Max input level of a PC soundcard?

Hello Folks,

What's the max for soundcards? I have read specs of some where it says
22dbu which sounds wrong to me. A Via AC'97 I have seems to max out at
10mV RMS but no clue if that is a standard. There was no datasheet on
the Via Technologies site. Also, anyone know the "frying level" where it
would go kaputt?

The motivation here is that I want to build a "noise sniffer" preamp.
20-40dB or so, plain old transistors. Here's hoping I'll get by with a
single AAA alkaline as the supply because that would allow a pen type
enclosure.

 

There seem to be absolutely no standards whatever for this.

Apparently Creative won't even publish (or tell you on request) the numbers for
the Soundblasters.

Graham

 

22 dBu sounds wrong to me also. Although, that "unloaded" thingy has never
been my favorite.

http://www.waypoint.com/users/~discobay/Sound_Card_Presentation.htm

 

There is no standard for it.

It all depends on the analog VGAs of the particular audio card and the
mixer setting. The mix. input can typically take the levels up to 100mV,
and the line input is up to several volts.

There was no datasheet on

I did that! The signal of 10Vrms kills the mic. input.

Why transistors? One good opamp would do.

Many audio cards do have the electric power at the mic. input, so you
can feed the preamp from it.


Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com

 

He talks about 1.5Vpp. That really doesn't sound right to me. AFAIK this
whole area on the laptop is supplied with only 3.3V.

BTW I was pleasantly surprised by the dynamic range I got. And not so
pleasantly surprised that the only way to remove hum (other than
notching) is an audio transformer. Or if you have to measure below 20Hz
a powerline transformer that has two individual secondaries.

 

I thought so :-(

Specsmanship also does not seem to be a strength in that field. I guess
none of those guys has ever worked on a defense or medical design.

Ouch! That's a lot of volts.

It's easy. Opamps aren't that good on a single alkaline that has dropped
to 1.3V. No problem with transistors. Of course this only makes sense if
I know that I never need more than 50mVpp or so.

Mine offers 2.6V. Not much to write home about and it begins to collapse
the millisecond you load it with only a few hundred microamps. Also, the
2nd channel does not provide any phantom power at all.

 

I hear "dBu" and immediately think: "dB above 1 microvolt per meter".
But I don't think audio and radiofrequency field intensity mix very
well.

-mpm

 

dBu is a voltage measurement specific to audio.

Graham

 

Perhaps, but it is also used in radiofrequency propagation
measurements.
Feel free to check out 73.215 of the FCC Rules, for a classic example.
(In this case, the grandfathering of short-spaced FM stations.)

http://a257.g.akamaitech.net/7/257/...s.gpo.gov/cfr_2006/octqtr/pdf/47cfr73.215.pdf

-mpm

 

I'd say it should be good for "line in" levels at the "line in" input. ;-)

I've heard somewhere that this can go to .7 or 1VRMS; I'm sure you or
someone could look up the standard for "audio line level" or some such.
If you're building a 20-40dB preamp, then you could just clamp its
output at 1.4VPk or so if you're paranoid or anything. ;-)

Good Luck!
Rich

 

Hmm, that would mean I need two AAA cells. Dang. But for some reason my
card starts to clip at 10mVrms. Thing is, I will use that on other
people's (client's) PCs as well so I want to play it safe.

No idea why mine clips so early. Maybe they have some kind of PGA on
there. Probably buried in a driver and there is next to nothing in
documentation. In the good old days PCs were extensively documented.
Nowadays you get a packing list and numerous lawyer pages, that's it.

 

No it isn't.

Incorrect usage. That's dB 'mu' btw not dBu. Furthermore it should be suffixed with a unit like V or W.

Graham

 

Which standard ? It's rather hard to have a standard level for a dynamic signal
too !

Graham

 

Mine tops out at about 80mV rms. (1dB compression point)

 

No, it can't be that hard to agree on a maximum signal level for
example. Sometimes such standardization works well, such as in the
medical DICOM format. Other times it's mostly hot air, such as in EDIF.

 

Interesting. That is quite a few dB difference there. Maybe a PGA is
involved but I guess in the absence of a standard or at least some
documentation I won't know :-(

 

Bullshit. It is audio industry standard (DIN among others).

Most likely STANDARD "line level" inputs.

They certainly do. Apparently you are not able to read or interpret
specs for audio gear.

Line-level... mentioned all over their specs for nearly ALL of the
analog input declarations. Since we are not discussing digital, we need
not go there.

From Wiki for the term "line-level":

"Professional audio equipment line levels are rated in dBu, and the most
commonly used nominal level for such equipment is +4.000 dBu, which
corresponds to a sine wave of about 1.228 VRMS. The absolute reference
voltage is 0.7750 VRMS. This reference voltage corresponds to 1.000 mW of
power at 0.6000 kOhms."

 

No shit!?

 

So how did you know this, yet not see how perfectly clearly Creative
declares their input specs?

 

Way I look at it is that the bottom noise end is bound to be a uV or so
(20kHz bw ) and the makers are always keen to point out their 100dB DR,
hence max' in has got to be in the 100mV area but yes, if there's a volume
slider lurking somewhere in the mutiple layers of soundcard baggage, then
all bets are off.