phantom blocking capacitors

[pucho812]

what's a good value for phantom blocking caps.

is 100uF too much?

also why is it I have seen the caps wired backwards meaning coming from the xlr jack into the positive side of the electrolytic on some schematics. Is that as simple as a miss print?

 

[gemini86]

The positive end should be connected to the phantom it's blocking.

 

[undertone]

The positive end should be connected to the phantom it's blocking.

What happens if it isn't?

 

[pucho812]

The positive end should be connected to the phantom it's blocking.

really?

so it would be like this or does the negative go to ground?
(jack)
Pin 2_________+[100uF]-_________Mic pre input +
Pin 3_________+[100uF]-_________Mic pre Input -

 

[gemini86]

Well, then the cap would be reverse biased...and die rather quickly. 48v on one side, 0v on the other, how would you hook it up?

 

[gemini86]

Can you show me a schematic that shows cap - terminal to xlr pins?

 

[pucho812]

now you got me corfused ??? never bothered before as with a transformer ya don't need it, no transformers this time around

 

[gemini86]

http://electronicdesign.com/Content/14978/59941_fig_01.gif

here's a random pic I found...goooooogle.

...just note the xlr connector in the top right corner...

Also see the THAT 1510/1512 app note...

 

[pucho812]

http://electronicdesign.com/Content/14978/59941_fig_01.gif

here's a random pic I found...goooooogle.

...just note the xlr connector in the top right corner...

Also see the THAT 1510/1512 app note...

cool thanks.

 

[undertone]

Well, then the cap would be reverse biased...and die rather quickly. 48v on one side, 0v on the other, how would you hook it up?

No connecting + out is the right way to do it. But here's a question: if you have a self powered microphone, like an ECM, and you put a DC blocking cap at its output you put it's positive end towards the mic and it's negative towards the external connection (because the internal biasing voltage of the ECM is positive). When you plug this into a mic preamp that supplies a bias voltage (much lower than phantom, less than 5 volts) you are now presenting the negative side of that cap to a positive voltage. So the cap has a different positive voltage on each side of it, but since it's not connected across a DC voltage it's fine and performs normal transfer of AC signals.

But what happens when the internal biasing voltage fails, through battery failure or omission? In the absence of the voltage source, the inner side of the mic could find a path to ground and the cap could find itself connected in reverse, with positive on it's negative, and it's positive grounded internally. Could this destroy the capacitor, even at these low voltages? And could damage be partial?

 

[gemini86]

at such low voltages, cap failure would be a sloooow booooring death. Most caps can withstand small amount of reverse bias. I'm not an expert here...so a pinch of salt goes a long ways (beside, I'm sure if I'm wrong, I can think of a few people who would have straightened me out already). There was thread about this a while back http://www.groupdiy.com/index.php?topic=44945.msg562708#msg562708 ...

 

[JohnRoberts]

what's a good value for phantom blocking caps.

is 100uF too much?

also why is it I have seen the caps wired backwards meaning coming from the xlr jack into the positive side of the electrolytic on some schematics. Is that as simple as a miss print?

A lot of people favor 47uF, I've been pretty happy with 22uf (or none at all). !00uF is better in some regards but worse in others. Accidental shorts on the input could dump several amps into sensitive circuity, and if it doesn't blow up it will have a longer settling time. On the good side LF CM should be a little better with 100 uF.

JR

 

[Rossi]

Bigger caps like 100 uF or even more will push the highpass frequency well below 20 Hz. Which presumably lessens the degrading effect of audio passing through lytics. Lytics are evil. Lytics will kill your family in their sleep, so don't give'em a chance!

Another, perhaps more important, consideration is that the bigger the cap, the less you have to care about their capacitance imbalance. After all, electrolytics are usually +/-20%.

Also, these days lytics are often undersized, i.e. near their -20% limit. You may want to go one larger than what you think you need.

One disadvantage of - physically - large caps is that they are more prone to pick up hum from a nearby transformer and other crap.

 

[JohnRoberts]

While a rather obscure concern, the leakage current in electrolytic capacitors is linear with C, so scaling up the cap value X times will (if design equations are to be believed) scale up leakage X times, all other terms equal. 

In practice this is generally a non-issue. I only recall one production problem with noise in a mic preamp due to phantom blocking cap leakage, and I associate that with a a single model of capacitor that got a little to clever in their engineering to reduce size/cost (no I don't remember the brand). 

But as already cautioned look out for other unintended consequences (I.E. scale up protection circuits to handle the extra current from transient shorts- especially if using a patch bay where input shorts are more common).

JR

PS: I share Rossi's distaste for electrolytic caps in low impedance paths (like mic gain pots, mic inputs, and passive speaker crossovers.) For line level relatively high impedance paths the design practice of scaling up generally keeps the electrolytic mostly harmless.