Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[richbreen]

Hi there - I have a pair of MKH80 microphones which I use frequently in many different recording studios and which in general work perfectly.  I have however run into a couple of places where they *don't* work, and I'm slightly mystified as to what's going on - I'm hoping someone might have seen this before.  In two different studios now, these mics sound very low level and band limited (no high end, no low end); it seems to be a phantom-power related issue, as inserting a stand-alone phantom power box in line restores the mics to normal operation plugged into the same preamps.  However, all other mics (many phantom powered) operate just fine, and measuring the voltage at the mic's xlr with the mic connected shows the phantom voltage at the correct 48VDC - there has not been time in any of these situations to troubleshoot beyond that.  Both studios where I've run into this are older consoles (one an old Neve 80xx series, the other a vintage API).  Has anyone else run into this?   Thanks in advance for any thoughts.

best,
rich

[David Satz]

Rich, the phantom powering for an MKH 80 requires 3 mA, which isn't very much in modern terms. But many "vintage" mixing desks (before, say, the mid-1980s) didn't/don't have a proper implementation of the phantom powering circuit according to the present-day standard. One of the main defects of older consoles is that the two 6.8 kOhm resistors on each mike input are often fed not directly from the +48 Volt supply, but through a series (current limiting) resistor that adds a further 2 or 3 kOhm to the circuit.

Thus when no microphone is connected, the correct 48 Volts will be measured but as soon as a microphone is connected, the voltage drops; some amount of drop is normal and expected by the microphone's circuit designer, of course, but the added series resistance causes the voltage to drop several Volts farther than the designer may have reckoned with.

If you can loosen the shell of an XLR connector on a mike cable enough that you can measure the voltage between (say) pin 1 and pin 2 with an MKH 80 connected, according to Ohm's Law you should see about 38 Volts (since each leg draws ca. 1.5 mA, and 1.5 x 6.8 = 10.2V expected drop; 48 - 10.2 = about 38). If you measure less than about 32 Volts, the microphone simply isn't being powered correctly, and frankly that wouldn't surprise me at all with many older consoles.

--best regards

[klaus]

David, could you elaborate further on this oddity:

...One of the main defects of older consoles is that the two 6.8 kOhm resistors on each mike input are often fed not directly from the +48 Volt supply, but through a series (current limiting) resistor that adds a further 2 or 3 kOhm to the circuit.

Can this shortcoming be remedied on these older consoles, and if so, how?

[soapfoot]

I've seen a few designs for phantom circuits that have a series resistance that feeds the 6.81k resistors. I've never been able to get a straight answer as to what it's for, or why I would want it.

I've built several preamps with phantom supplies, and I always just feed the +48 directly to the junction of two 6.81k resistors. I've never had any issue and it always seems to work perfectly.

[richbreen]

...
If you can loosen the shell of an XLR connector on a mike cable enough that you can measure the voltage between (say) pin 1 and pin 2 with an MKH 80 connected, according to Ohm's Law you should see about 38 Volts (since each leg draws ca. 1.5 mA, and 1.5 x 6.8 = 10.2V expected drop; 48 - 10.2 = about 38). If you measure less than about 32 Volts, the microphone simply isn't being powered correctly, and frankly that wouldn't surprise me at all with many older consoles.

--best regards

Hi David - thanks very much for the informative reply; the same thought occurred to me, and as I said, I did measure at the XLR pins with the mics connected and it was a solid 48VDC, which is what makes this so mysterious.  I'm at a loss right now to explain it, though the possibility always exists that I simply made a mistake in this measurement.

Thanks again,
rich

[gsherman]

Are you sure that you have the hot pin (I think its pin 2) connected to the +ve voltage?

[David Satz]

rich, if you had 48 Volts at the pins, then the microphone couldn't have been receiving power. When a microphone draws 1.5 mA through each modulation lead (for a total of 3 mA), then that 1.5 mA is flowing through (at least) the 6.8 kOhm resistor leading to that pin, which creates a voltage drop of just over 10 Volts.

As I said, the microphone designers are well aware of the inherent voltage drop, so when a microphone is specified for 48 Volts, in reality that means "connect this mike to an input where the _open-circuit_ voltage is 48 Volts." Well, that plus the other things that are defined in the standard--but the point is that microphones don't require, and in some cases may not even be able to tolerate, 48 actual Volts while operating.

But they do have minimum requirements, and any substantial extra resistance between the +48 DC supply and the pair of 6.8 kOhm resistors will further reduce the voltage that reaches the mike. The standard specifies a tolerance of 6 Volts either way, and microphone designers should take that into account as well, which is why I said that anything below about 32 Volts would be incorrect.

--best regards

P.S. (October, 2013): I looked again at the printed DIN standard last night and realized that in the above message, I had overstated the voltage tolerance. It is actually ±4 Volts, not ±6. So I should have said in this message, "anything below about 34 Volts would be incorrect." My apologies.

Small differences can be quite important especially in microphones that use internal DC converters, which not only provide the capsule polarization voltage but may also power the rest of the circuitry including the output stage. Depending on the converter in a given microphone, even a small reduction in powering below the expected level can cause large reductions in both the overall sensitivity and the clipping point (max. SPL) of the microphone. --D.S.

[Uwe]

... and to test a phantom supply if it complies with the standard, simply connect two resistors of 6.8 kΩ one between pins 1 and 2, the other between pins 1 and 3 of an 3-pin XLR male plug insert. When this is plugged into a 48 V phantom powered microphone input you should be are able to measure 24 Vdc (±10%) across each of these resistors. I have long ago added this simple plug to my essential tool kit ...

[Kai]

connect two resistors of 6.8 kΩ one between pins 1 and 2, the other between pins 1 and 3 ... measure 24 Vdc (±10%) across each of these resistors...

The basic idea is correct, but some devices, like standalone preamps might not be capable to supply sufficient current (7mA) into this.
Few mic's need 7mA, so e.g. a Neumann KM84 might perfectly work on a preamp that fails this test.

When testing P48 one more thing is important: The phantom power should be clean, straight DC.
The test is simple: connect a single resitor ca. 6K8 to only one side of the balanced input (e.g. from pin 1 to pin 2).
If you listen to the preamp output you can hear the hum and noise that might be in the phantom power. There shouldn't be much compared to open circuit.

(Explanation: by loading only one side of the balanced input hum and noise in the phantom power does no longer cancel out and becomes audible.)

Regards
Kai

[klaus]

(...) switch a resitor 6K8 to only one side of the balanced input (e.g. from pin 1 to pin 2).

This statement is not clear to me. Do you mean: connect only one resistor instead of the usually two, as Uwe suggested for his test? Please rephrase.

[Jim Williams]

The problems mentioned here are a result of current drop, not necessarily a voltage drop. That is why you will still see 48 volts if you measure pin's 2 and 3 and your U-87 or Schoeps still won't power up properly.

Yes, some preamp designs do less than 48 volts, some Symetrix preamps do 43 volts through reduced 4.99k resistors.

As David mentioned, it's usually a series resistor added before the 6.81k phantom distribution resistors. I find 1 or 2k ohms commonly added before with a 47 uf cap added to filter that voltage to ground.

As a general rule here, I change that to a 100 ohm series resistor followed by the 47 uf cap and then the 6.81k resistors. If the supply is an honest 48 volts, all mics will power up properly. I usually set my regulator voltages to 50 volts to allow for some droop and to give the mics an honest 48~49 volts to keep all of them happy.

Every phantom powering scheme should be examined to determine if that series added resistor is too high in value. Eliminating it also isn't a good idea as it creates a R/C network that helps filter out noise from the 48 volt supply. Adding a small .1 uf cap can help reduce high frequency power line noise as well. Like your car, audio gear is also dependent on the quality of your fuel supply.

[Uwe]

I fully endorse Jim's observation to check every purported phantom power implementation. This is exactly the reason for my previously suggested test adapter. I further endorse his additional test for any undesirable noise from the phantom supply.

The apparent problems with 48 V phantom power may stem from the changes in the standards governing it. These changes were made necessary by the evolution of microphone circuits over time. Early phantom powered microphones did require modest operating currents of between 1 and 2 mA, while many modern versions may need up to 8 mA to meet their full technical specifications.

The early DIN 45496 standard did require a minimum current capability of 2 mA, but the current IEC 61638 standard calls for a minimum current capability of 10 mA. Assuming short circuit conditions at the microphone connection, to meet the present IEC-standard requirements, the total internal DC-resistance for a phantom supply can not be less than 4.8 kΩ, which leaves a maximum series resistance of 1.4 kΩ (typically <1 kΩ to allow for the permissible tolerance of ±10% of the 48 Vdc source) for the filter section between the 48 Vdc source and the 3.4 kΩ due to the two mandatory parallel 6.8 kΩ series resistors.

My earlier post simplified the phantom source without a possible additional filter section when it gave the resulting measurement as 24 V ±10% (21.6 V to 26.4 V) . To be completely correct, the acceptable voltages measured may be between 26.5 V (phantom source at upper tolerance range of 53 V with near 0 Ω source resistance) and 19 V (phantom source at lower tolerance range of 43 V with around 1 kΩ source resistance).