How do I measure - 140 dB noise levels?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
G

gentlevoice

Member
Joined 2007
Paid Member
Hello,

I'm looking for a cheap and straightforward way of measuring noise levels down to about -140 dBs (ref. 2 VDC). Any of you who knows how this can be done? A soundcard would be prefered but I haven't seen any with this kind of specs at a reasonable price .....

Regards,

Jesper
 
ostripper

ostripper

Member
Joined 2008
I think Nelson Pass "Passlabs" is/was working on a ultra low
noise precision amp to allow for testing sub 110DB noise.
Do a search in "passlabs" you might find it.

My soundcard (M-audio/envy 24) is -122 at 20khz -126 at 1khz
I use FFT and can see this, but it is good enough.(can see
junction noise in transistors... yikes)

BTW .. Why??? -110-120 noise is spectacular ,if it is below that
why the fuss?:D
OS
 
G

gentlevoice

Member
Joined 2007
Paid Member
Hi again,

- and thanks for your replies - they made me think further into what my needs are.

I would like to measure very, very low noise levels, e.g. the noise of batteries, however, would need to do so also at very low frequencies, i.e. down to <1 Hz. At the same time I will be measuring noise (at a high resolution) up to ~ 100 kHz.

I guess this could be done using a very low noise amplifier feeding into a quality (192 kHz/24 bit) soundcard, possibly with the exception of the ability to measure down to 1 Hz (I haven't seen any capable of measuring at these frequencies..?). The soundcard needs to be able to connect to my PC via USB or firewire. Also it should be able to record and playback at the same time as I intend to use it with the Praxis software (http://www.libinst.com/).

@ Ostripper: BTW I did a search on Nelson Pass' ultra low noise amp but did not find anything ... nor on M-audio/Envy ... ??

Suggestions are appreciated ... :angel:

Best regards,

Jesper
 
K

kristleifur

Member
Joined 2006
gentlevoice said:
Thanks - it would be an interesting design for me, however, I don't have a source for the AD797 - which by the way seems to be very low noise...

Do you know where to source it?
Click to expand...

Sorry, don't know, but you may want to check with Tangent himself, and at the Headwize.com DIY headphone amp forums. I think they were discussing these amps there a while ago.
 
lineup

lineup

Member
Joined 2005
ostripper said:

BTW .. Why??? -110-120 noise is spectacular ,if it is below that
why the fuss?:D
OS
Click to expand...

Agree.
Why bother with noise levels below -100dB

A good phono RIAA has got -70 to -80 dB noise.
And this is well below anything effecting music from a vinyl record.
Blowtorch topic is the right place for all thinking inaudible noise will kill them.

Modern CD-players and opamps and most power amplifiers are so low on noise
it has got no longer pratical importance,
after passing the weakest link = Loudspeaker/Headphones.

And in my opinion, low noise is a very overrated parameter.
Because of those nerds that day out and day in discuss it like it was a deadly enemy.

Noise is natural.
We are constantly surrounded by a background noise.
Even in the middle of the quietest forrest.
We are so used to this, that only if all noise was removed, we would notice.

And be serious scared!
Make a visit into a room without noise.
Those who have, will tell you it is a horrible experience ;)
 
G

gentlevoice

Member
Joined 2007
Paid Member
@cuibono - thanks for the suggestions. Will check it out ;)

@lineup. Hmmm.... I actually agree with you. Noise is not really a crucial parameter to me. My general approach is to use measurements, e.g. noise, as a guideline, however, I am a bit intrigued by the prospect of looking at e.g. battery noise under dynamic conditions. And it is my impression that batteries have very low noise floors meaning that the measurement equipment must be similarly quiet.

Anybody knows a reasonably priced source for e.g. AD 8599 (dualchannel opamp with noise specs close to AD797)?

Enjoy your day DIYs :)

Jesper
 
syn08

syn08

Banned
Joined 2005
synaesthesia.ca
C

Conrad Hoffman

Member
Joined 2007
conradhoffman.com
Actually, batteries aren't as quiet as sometimes thought. It depends on the type, the state of discharge and the current draw. IMO, the two approaches are to ac couple an amplifier or preamp, the look at the output with a scope, or to set up an even quieter reference voltage and do a differential measurement. That gets you down to DC, but building the reference is a fairly difficult task. If you ac couple the battery with a nice low leakage film cap, and use any decent low noise opamp with high gain, it's a trivial problem. An LF411, NE/SE5534, or dozens of other opamps will probably get you there. I'd use the National LME series opamp, as it's become an almost universal part for me. It doesn't seem to suffer from popcorn noise, which is important, because batteries do. Though it has high bandwidth, you'll want to filter out HF noise to see the LF stuff clearly. That will be true regardless of what you use. BTW, if you look at primary cells, be sure to compare a standard carbon zinc battery with an alkaline. The difference can be quite remarkable.
 
G

gentlevoice

Member
Joined 2007
Paid Member
@syn08: Thanks for the tip about analog devices samples. I just ordered two AD8599s. That is indeed a very fine opportunity & service for people needing just a few items. I look forward to measuring batteries etc.

BTW I have two PCM4202 from TI that I would like to send data to my laptop via USB (at the high 24 bit 192 kHz resolution).

Anybody has suggestions on how this is done?

Regards & thanks to all of you for ideas!

Jesper
 
peranders

peranders

Member
Joined 2002
sjostromaudio.com
Paid Member
lineup said:
Why bother with noise levels below -100dB

A good phono RIAA has got -70 to -80 dB noise.
Click to expand...
Mr Lineup, such statements will only lead to that you'll be ridiculed at some other forum. Jesper did ask for how to... not you don't need to.

Jesper you'll need a preamp which has lower noise than the test object but technically I'll doubt that the test object has more than 140 dB dynamics. Can you explain what you are up to, then it's easier to give advices.
 
S

starstruck

Member
Joined 2008
richie00boy said:
-140dB is the realm of high quality test equipment, i.e. not reasonable price.
Click to expand...


Yip , you will need some pretty expensive test gear to do this !

Whilst opamps etc may quote noise figures in that region, by the time you build them into a measurement system you have introduced loads of noise.

Same with PC sound cards, the PC generates so much noise that you would never get an accurate measurement.

Also if you build you're own / use a sound card how are you going to prove you're measurement is accurate. Do you have a calibrated noise source as a reference.

I don't want to sound negative here but the likes of Agilent, LeCroy, Tektronix etc etc charge a fortune for kit that can measure noise in that region. If it was easy to build your own circuit to do it then they would go out of business.

Good luck mate.
Kev
 
PMA

PMA

Member
Joined 2002
pmacura.cz/audiopage.html
Paid Member
I use +60dB amp before soundcard.
Hereby -120dB from SACD, RBW=1.5Hz.
 

Attachments

  • -120db +zes60db_s.gif
    -120db +zes60db_s.gif
    48.3 KB · Views: 274
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom