|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
151.33.226.172
In Reply to: RE: You may be posted by fmak on June 08, 2016 at 06:32:55
Hi !
main problem is to find reliable measurements
I bought the one in the link in the 12VDC version to use with a usb interface based on instinct
I should receive it soon and be able to get some noise spectra with Arta software
Kind regards,
bg
Edits: 06/08/16 06/08/16Follow Ups:
Probably equally important to find relevant measurements.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
HI ! Thanks a lot for the kind and helpful reply.
yes ! absolutely.
But almost never the manufacturers provide clear specifications and even more rare are noise spectra.
So it is a long and costly buy and test method.
As i said i have an interface that needs 12VDC/2A
The stock psu is a cheap and little smps thing.
With the stock psu i see spikes (honestly low in level but still evident) standing out above the nicely low noise floor.
I am about to test other and supposedly better psu (switching and linear) to see if the spikes go away. Maybe not.
So in the end i need a very clean 12VDC/2A power supply.
I understand that usually the problem with many psu is some residual high Hz ripple.
Unfortunately i get only spectra up to 48kHz. Better the nothing.
The spikes are always at the same frequencies even changing psu (i have tried only different cheap smps by the way. I will start soon with the linear one depicted in the post above).
Thanks again.
Kind regards,
bg
"The spikes are always at the same frequencies even changing psu"
If various power supplies you are using are not identical then this is likely a clue pointing to the noise originating somewhere outside of the power supply.
If you post a screen capture of your noise floor samples, along with details about your setup, perhaps forum members could make useful comments about the noise you are seeing.
Hi !
i am attaching the noise floor spectrum of the interface
It looks almost the same with different cheap smps.
Tomorrow i will try a better linear psu just out of curiosity.
Thanks and regards.
Kind regards,
bg
The biggest peak, around 37kHz could be emi from some switching power supply in your environment. Though, often times it is the audio systems immunity to these types of interference which is more the blame than the switcher when quality switching power supplies are being utilized.
Note that a low level, high Q spike at ~37kHz is not likely to be related to MHz band instabilities in the power supply output of your usb cable wart.
When doing loopback testing such as you seem to be doing, especially in a noisy pc environment where single ended audio gear is being utilized, the ground loop problem is a huge concern.
In order to believe what you see in these types of setups is you have to beak that loop in the D/A to A/D chain since it may even be the culprit you are now chasing!
Do you have some type of loop breaking isolation inserted into your loopback chain if you are using single ended signalling? Or, are you using all truly balanced audio hardware?
If the answer to both those questions is no then the loopback connection itself can be a major contributor if not the sole contributor to these types of noise entering the audio signal.
This is the technical difficulty of doing believable loopback.
Hi ! thanks a lot for the very kind and valuable reply." The biggest peak, around 37kHz could be emi from some switching power supply in your environment.
Though, often times it is the audio systems immunity to these types of interference which is more the blame than the switcher when quality switching power supplies are being utilized. "Exactly ! Other interfaces I tried with the same exact set-up (only the interface changed) have lower noise and almost no spikes to speak of ... or at least not evident like the one in the picture.
However, and just for comparison purposes, i am attaching the noise powering the interface with a linear psu ... same spikes more or less than with the cheap smps.
So it is noise generated inside the interface, maybe for not enough filtering. Usually these interface take the 12VDC supply and generate the different voltages needed with dc to dc switching converters.
There must be a significant difference in the quality of this power generation between cheap units and more expensive ones.
Even if some exceptions could exist." Note that a low level, high Q spike at ~37kHz is not likely to be related to MHz band instabilities in the power supply output of your usb cable wart.
When doing loopback testing such as you seem to be doing, especially in a noisy pc environment where single ended audio gear is being utilized, the ground loop problem is a huge concern.
In order to believe what you see in these types of setups is you have to beak that loop in the D/A to A/D chain since it may even be the culprit you are now chasing!
Do you have some type of loop breaking isolation inserted into your loopback chain if you are using single ended signalling? Or, are you using all truly balanced audio hardware?
If the answer to both those questions is no then the loopback connection itself can be a major contributor if not the sole contributor to these types of noise entering the audio signal.
This is the technical difficulty of doing believable loopback. "Actually what i am doing is just measuring the noise floor with open input. There is no analog connection between the DA and AD section.
So i guess it is not a complete loop-back. Both ins and outs are left open, not connected.My conclusion is that changing PSU has no significant impact on the noise floor of the interface, as long as the PSUs used are of decent quality.
The best interfaces have a very low and flat noise floor, evidence of a superior design and construction.
Of course there is also the distortion ... but this is more difficult to measure.
I hope to be able to carry out distortion measurements soon.
Thanks a lot again.
Kind regards,
bg
Edits: 06/11/16 06/11/16 06/11/16 06/11/16 06/11/16 06/12/16 06/12/16 06/12/16 06/12/16 06/12/16 06/12/16 06/12/16
Hi!
thanks for the helpful advice.
I will do it soon. I am away from home now.
Yes i tried different 12VDC smps (all cheap things by the way).
But i found interesting that the spikes were always at the same Hz and more or less of same levels.
I have a linear ps to try also.
I will post the results.
Kind regards,
bg
With that power supply, there is an instability issue in the MHz range. Basically, to get rid of most of it, you will need to decouple the power supply pins of the NE chip with 0.1uF npo ceramics and a 4.7u PP cap across the power pins.This occurs not only on one, but three units. You'll need a wide band scope to see all this clearly and also put the unit on load to make sure.
Edits: 06/09/16
"With that power supply, there is an instability issue in the MHz range."
Without you providing supporting evidence for your bold proclamations they cannot be useful to anyone.
You will need to provide evidence of your findings for them to be taken seriously.
Who do you think you are? Don't believe me and buy one of these.
Hmm. The exact same answer someone who was trying to hide the fact they have absolutely no clue might have given....Hmm
Yeah, who do you think you are anyway? Everyone here makes purchasing and modification decisions based on posts by internet buffoons. What's wrong with you?!
Hi thanks a lot for the very kind and precious advice
I think I will be able to do what you recommend
I can only see noise floor up to 48kHz
Are you aware of very good 12V/2A psu
maybe?
Kind regards,
bg
Edits: 06/09/16
It'll cost you bought : ie Paul Hynes.
But you can make one from the ALW version of the superregultor (with heatsinking) or use a Bellioson ready made unit (with heatsink). The former needs quite a lot of work.
Thanks a lot again for the helpful advice.Usually i see in the specifications noise figures given up to 100 kHz.
I wonder what kind of issues noise located above 100kHz (i.e. in the MHz region) can cause.
Some clock irregularity/instability maybe ?
with my current set-up i can only check the noise floor up to 48kHz (using 96k sampling rate)I looked at the usual regulator chips and as you said before they are quite uneffective at suppressing higher Hz noise.
There are some a little better but kits using them are very rare.
I will look for the brands you mention.
In the meantime have found the Belleson site.
Thanks a lot again.
Kind regards,
bg
Edits: 06/09/16 06/10/16 06/10/16
Have you done wiring separated power line and grounding? That may fix your problems.
Regards,
Windows X
Hi thanks for the valuable advice
I am not sure to understand rightly
The device has only a two contacts dc socket without ground
I would try a battery but it's too un convenient
I wonder if the spikes are typical of the unit as i tend to believe
I will check
Kind regards,
bg
Edits: 06/10/16
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: