/power hungry would work to their full capability. Didn't seem to hurt the ones that didn't need the full 48 volts either...

--Ben

[nosebleedaudio]

Originally Posted by David Rick

Changing the ballast resistors by 10% is not going to overwhelm the output stage of any competently-designed microphone. Mics already see much greater impedance differences between one preamp and another. For instance, the input impedance of a Millennia preamp is significantly higher than that of many other preamps, and this causes some mics to sound different (for better or worse). There are several other preamps that have selectable input impedance to exploit this effect.

For microphones with constant-current biasing, raising the phantom voltage raises the supply voltage seen by the microphone by an identical amount (4V). Reducing the ballast resistor raises the supply voltage seen by the microphone by a smaller amount. For a microphone drawing 6 mA, the increase is about 2V. If you really want a 4V increase, you could reduce the ballast resistors by 20% (put 34k resistors in parallel). I can't seem to find any record of what I did on the Millennia preamp I use with my Josephson mics, but I did feel it improved the sound.

There's certainly nothing inherently more dangerous about changing the resistance compared with changing the open-circuit voltage. The big advantage of the former is that you can easily have a switch to disconnect the parallel resistors when using a mic that you're concerned about. This should be a very easy modification.

Raising the voltage can be easy or hard depending on the design of the phantom power supply. In the best case, you'd only have to change one resistor, but the new value might be hard to calculate if you don't have full service documentation. In the worst case, you'd have to rewind a power transformer.

David L. Rick

I think you just answered this, the mic pre impedance can have more effect on this than lowering the resistors.
The fact is the more you load the mic with a given level out the more current it demands from the mic. This is ohms law...
I spent over an hour the other night playing around with different levels/loads and simply put, if you want you mic to perform at its best use a higher input impedance and the max 52 volts...

[Testing123]

Originally Posted by huub

With long cable runs ( 300 meters or more) something is lost; on a Qbox the fantom power lights become less bright, and some mics, that are more critical to phantom power, stop working, or become problematic..

I'd be curious 300 meters of what? I grabbed the spec for a typical snake cable (Belden 1514C - 16 shielded pairs) and it states 23 ohms per 1000 feet for the conductors. The shields are only 16 ohms per 1000 feet. Since both conductors are used to pass the phantom power, the effective conductor resistance is half of the resistance of one. That gives us a loop resistance of (23 * 0.5) + 16 = 27.5 ohms for 1000 feet.

The spec that gives the paramaters for phantom power is IEC 61938, and it specs a max of 10mA current draw for a phantom powered device (which most mics are well under). So at the max current draw we'd have a 27.5 * 0.01 = 0.275 volt drop. Our phantom powered device would see 48 - 0.275 = 47.725 volts. Doesn't seem like that should make much difference.

As for the QBox, I'd wonder if they are exceeding (maybe by a lot) the 10mA current draw to light the LEDs and do whatever else it is they do in that box.

Edit: you guys might find this article interesting:

http://www.reddingaudio.com/schoeps/...le_phantom.pdf

[nosebleedaudio]

Originally Posted by Testing123

I'd be curious 300 meters of what? I grabbed the spec for a typical snake cable (Belden 1514C - 16 shielded pairs) and it states 23 ohms per 1000 feet for the conductors. The shields are only 16 ohms per 1000 feet. Since both conductors are used to pass the phantom power, the effective conductor resistance is half of the resistance of one. That gives us a loop resistance of (23 * 0.5) + 16 = 27.5 ohms for 1000 feet.

The spec that gives the paramaters for phantom power is IEC 61938, and it specs a max of 10mA current draw for a phantom powered device (which most mics are well under). So at the max current draw we'd have a 27.5 * 0.01 = 0.275 volt drop. Our phantom powered device would see 48 - 0.275 = 47.725 volts. Doesn't seem like that should make much difference.

As for the QBox, I'd wonder if they are exceeding (maybe by a lot) the 10mA current draw to light the LEDs and do whatever else it is they do in that box.

Like I said several times the capacitance is what the mic sees and LOADS the mic down, add that up and perform the Ohm law and see what happens to the current at high frequencies...
I did the math: if a cable has 20pf per foot (Mogami 2552) and a length of 300' and a signal of 20000hz the load into a 1500 ohm mic pre will result in a load of 1.4ma. Compare this to a short cable of Mogami 2549 (3.4pf per foot) will have very little difference... Or .66ma vs. 1.4ma or 2.3 times the current...
This is referenced to a 1volt signal.
The load the mic sees now is a 704ohm @ 20khz.
This would be equivalent to a 6000pf cap at the end of the cable...
Check my math if you want...

[Testing123]

Originally Posted by nosebleedaudio

Like I said several times the capacitance is what the mic sees and LOADS the mic down, add that up and perform the Ohm law and see what happens to the current at high frequencies...

The capacitance has nothing to do with phantom power. It will affect the frequency response, yes. The phantom power issue is not related to that though.

[nosebleedaudio]

Originally Posted by Testing123

The capacitance has nothing to do with phantom power. It will affect the frequency response, yes. The phantom power issue is not related to that though.

You missed the point...
This is ALL related, if you load the mic MORE it draws more current, if you lower the phantom resistors it has more current available...
You can not change one and not effect the other.

[huub]

Kay, all very interesting stuff, but I have no idea what you are talking about
Anyway, when broadcasting golf, most mics stop working or become problematic (cable runs up to 1 kilometer or so.. sennheiser 416p or neumann km140), thus we feed phantom power locally..

[Testing123]

Originally Posted by nosebleedaudio

You missed the point...
This is ALL related, if you load the mic MORE it draws more current, if you lower the phantom resistors it has more current available...
You can not change one and not effect the other.

You're making a major assumption - that the amplifier inside the mic is servo controlled (i.e. has a feedback loop). A normal open ended amplifier will not behave this way, it will draw the same current regardless of load since it has no way to compensate. The schematics I've seen (e.g. Schoeps) do not appear to be servo amps, although I'm sure there are mics that are built that way.

That still wouldn't explain why the lights on a QBox cable tester would be dim - that has to be purely the DC resistance of the line.

[Testing123]

Originally Posted by huub

Kay, all very interesting stuff, but I have no idea what you are talking about
Anyway, when broadcasting golf, most mics stop working or become problematic (cable runs up to 1 kilometer or so.. sennheiser 416p or neumann km140), thus we feed phantom power locally..

OK, so you ran the mic lines for a kilometer and just fed the phantom power closer to the mic? That's one LONG cable run!

[huub]

Originally Posted by Testing123

OK, so you ran the mic lines for a kilometer and just fed the phantom power closer to the mic? That's one LONG cable run!

Yeah, but I might be confusing things here, because the problems we have with long cable runs might not be phantom related, but just sending mic level over long cable runs.. We both feed phantom and add gain directly where the mic is..

And the Qbox light thing I have seen on more than 1 occasion, but hardly under scientific circumstances, other factors could have been involved..
I, and many of my engineer collegues have always thought that you lose voltage over long runs, but I have now learned that this might not be the case...

So thanks for enlightening me

[David Rick]

Mike Keith,

Your math is correct, but you're calculating the wrong thing.

Assume, as you did, that the mic preamp has 1.5k differential input impedance. Now let's add a phantom power circuit and calculate how this changes. By symmetry, no differential current flows into the phantom supply, so it is as if we put 2*6.8k=13.6k in parallel with 1.5k.

1.5k // 13.6k = 1.351 kohm

Let's use that as our baseline, and recalculate with ballast resistors which are 10% and 20% lower.

1.5k // 12.24k = 1.336 kohm

1.5k // 10.88k = 1.318 kohm

This will have a negligible effect at low frequencies. If the microphone has a 200 ohm differential output impedance, then the gain will change by less than 0.03 dB.

What about high frequencies? Your example of 6 nF cable capacitance (300 feet) in parallel with 1.351 kohm yields a pole at 19.63 kHz. Repeating the calculation for 1.318 kohm, the pole moves out to 20.13 kHz. I think we can agree that this is a good thing.

In summary, cable loading matters a lot, while loading by the phantom power resistors is not very important.

Have a great weekend,

David L. Rick
Seventh String Recording

[nosebleedaudio]

Then David I say you missed my point as well....
The capacitance (long cable run) loads the mic (at high Frequency's) far more than the extra current (lowering the 6.8K) will compensate for...
By the way, do you think that who ever decided on the value of 6.8K did not realize it is a compromise?
It's not some magical value either.
Those resistors accomplish several things...

[David Rick]

Originally Posted by nosebleedaudio

Then David I say you missed my point as well....
The capacitance (long cable run) loads the mic (at high Frequency's) far more than the extra current (lowering the 6.8K) will compensate for...

Mike,

I'll concede that this is true in your example, though it obviously depends on the cable length. If the problem is long cable runs, then clearly a better solution is to move the preamps to the other end of the snake. That's why Millennia, Grace, Gordon, etc. offer models with remote gain control.

But cable capacitance isn't the primary reason that folks like me "hot rod" the phantom supply. Some mics just seem to sound better with more current. There are a number of us who like to run the Josephsons that way, and apparently there's some benefit with Sennheisers as well.

Anyway, the modification isn't that difficult, and Plush has good ears. He can just try it out with some different mics and decide if any seem to benefit.

Cheers,

David

[nosebleedaudio]

The original post was about raising the voltage to 52V.
Not lowering the resistors, big difference...
Like I mentioned already this totally depends on the mic and what the mic does with that 48V/limited current.
Some mics increasing the voltage and lowering the resistors will have ZERO effect on the output.
Just wanted to mention that...