1audio said:
I can't find the original paper, I think there was a link on Dan Duggan's nature recording site. B&K actually has some specialty mikes. The paper below is research but I have seen a 0dB SPL one for sale. There are also RF mikes that do 7dB which are what I think the survey was done with.
MICROPHONE SYSTEM FOR EXTREMELY
LOW SOUND LEVELS*
by
Erling Frederiksen
Conclusion
An experimental one inch microphone system with a flat frequency response up to 13 kHz has
been developed; a system which is able to detect extremely low sound pressure level – the
inherent A-weighted noise of the system is –4 dB or about 15 dB lower than the noise of any
corresponding systems.
Edmond Stuart said:
Your sonic acuity will not be enhanced by sitting 10cm away from a 110dB/2.83V/M loudspeaker.
What is the ambient SPL where you are?
SY said:
Hi Stuart,
That's right, but I thought we were talking about ultra low noise, not acceptable noise. So I won't hear any hiss from any distance from my speakers. 🙂
Cheers,
Edmond.
I am unsure about noise floor discussions. Noise floor is usually about white noise. I have no idea what the ratio is , but I'm reasonably sure that the ear can hear coherent signals well below the noise floor. As an example the ability to focus on and decipher a conversation at a loud party or during loud music.
Or perhaps a similar point, hearing a Leopard step on a twig near a pack of noisy monkeys.
I am not advocating that your system should hiss happily, I find that quite annoying. It's just that like sine waves, white noise doesn't seem to have a lot of relation as to how we hear reproduced music.
Useful tool, yes. As a measurement of a given system's resolving power, no.
Or perhaps a similar point, hearing a Leopard step on a twig near a pack of noisy monkeys.
I am not advocating that your system should hiss happily, I find that quite annoying. It's just that like sine waves, white noise doesn't seem to have a lot of relation as to how we hear reproduced music.
Useful tool, yes. As a measurement of a given system's resolving power, no.
People are talking about noise in a very haphazard way.
Pot position is almost everything in determining how much noise that it will contribute. IF the volume control is located at the input of a power amp, it usually will not contribute too much noise, because the gain FOLLOWING the pot is limited.
However, if the volume control is located in front of the preamp, (as it usually is) then the actual resistance of the volume control can be a determining factor as to its audibility. This is because the preamp gain multiplies the noise considerably.
Putting a volume control at the input of a phono stage is not realistic, and will almost always ruin the S/N ratio.
The value of the total resistance of a volume control is set by several tradeoffs. The first, being the loading effect on the sources. For example, if you use tube electronics, then anything below 50Kohms could seriously effect the distortion and low frequency response from any tube source.
Sometimes designers will have a special 'tube' input in quality preamps that will add an extra resistor in series with the pot to increase the load resistance for that input.
In any case, the worst case noise (in general) is 1/4 the nominal value of the pots total resistance. This occurs at 6dB below max output only. All other positions on the pot give lower noise.
The rule of thumb of the noise generated by a resistor is:
10K is 13nV/rt Hz
100K is 39nV/rt Hz
This a a 10dB difference between the 2 pots. However, please remember that the EFFECTIVE worst case noise is 1/4 the maximum resistance or 1/2 the noise voltage.
Generally volume control resistance below 10K is not practical as it adds distortion to the audio system due to excessive loading of the sources.
Pot position is almost everything in determining how much noise that it will contribute. IF the volume control is located at the input of a power amp, it usually will not contribute too much noise, because the gain FOLLOWING the pot is limited.
However, if the volume control is located in front of the preamp, (as it usually is) then the actual resistance of the volume control can be a determining factor as to its audibility. This is because the preamp gain multiplies the noise considerably.
Putting a volume control at the input of a phono stage is not realistic, and will almost always ruin the S/N ratio.
The value of the total resistance of a volume control is set by several tradeoffs. The first, being the loading effect on the sources. For example, if you use tube electronics, then anything below 50Kohms could seriously effect the distortion and low frequency response from any tube source.
Sometimes designers will have a special 'tube' input in quality preamps that will add an extra resistor in series with the pot to increase the load resistance for that input.
In any case, the worst case noise (in general) is 1/4 the nominal value of the pots total resistance. This occurs at 6dB below max output only. All other positions on the pot give lower noise.
The rule of thumb of the noise generated by a resistor is:
10K is 13nV/rt Hz
100K is 39nV/rt Hz
This a a 10dB difference between the 2 pots. However, please remember that the EFFECTIVE worst case noise is 1/4 the maximum resistance or 1/2 the noise voltage.
Generally volume control resistance below 10K is not practical as it adds distortion to the audio system due to excessive loading of the sources.
john curl said:
That is EXACTLY what I spoke about. My 50k pot example is 12.5k effective resistance and it ADDS 15nV/rtHz at the input of low-noise preamp. (Speaking in terms of 4kTR contribution, there is still input current noise density to be taken into account).
Howgh 😀
My apologies, I'm off in my own world and forgot standard practice. Phono is my only non-line level source so I use a passive preamp (just the stepped attenuator). I set my phono up to give line level out.
dimitri said:
I think you got the basic idea and yes you can dispense with the extra bias sources like you did with only a small hit on distortion. When I simmed your way (optimized values) I got .004% distortion at 10V p-p, not too shabby for open loop.
john curl said:No op amps Scott? Less is best? What does ADI think about this?
I take some flak, but in the end nobody really cares.
MY zen phono (with due apologies to Nelson)
A few people have asked so here is the phono front end that I have been using. I had, due to the kindness of a friend, 4 pairs each of 2SJ73’s and 2SK146’s and I had just seen Nelson’s ZEN article. So I said let simple rule the day and what do you get. As I said I wanted differential in and the P first stage and N second stage is a nod to good karma if nothing else. This set up has given me some beautiful music and to be honest it has performed as well as anything else I have owned. It was in a well shielded box with a super regulated low noise supply.
So this is my ‘ZEN’ phono front end. BTW I study traditional zen painting as another hobby.
EDIT - All high quality film caps in the signal path, no servos, electrolytics on the supplies only.
A few people have asked so here is the phono front end that I have been using. I had, due to the kindness of a friend, 4 pairs each of 2SJ73’s and 2SK146’s and I had just seen Nelson’s ZEN article. So I said let simple rule the day and what do you get. As I said I wanted differential in and the P first stage and N second stage is a nod to good karma if nothing else. This set up has given me some beautiful music and to be honest it has performed as well as anything else I have owned. It was in a well shielded box with a super regulated low noise supply.
So this is my ‘ZEN’ phono front end. BTW I study traditional zen painting as another hobby.
EDIT - All high quality film caps in the signal path, no servos, electrolytics on the supplies only.
Attachments
Scott,
Correct me if I'm wrong, but are you accepting balanced input from the cartridge, then using a single-ended output? It appears, though I'm not certain, that your second stage (right-hand JFET of the differential) is taking a balanced input from the first stage. There's what I'm assuming is an input arrow partially hidden beneath the degeneration resistor. Yes?
Grey
Correct me if I'm wrong, but are you accepting balanced input from the cartridge, then using a single-ended output? It appears, though I'm not certain, that your second stage (right-hand JFET of the differential) is taking a balanced input from the first stage. There's what I'm assuming is an input arrow partially hidden beneath the degeneration resistor. Yes?
Grey
GRollins said:
Yes sorry for the less than perfect drawing, I split the phono termination (300pF and 23.7k on each side in my case) and fed differential from the first stage to the second stage. The second stage does diff to single conversion with one FET doing what I hope is a little bootstrap/cascode at the same time. The idea was to not mess with the lowest (highest gain) RIAA time constant with the output resistance of the FET's. I never actually measured the effects. An exercise in minimalism.
BTW it's the standard Jung all in one passive RIAA.
- Status
- Not open for further replies.