Love this paper Gerhard!
The tested C-form is both different capacity and brand, so a bit difficult to conclude reliably if what count most is the brand or the load vs. capacity.
Maybe the analysis should also include a LIPO4 cell for direct comparison.
What's the better preamp you used?
A bit OT - sorry!
The tested C-form is both different capacity and brand, so a bit difficult to conclude reliably if what count most is the brand or the load vs. capacity.
Maybe the analysis should also include a LIPO4 cell for direct comparison.
What's the better preamp you used?
A bit OT - sorry!
I don't have LIPO cells here, but I have measured 18650 Li-Ion and published it
somewhere in a discussion in the power supply section here on diyAudio.
The Li-Ion were quite good, actually.
It seems that without/small load most batteries are OK, but with load,
the bigger cells rule.
The description of the preamp is in the same directory on my web site.
It has 20 paralleled ada4898 op amps, ca 220 pV/rt Hz. It needs a larger
input coupling capacitor, 4700uF are about right.
I have burnt its successor accidentially, ( therefore the suggestion of the fuses
)
Now I'm working on a FET solution, but currently no time for it. A populated
board is done already.
I plan to repeat the measurements with the new amplifier. There will be
better results below 100 Hz, probably.
Gerhard
somewhere in a discussion in the power supply section here on diyAudio.
The Li-Ion were quite good, actually.
It seems that without/small load most batteries are OK, but with load,
the bigger cells rule.
The description of the preamp is in the same directory on my web site.
It has 20 paralleled ada4898 op amps, ca 220 pV/rt Hz. It needs a larger
input coupling capacitor, 4700uF are about right.
I have burnt its successor accidentially, ( therefore the suggestion of the fuses
)Now I'm working on a FET solution, but currently no time for it. A populated
board is done already.
I plan to repeat the measurements with the new amplifier. There will be
better results below 100 Hz, probably.
Gerhard
Update- the freeware REW REW - Room EQ Wizard Room Acoustics Software does have a rt/Hz scaling and can be adjusted for correct sensitivity. Its in a menu on the left side for vertical scaling. So one more piece for an opamp noise analyzer. (A transistor noise analyzer can also be made but has some much more complicated components to create.)
- Sorry to say this, but we can't know your purposes, because you don't know them. You only say "you want to measure noise" - but it is very wide area.Hi ! very impressive noise performance indeed
This is a very professional device i guess ... maybe too much for my purposes ? i think i will not need such quality ?
To you could understand what you need you have to start to measure something - then you'll know if it is good or not, and you'll know where to go (and we may tell you something concrete). Now we can tell you that you may measure noise with your present card. And the presented QA401 is one of the best and is wide and succesfully used for DIY audio measurements.
You may use your actual card with low-noise preamp. You may use QA401 without preamp. And of cause you may use QA401 with low-noise preamp.
Last edited:
- Sorry to say this, but we can't know your purposes, because you don't know them. You only say "you want to measure noise" - but it is very wide area.
Now we can tell you that you may measure noise with your present card.
Hi again
i would like to get a noise spectrum in the audio band And as i understand good scopes are very expensive (aren't they ? because if they are not i could buy one ... i love scopes) i have been told to use a usb sound card and i laptop. The first thing i have done is to see a spectrum of the noise floor of the sound card. My thinking is that the origin of the noise is mainly the power supply. The sound card is powered with the low quality +5VDC voltage coming from the laptop through the usb cable.I am 100% that addressing this issue will lead to lower noise floor for the card. There are indeed special cable that cut the voltage coming from the pc and allow for the use of an external +5V source. At that point we can use a very low noise power supply. That is very interesting to me. Noise comes from the power supply. The very best card have a better noise rejection of course. But even very cheap cards with this issue addressed can provide great results noise floor wise. I am sure of that.
I have bought this one ...
TeraDak U9 DC8.5V USB DC5V Power Iniezione Hi-Fi Alimentatore lineare | eBay
this psu comes with a special cable that cuts the power supply from the laptop and allows for the use of an external +5VDC power source
I would have liked better a DC plug instead of a usb plug but it is a start
If the power supply is very low noise i should see the result on the usb card noise floor. I have seen (like in the case of the QA401) that it can be very low indeed.
I will see in some weeks time ...
If i will not get -120dB across the audio range i will move upwards. Again noise can only come from the power supply. Lately i have become quite obsessed with noise. Noise is not music. Instead sometimes music can be just noise
Of course the power supply will still have noise. It looks decent but it cannot be great. I will soldered a usb female to the splitting cable to be able to use any +5VDC i can find and in this way i will test indirectly also the quality of the power supply. The lower the sound card noise floor the better the psu
Thanks a lot again
Last edited:
Hi i am quite decided on using Arta and a usb sound card now.
To get a noise spectra in the audio range.
Noise spectrum of what? A resistor, a Zener, or an amp/preamp? This needs a different "depth" of research.
- But you wanted another better sound card?
... But QA401 is USB-powered too, so that is ok.
Last edited:
The main thing which corresponds with a noise level is an ADC, but not power. Power usually is made such a quality to be just good (+-) for an ADC and DAC.
Line preamps only. Or power supplies if possible ... power supply ripple.
QA401 is really nice there is a wide consensus on this. But it is just a measurement device. I would like a usb card that can do both ... measuring and playing back music.- But you wanted another better sound card?
this i did not know. So i have to look for cards with ADC with low noise ? like ? are they very expensive ?
I think a better understanding of the basics of noise would be important here: Tom's Circuits - Reducing Random Noise in Analog Circuits | Tempo
At times its clear we are talking apples and oranges so the understanding gets lost.
To make sense of a noise measurement you need a lot of extra info: you need source impedance (preferably zero for noise voltage measurement), the gain between that input and the place where you are measuring, the noise bandwidth you are measuring, is it weighted or unweighted etc. Its possible to correct for differences but it gets complicated quickly. And its important to put the noise into context. Noise 120 dB below a signal at 90 dB SPL will be 30 to 40 dB below anything a human or other creature could hear. Actually noise in subsequent stages will dominate what you will hear.
Power supplies are important but good circuitry can reject a lot of power supply noise. Bad ground design or connections can render pointless really ambitious power supply designs. What comes out is a composite of all the optimizations and compromises of the complete system.
At times its clear we are talking apples and oranges so the understanding gets lost.
To make sense of a noise measurement you need a lot of extra info: you need source impedance (preferably zero for noise voltage measurement), the gain between that input and the place where you are measuring, the noise bandwidth you are measuring, is it weighted or unweighted etc. Its possible to correct for differences but it gets complicated quickly. And its important to put the noise into context. Noise 120 dB below a signal at 90 dB SPL will be 30 to 40 dB below anything a human or other creature could hear. Actually noise in subsequent stages will dominate what you will hear.
Power supplies are important but good circuitry can reject a lot of power supply noise. Bad ground design or connections can render pointless really ambitious power supply designs. What comes out is a composite of all the optimizations and compromises of the complete system.
Then why almost all TOTL usb sound cards do have a AC socket or a DC socket for an external power supply. But this is not important.
The very important thing that i did not know is that the actual ADC used in the card fixes the noise floor. Therefore if an ADC is noisy no power supply can tame that noise
This changes completely my point of view. I will try to see which sound card uses very low noise ADC.
Thanks again. I will stop focusing on power supplies noise then.
Last edited:
Yes ! but i understand now that he was overrating the impact of power supply noise while the noise of a card is fixed by the ADC noise.
I was looking at the wrong thing
I cannot find a comparison of different ADCs noise wise. Once the less noisy ADC are clear i can look for a card with that ADC and be done. Thanks again.
Thanks a lot for the useful advice. I will read the paper mentioned.
I am sure that different usb sound cards can provide pretty different performance even using the same usb voltage source thanks to better internal circuitry.
And the measurements show that quite clearly.
The special cable i will get will allow to use different power supplies. If i will get same noise floor i will look more to the usb cards than try to transform a ugly duck in a swan
I honestly did not know about ADC own noise being the fundamental issue in this case.
Now i understand because manufacturers often promote their new ADC chips mentioning their ultra low noise
I am trying to fix some principles ...
Now my new obsession is ADC noise ...
Last edited:
Did you look at this thread?
USB sound card for audio spectrum analysis
AK5394 seems still the best affordable ADC around.
USB sound card for audio spectrum analysis
AK5394 seems still the best affordable ADC around.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.