Phantom Power and Y-Splits

[smalljoshua]

2 mics into 4 channels Passively splitting each mic into 2 channels, both of those channels with phantom applied.

 

Josh

[Dan Appleby]

Wow, didn't expect such an in-depth response.

 

This is a simple concept! 1 mic (phantom powered) y-split into two mic inputs on a desk, both have phantom power applied (I.e no selectable phantom power switch on each channel).

 

Anyway... all you need to know is that I did it and it worked fine. All is well in the world.

 

For those who are interested, the reason I did this is so I could use a different gain structure on the two separate channels, one for the PA and one for a VT record feed. I couldn't seem to get enough level from my pre-fade aux for the videots to be content with!

 

I often y-split channels if I'm running monitors from FOH and want un-compressed/different EQ for vocals etc in monitors. It's good practise if you have spare channels on your desk!

[paulears]

Well - people have given some really exact predictions about the voltage increase from plugging a Y-Split into two desk channels.

 

Having just been clearing up, I found two XLR males and a female that had been hacked off an old loom - so I've soldered up a Y-Split. Here is what I found.

 

Y-split plugged into a Soundcraft desk. No Mic attached.

 

Between Pins 1 and 2 = 48.9V

Between Pind 1 and 3 = 48.9V

 

Audio Technica AT-815 Mic attached. Fader raised, mic working.

Between Pins 1 and 2 = 48.9V

Between Pind 1 and 3 = 48.9V

 

Dangling XLR plugged into channel next door

Between Pins 1 and 2 = 48.9V

Between Pind 1 and 3 = 48.9V

 

Anyone care to comment on why my practical tests of a functioning split appear to make all the physics make no sense whatsoever?

[Mixermend]

Although this AT mic is a condenser type - it is clearly consuming no current from the Phantom supply.......

 

Was it actually working?

 

It can be powered with a battery, although normally Phantom would 'take over' from it.

 

There must be an answer!

[bruce]

Well, first of all, the spec for that mic http://www.audio-technica.com/cms/wired_mi...7f6b/index.html suggests that the typical phantom current is of the order of 2mA - which means the assumptions made earlier about resistance are an order of magnitude out. But it still suggests that there should be a voltage swing of around 3V.

 

I wonder if the limiting resistors in the mixer are common to all mic channels - so there's not one per channel, just a global limit?

[paulears]

As I said, raising the fader showed the mic working, and there is no battery fitted. I have an Oktava condenser in the studio next door, I know this is rather more current hungry, being a large diaphragm condenser, not an electret. I shall pop off and do the measurement on that one.

 

AFTER MEASUREMENTS

 

There is a difference - no load voltage from Pin 1 to 2, and Pin 1 to 3 still =48.9V

 

However - with The Oktava (An MK319 for reference)

 

One mixer channel connected. Pin 1 to 2 and Pin 1 to 3 =45.5V

 

Connecting the two did increase the voltage - or perhaps more correctly, reduced the no load voltage to 47V - again, the same on Pin 1 to 2 and Pin 1 to 3.

 

This is rather different from the 17V and 26V mentioned earlier. If anyone else has a splitter and fancies trying this out, it's quite interesting. Drawing out the circuit - I get this. I've tried to account for some protection found in some circuits for when loads are applied without turning off the power. I've included an indication of some feeder resistance, but that's pretty low. All the phantom PSU circuits seem to have similar designs, but small variations in component values.

 

Looking at the circuit, and using what maths I remember - I can't make my measured results work. So, what's going on?

http://www.eastanglianradio.com/phantom.jpg

[Rob_Beech]

It would make sense as Bruce says for there to be global limiting resistors on global phantom. Sadly my knowledge is limited in this instance to thinking logically rather than electronically.

 

I have the necessary leads. I shall perform the test myself. On both a desk with global phantom, and one with individually switched phantom, switching on both mics. I shall try a Sennheiser lecturn mic, and a U87.

[Jivemaster]

As my battery/phantom mics had a battery life approaching the shelf life of the AA cell the actual current must ba so small as to me almost negligible hence the lack of volts drop over the phantom power system.

[Dave]

OK, just did a quick test myself, using EMO E725 two channel phantom power supply and Shure MX418 lectern mic.

 

No load: 48.0V both channels both pins

Mic connected to one channel: 40.4V both pins

Mic connected to both channels: 44.1V both pins

 

The PSU has 6.8K resistors on each pin (repeated for each channel)

 

If you do the math for both cases you'll find the mic draws about 1.1mA on each pin, consistent with the data sheet value of 2mA (which will be for both pins).

 

This link has some useful info on phantom supplies. It says the standard allows for a +/- 4V tolerance on a 48V supply so I can't see the increased voltage presenting a problem. There is an issue in that the source impedance of the power is halved but I don't think that will matter.

[David Duffy]

Keep in mind that the voltage at the mic can't rise to more than 48V (that the desk provides), no matter how many channels you common up.

[Rob_Beech]

Interesting result from a dedicated +48v supply. Shows that things are different.

 

Test done with Z leads to allow easy measurements

 

So,

 

No load 47.9v on the MX12/4 48.6 on the GL2800

Sennheiser lectern mic connected to 1 channel, 46.2 and 48.5

Connected to 2 channels 46.2 and 48.6

 

U87 1 channel 44.5 and 48.5

and 2 channels 44.1 and 48.5

 

 

As a silly test but one that I found interesting....

 

I connected the lectern mic through a 10mXLR, into a stage box that's 15m into a multicore that's 100m and into the desk. I lost about 1V, yet if I put a Y split (Z actually for measurements sake) on the tails at the desk end and another for measurements sake on the end of the XLR that plugged into the mic, I actually measured over 48v with both desks.

 

So it appears to make a difference over significant cable length but less of one on short cables.

 

 

The multicores were coiled in their cases.

[Dave]

Rob,

 

Something funny is going on with your gear :P Not sure what, though. I'll do some tests with my own mixers later.

 

However, I don't think you should see any noticeable drop via a long cable. 0.22mm copper (typical size of most mic cables and multis) has a resistance of about 0.08 ohms per meter, which is 10 ohms for your 125m setup (ignoring connector resistance). 1mA through 10 ohms gives a voltage drop of 0.01volts, which I think anyone would struggle to measure outside of a laboratory!

 

EDIT: I meant to add that I tried the EMO phantom psu simply because I had it to hand, not because I expected it to behave differently to a mixer.