I've seen some specifications on user/service manuals (while hunting for specs on vintage stuff) where the signal to noise ration is presented like this:
SN ratio (short-circuit, A network):
Phono MC 86 db
Tuner, CD, ... 106 db
Can someone comment on how this compares to say, IHF 66 for example, which has normally much higher values in some stuff
For example:
S/N 100 db (106 db, IHF, 66)
other times it can be like:
S/N 100 db (106 db, IHF, A)
How do these compare in practice and in the last two examples how can one have a real idea of the 1st value presented (I assume it's the manufacturer's but it doesn't say anything else regarding the test conditions so probably only the values between brackets are of any meaning for comparison purposes ?)
THanks
SN ratio (short-circuit, A network):
Phono MC 86 db
Tuner, CD, ... 106 db
Can someone comment on how this compares to say, IHF 66 for example, which has normally much higher values in some stuff
For example:
S/N 100 db (106 db, IHF, 66)
other times it can be like:
S/N 100 db (106 db, IHF, A)
How do these compare in practice and in the last two examples how can one have a real idea of the 1st value presented (I assume it's the manufacturer's but it doesn't say anything else regarding the test conditions so probably only the values between brackets are of any meaning for comparison purposes ?)
THanks
Short circuit, A must mean that it is measured with the input shorted, and the output through an A-weighting filter.
A-weighting corresponds better to our hearing sensitivity versus frequency but also gives higher values than without filtering/shaping.
But what is also important is the reference input it is measured against.
For instance, if your design input level is 1V and the output 'noise' level is 100uV than the S/N ratio is 1/10,000 which is -80dB.
But if your design input level is 100mV, than with the same noise level, that would be only -60dB.
So you need to know the levels in each case for a meaningfull comparison.
jan didden
A-weighting corresponds better to our hearing sensitivity versus frequency but also gives higher values than without filtering/shaping.
But what is also important is the reference input it is measured against.
For instance, if your design input level is 1V and the output 'noise' level is 100uV than the S/N ratio is 1/10,000 which is -80dB.
But if your design input level is 100mV, than with the same noise level, that would be only -60dB.
So you need to know the levels in each case for a meaningfull comparison.
jan didden
When quoted in this way, S/N is usually referred to nominal full power (typically at 8 ohms) with an input signal at nominal sensitivity.
Measurements referred to 1 W / 8 ohms (2.83 Vrms) or 50 mW / 4 ohms (447 mVrms) of output are where the wheat and the chaff separate. Amplifiers with a straightforward gain structure (read: volume pot followed by about 45 dB) rarely exceed 70 dB referred to 50 mW / 4 ohms. This works out to about 0.8 µV input-referred or 5.6 nV / sqrt(Hz) of equivalent input voltage noise density - some of that comes from the amplifier itself, some from the source impedance it sees (depending on volume control pot). It's not a terrible value at all, but still amounts to about 140 µVrms of output noise... average performers tend to be around 300, noisy ones up to 500 µV.
If you find an amp that's a lot less noisy, chances are it has better gain distribution. Here's a little article of mine on the subject of noise in audio amplifiers that discusses such variations.
Measurements referred to 1 W / 8 ohms (2.83 Vrms) or 50 mW / 4 ohms (447 mVrms) of output are where the wheat and the chaff separate. Amplifiers with a straightforward gain structure (read: volume pot followed by about 45 dB) rarely exceed 70 dB referred to 50 mW / 4 ohms. This works out to about 0.8 µV input-referred or 5.6 nV / sqrt(Hz) of equivalent input voltage noise density - some of that comes from the amplifier itself, some from the source impedance it sees (depending on volume control pot). It's not a terrible value at all, but still amounts to about 140 µVrms of output noise... average performers tend to be around 300, noisy ones up to 500 µV.
If you find an amp that's a lot less noisy, chances are it has better gain distribution. Here's a little article of mine on the subject of noise in audio amplifiers that discusses such variations.
- Status
- Not open for further replies.