diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Solid State (http://www.diyaudio.com/forums/solid-state/)
-   -   Help interpreting spectrum analyzer results (pic attached) (http://www.diyaudio.com/forums/solid-state/63189-help-interpreting-spectrum-analyzer-results-pic-attached.html)

percy 26th August 2005 01:01 AM

Help interpreting spectrum analyzer results (pic attached)
 
I dont have much experience with power amps but recently I used RMAA to do some basic measurements on my power amp and there were some things I noticed in the charts that I am not quite sure if I am interpreting correctly or not.

This is the noise chart. Are those peaks you are seeing (in white) multiples of the 120hz rectifier(bridge) ripple ?

http://home.earthlink.net/~percy.mis...echnics15w.jpg

Secondly, as the load increases (low impedance) the Crosstalk increases quite a bit from -80db to about -70db to -65db.

I am trying to find out if this ripple(?) and drop in crosstalk normal or within acceptable limits ? Do both these things indicate a problem with the psu filter caps ? i.e. inadequate value or leakage or low quality etc...??

Please advise.
Thanks.

SY 26th August 2005 01:29 AM

There's ripple in the left channel. It shows as a spray of 60Hz harmonics. And in the right channel, you've got no ripple, but some hum and noise pickup- if it's supposed to be the same as the left, you've got some troubleshooting of wires to do.

Is it too much? I don't know- there's no magnitude scale that I can see.

PRR 26th August 2005 01:33 AM

> Are those peaks you are seeing (in white) multiples of the 120hz rectifier(bridge) ripple?

Yup.

Is it audible?

> as the load increases (low impedance) the Crosstalk increases quite a bit from -80db to about -70db to -65db.

Could be the common ground wire for the output terminals. 0.01 ohms shared between the two black terminals will give ~-60dB crosstalk for typical speaker loads.

But could also be measurement artifact.

Not sure that -70dB has any meaning for stereo. Might be an issue for commercial work where two different signals are going to two different rooms, and leakage is exceptionally annoying. Background music for a fine restaurant in one side, a shout-out system to page the parking attendants in the other side.

> there's no magnitude scale that I can see.

Lean on the right-arrow key a while. There ARE numbers, way-way over there. (Rant about re-sizing over-large {1000 pixels, 100KB} images omitted.)

percy 26th August 2005 01:46 AM

sorry about the large image. :o

I did another boo-boo. I forgot to mention that since I had only one attenuetor(to prevent frying the sound card's inputs) I put it on only one channel (left). That is why the right channel looks different. The right ch output was not connected to the sound card's inputs at all.


No the ripple isnt really audible as 'hum' if thats what you were asking.

percy 26th August 2005 04:36 PM

Quote:

Originally posted by percy

No the ripple isnt really audible as 'hum' if thats what you were asking.


However my actual doubt is whether this amount of ripple 'noise' has the potential of modulating the sound in any way.

About the crosstalk, all I want to find out is if the drop in crosstalk is normal behaviour or not ? I am not quite concerned about the absolute value of the crosstalk as such - be it -80db or -60db.

bear 28th August 2005 03:01 PM

Better test:

run the "other channel" within 3-6dB of max output.
If you use a suitably high frequency signal (say ~10kHz. to be not too hard on ur amp) the signal's harmonic products should be all out of band with respect to the following:

test the "quiescent" channel and see what the xtalk and ripple components look like.

That will give you more information.

Also, is this a commercial amp, or homebrew??

_-_-bear


All times are GMT. The time now is 12:57 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2