Not strictly DIY but I'm considering buying a non-functioning preamp to rebuild (Audible Illusions Modulus 3) with a 1.2Kohms output impedance. My favorite amp (rebuilt N.E.W. A20.1) has a 23.1Kohms input impedance. IN THEORY this is borderline, probably OK given the 19X difference between the pre's output impedance & amp's input impedance. I also called Audible Illusions and he was confident it's OK given the amount of current drive the Modulus 3 puts out. What are your thoughts/experiences here?
Perfectly fine.
Ideally the preamp also should have an output coupling capacitor of at least 3.3uF,
and it does appear to have at least that.
Ideally the preamp also should have an output coupling capacitor of at least 3.3uF,
and it does appear to have at least that.
Last edited:
For greatest power transfer between source and load, the output:input impedance ratio should be 1:1; however...
For the lowest noise floor the output:input impedance ratio should be ~1:3; however...
For active sources for the lowest THD the output:input impedance ratio should usually be >1:10.
So theoretically if you are trying to get a telephone to work from one side of the country to the other use 1:1 or more specifically 600Ω:600Ω (which also matches the transmission line impedance of the spaced apart telephone wires).
If you want the lowest noise from a microphone, for example, use 1:3, which is a bit indeterminate because microphones vary widely in output impedance from ~50Ω up and sometimes sound better with a higher load than that optimal for noise.
And if you have a consumer grade audio component with the typical output impedance of around 2kΩ and you want the lowest colouration use ~1:10, i.e. 20kΩ or greater. Your source impedance of 1.2kΩ is going to be in love with the 23.1kΩ load.
For the lowest noise floor the output:input impedance ratio should be ~1:3; however...
For active sources for the lowest THD the output:input impedance ratio should usually be >1:10.
So theoretically if you are trying to get a telephone to work from one side of the country to the other use 1:1 or more specifically 600Ω:600Ω (which also matches the transmission line impedance of the spaced apart telephone wires).
If you want the lowest noise from a microphone, for example, use 1:3, which is a bit indeterminate because microphones vary widely in output impedance from ~50Ω up and sometimes sound better with a higher load than that optimal for noise.
And if you have a consumer grade audio component with the typical output impedance of around 2kΩ and you want the lowest colouration use ~1:10, i.e. 20kΩ or greater. Your source impedance of 1.2kΩ is going to be in love with the 23.1kΩ load.
I would suggest:
- yes, the combo will work OK (with the 19x Zout/Zin ratio)
- but it would sound better if there was a 30x ratio.
Is there any reason why the 23K1 Zin of your NEW A20.1 can't be increased to, say, 33K ... or even 47K?
Andy
higher ratio less likely to cause device distortion, or filter distortion from coupling capacitors being too small.
this case active devices less likely a issue, passive devices different story
this case active devices less likely a issue, passive devices different story
What's the characteristic impedance of the cable you are using?
There's actually a good discussion of source and load impedance, the losses involved from mis-matched impedances on https://en.wikipedia.org/wiki/Standing_wave_ratio
There's actually a good discussion of source and load impedance, the losses involved from mis-matched impedances on https://en.wikipedia.org/wiki/Standing_wave_ratio
For optimum impedance matching calculate the RC of the AI's output stage (IIRC stock - 3.3uf / 540KR) with the input Z of the Amp. Shoot for a low corner Freq.
The rp is closer to 1K5 as AI runs the plates in parallel. If someone can find the actual plate resistance of the 6922 please post it. IIRC it's 3K perhaps 3K3. I'm not certain of this.
Cheers,
W