48V phantom resistor, exactly 6k81?...

[ibvee]

I've come across some phantom power supplies which didn't put out exactly 48V at all. More something between 45V and 50V. So I was wondering why 6k81 is the standard resistor, instead of a 6k8, which works fine too, and is usually cheaper?

 

[dmp]

It is important that the two phantom resistors in the circuit match exactly to affect the balanced signal evenly. The absolute value isn't terribly important - I think it was defined as a standard for phantom power. The resistance limits the current the mic can take by creating a voltage drop. The resistor value is written 6.81k to probably indicate a high accuracy part in some schematics, but you can do just as well selecting a closely matched pair from 6.8k resistors.
From my understanding, any unbalance in the phantom resistors degrades the common mode noise performance. I do not think it affects the load impedance that the mic sees or any other characteristic. 

 

[ibvee]

That makes sense. Thank you.

 

[andre tchmil]

measure them on a JLM go between kit and you will be very very surprised  ;D

 

[Andy Peters]

I've come across some phantom power supplies which didn't put out exactly 48V at all. More something between 45V and 50V. So I was wondering why 6k81 is the standard resistor, instead of a 6k8, which works fine too, and is usually cheaper?

The P48 spec is 48V +/- 4V, so any stable voltage between 44V and 52V is acceptable.

The larger point is that the resistors need to match well for best circuit balance performance. And they need to be of adequate power rating.

-a

 

[PRR]

> why 6k81 is the standard resistor, instead of a 6k8, which works fine too, and is usually cheaper?

http://www.hobby-hour.com/electronics/resistor_color_code.php

"6K8" can be a 20% part. You could have 40% un-balance between two 6K8 resistors. This may-or-may-not cause troubles.

"6K81" implies a 2% or better part. Send your idiot assistant to the stockroom, insisting on the numbers 6-8-1, and you won't get a 10% part, you get a 2% part.

 

[Rossi]

P48 Phantom power resistors should actually be 0.2% to be matched within 0.4%. Not all manufacturers will be that accurate, but that's what the specification suggests.

As was said, actual voltage can differ +/- 4V from nominal 48V. Microphones are allowed to draw up to 10 mA, which more or less is the same as the maximum possible short circuit current of 14 mA.

I think this was discussed in the Drawing Board a while ago: It's a good idea to use 0.6 W resistors, as 1/4 or 1/8 W may drift from overheating or even burn when the input is shortened over a longer period with P48 engaged.

 

[PRR]

> P48 Phantom power resistors should actually be 0.2% to be matched within 0.4%.

Ah. (I never get that precise.)

Then specifying a 3-digit value (possible 2% tolerance) means:

1) With a couple dozen 2% parts you can be sure of hand-sorting many 0.4% matched-pairs.

As a worst-case, "6K8" resistors non-randomly spaced 1.004 apart:
6800 6827 6855 6882 6855 6882 6909 6937

Real resistor production will never give a distribution like this.

So fifteen "6K8" within a 2% spread is sure to have some 0.4% pairs. If distibution is more random you will have <0.4% pairs. Buying fifteen 2% resistors to get for-sure four 0.4% pairs is "cheap" compared to four mikes, four cables, four jacks, four amplifiers....

2) 2%-ers probably drift less than old-old carbon-composition, so whatever it is today it will still be very-near tomorrow and next year

 

[Rossi]

I usually reach into a 100 pieces bag of 1% resistors an measure about 10 until I find two that are identical within my DMM's accuracy. I do that both with the 6k8 resistors in a preamp and the phantom sourcing resistors (often around 2K) within a mic I'm building.

I know that 1% unmatched will do in most cases (and many manufacturers be content with that), but I like to know that my stuff is a little better than absolutely necessary. And why not, if all it takes me is a minute or two of measuring?

 

[justanalogue]

Same procedure here, simple, cheap and you're sure you've matched pairs!

Willem.

I usually reach into a 100 pieces bag of 1% resistors an measure about 10 until I find two that are identical within my DMM's accuracy. I do that both with the 6k8 resistors in a preamp and the phantom sourcing resistors (often around 2K) within a mic I'm building.

I know that 1% unmatched will do in most cases (and many manufacturers be content with that), but I like to know that my stuff is a little better than absolutely necessary. And why not, if all it takes me is a minute or two of measuring?

 

[sisepuede]

I usually reach into a 100 pieces bag of 1% resistors an measure about 10 until I find two that are identical within my DMM's accuracy. I do that both with the 6k8 resistors in a preamp and the phantom sourcing resistors (often around 2K) within a mic I'm building.

I know that 1% unmatched will do in most cases (and many manufacturers be content with that), but I like to know that my stuff is a little better than absolutely necessary. And why not, if all it takes me is a minute or two of measuring?

Hello Rossi,

What happens if those are unmatched or are not exactly the same?

I am asking because I am experiencing some Phantom power noise and I can't find where the problem is, it sounds like boat's motor every tiem I turn the P.Power on. Could be that?

When I did my preamp I matched those but my multimeter is not "that" precise, I only could read 6.80K in both cases, those were metal film 1% vishay resistors. My preamp is a NV73 a Neve 1073 clone that sounds amazing, but I can't use condenser mics there because of that.

Thank you!

 

[[silent:arts]]

Your NV73 needs either the input or output pin1 connected to GND, not to chassis like we & Rane always recommend.

 

[sisepuede]

Your NV73 needs either the input or output pin1 connected to GND, not to chassis like we & Rane always recommend.

Thanks mate, yeah I did that, David ( NV73 Maker) told me that because if not the NV doesn't work at all, so I did that (audio GND, Chassis and PSU GND all together) and it worked fine with dynamic microphones, but at that time I heard some noise in the background, kind of 50Hz, then I connected a condenser and when I turned the 48v on that was a massive noise like a "boat's motor" (Thanks Cemal for that expression )... since then I am trying to figure out what the hell is causing that fu**ing noise.

I've uploaded some pics in my gallery, you could check those out to see how I did it, I have a frame where the NV73 is in test mode, with a card edge connector and it really works fine (no funy noises, GND loops or static noises) with the DIY51X PSU, but the phantom power, that's the only bad thing.

By the way tomorrow or on Friday my 511 Rack will arrive from Scotland hopefuly I'll have this sh*t sorted out.

cheers.

 

[JohnRoberts]

The matching of phantom power resistors can degrade common mode rejection so main symptom of poor match will be extra hum pickup.

6.8k value is based on nominal 48V phantom, but many industrial applications use lower voltage and lower value resistors, FYI.

JR

 

[sisepuede]

Your NV73 needs either the input or output pin1 connected to GND, not to chassis like we & Rane always recommend.

Thanks mate, yeah I did that, David ( NV73 Maker) told me that because if not the NV doesn't work at all, so I did that (audio GND, Chassis and PSU GND all together) and it worked fine with dynamic microphones, but at that time I heard some noise in the background, kind of 50Hz, then I connected a condenser and when I turned the 48v on that was a massive noise like a "boat's motor" (Thanks Cemal for that expression )... since then I am trying to figure out what the hell is causing that fu**ing noise.

I've uploaded some pics in my gallery, you could check those out to see how I did it, I have a frame where the NV73 is in test mode, with a card edge connector and it really works fine (no funy noises, GND loops or static noises) with the DIY51X PSU, but the phantom power, that's the only bad thing.

By the way tomorrow or on Friday my 511 Rack will arrive from Scotland hopefuly I'll have this sh*t sorted out.

cheers.

All sorted out, it was the TL783C, once I replaced it the problem disappeared.
Cheers.

 

[sisepuede]

The matching of phantom power resistors can degrade common mode rejection so main symptom of poor match will be extra hum pickup.

6.8k value is based on nominal 48V phantom, but many industrial applications use lower voltage and lower value resistors, FYI.

JR

Thanks a lot John, I didn't know that.
cheers.