Phantom power and transformer - how does that work?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

hg_man

Well-known member
Joined
Jan 21, 2005
Messages
56
Location
Indianola WA
Looking at some mic schematics, I've found that pin 2 and pin 3 of the XLR are tied to the outer sides of the transformer, and the phantom power supplied to the mic then comes off a center tap.

How does this work? I thought that a transformer couldn't be used with DC, yet here it is. Or is the amount of DC current so small that it doesn't saturate the transformer? I calculate 7 mA flowing through each leg of the transformer to the center tap (48V/6810 Ohms)
 
calculate 7 mA flowing through each leg of the transformer to the center tap

That's the crucial thing: the DC is balanced out, the magnetizing current is zero.

Another thing: you'd only be drawing 7mA through each leg if you were feeding a short!
 
another point to add to NYD's post is that the 6k81 resistors *should* be highly matched to ensure that there is no current flow.

Think of it this way.. if you apply 12VDC to BOTH pins on a light bulb, what happens? nothing at all. Same for the trafo, if matching DC is applied to the trafo then the current does not flow through it at all.
 
I realize that the 7 mA would only flow in each leg if it's feeding a short - so of course the actual draw of the amplification stage would be much less (we hope!).

However - let's imagine that the amp draws 4 mA. Thanks to Kirchhoff's Current Law, and assuming that the resistors are balanced, each side of the transformer would be drawing 2 mA DC through it. Right?

A
 
It's helpful to draw out a schematic of this, complete with the mic and it's internal circuitry. Then you can analyze different situations, shorted resistor, open cable pin 2, grounded centertap, etc.

It's kind of hard to understand with half the circuit miss ..... ing.

Start out with something lame like this, then work your way up to an actual circuit diagram with numbers and stuff, ...



phantom1.jpg
 
[quote author="Svart"]the 6k81 resistors *should* be highly matched to ensure that there is no current flow.[/quote]

How well should they be matched? The Jensen design i came across said 0.1% min, up to how much more is still noticably better?

And how about noise versus matching? I mean one could "build" two 6k81 resistors out of different others values get closer matching, right? Would the better matching be kind of eaten up by the acummulated noise of the parts?

i hope these questions aren't too theoretic...
 
Same question, can you "build" them?

6,800 ohm resistor + 10 ohm in series = 6,810 ohm

or if your resistors test as 6,780....add a 30 ohm = 6.810

??
(edit)...checked Mouser...prt#66-RC55-D-6.81K...95 cents each (tol .1%)..
...whooptee-do, go buy a couple.

=FB=
 
> the 6k81 resistors *should* be highly matched

Yeah, but the question is about the other way: one 3.4K resistor feeding the center-tap of the transformer.

There is DC in the transformer: two current paths flowing opposite ways in the winding. If the two currents are exactly equal (but opposite), then the magnetic field from their sum cancels and the core is happy and free of DC effects.

Are they equal? In this case, there is little we can do about it. The 3K4 resistor is common to both paths, it can be 5%, it has no effect on balance. The transformer DC resistances should be matched, but they usually are, and we don't have much control of that. (Beware of using this connection on older low-price pre-Phantom windings: I've seen dual-winding PA mixer trannies with significant unbalance.) Anyway the main current determining factor is the current limiter(s) inside the Phantom-power mike. Good mikes will have fairly good balance. Ya never know what's in a cheap mike.
 
There is no need to buy precision resistors for Phantom power feeding - with a normal digital multi meter just "pair match" from a selection of ordinary ones.

If you normally use 6K8 1% resistors (and these are inexpensive - so why not?) - just measure the actual values and sort them into pairs of identical value. This is quick and easy to do - some would say it is "gilding the lily" - but it is the sort of thing that DIY 'ers can do.

Commercially manufactured equipment is normally fitted with standard 1% resistors - and they clearly find that this is accurate enough for real world CMMR measurements, or balance.

John Godsland
 
And it does not have to be 6.81K. It can be 5.02K, 5.11k, whatever you have in a precision that is close.

Or just pick a couple of carbon comps out of the junk bin that are close, like a 6.2K or something.



Note that if going thru a transformer CT with phantom, that you get some choke like filtering on the dc.
 
Note that if going thru a transformer CT with phantom, that you get some choke like filtering on the dc.

I don't think that's right. You only get the benefit of leakage inductance, which is effectively really small (thanks to good magnetic coupling). The large differential inductance (seen by the mic) is not seen in common mode.

Note, even with 6.8k matched resistors, there can be a dc current imbalance in the primaries. Most mic trannys have <100 ohm dc on the primary winding, right? That's effectively a short. So the 6.8k resistors end up essentially in parallel, meaning they each have the same current value flowing through them. An unbalanced mic will lead to an unbalanced primary. Only way to fix this is to use a good mic or capacitively couple.

jh
 

Latest posts

Back
Top