Obviously Terry9 has way more experience listening to resistors than i do! Sadly, in commercial products "fragile" is a bad word.
Anyway, he also touched on the basics of how a typical RIAA phono stage is designed. Much of the circuit is irrelevant to this discussion, but what IS germane is:
- For MC the gain will be between 60-70dB (1,000 X to 3,000 X)
- All cartridges demand different loadings for optimal linearity (sound) and to a lesser degree for impedance matching (not true in the circuit i will now describe)
The input from a MC cartridge is fairly low impedance (high current, relatively) but a tiny voltage - typically maxing out at 1/2 of 1/1000th volts. Maxing mind you - if you want 60 dB of signal to noise from the cartridge, this means the lowest signal is 1000 X lower. Think about that for a moment. To me, this is one of the two most challenging points to design in any audio chain since it must be so quiet and so linear.
Putting aside currently popular (and potentially excellent) transformers for the first stage, almost every RIAA phono stage begins with one or more amplifying devices, often in a balanced pair, with the signal (wires from c art) applied either across the differential input or from the single ended input to ground.
For duplicative redundancy, i'll repeat - the resistors in question go from the input (base, Gate,...) of these first devices to ground. They determine the linearity of the signal and the proper termination of the cartridge. Any deviation is applied directly to the input.
Until a voltage is generated across this/these resistors there is no signal, no sound. The sound that is amplified IS this voltage drop ( the "voltage divider" referenced above - that much was kinda true). This first stage amplifying device is either a bipolar transistor (BJT), a Field effect transistor (FET, whether MOS or Junction, who cares? at the moment?) or a Triode. In this application they can be made to work pretty much the same. Most designers will choose a FET or Triode since it has the highest input impedance and therefore the least loading effect. The divider, in this case, is 100% through the resistor. Let me repeat - 100% through the resistor, in the case of the BJT with a gain of 200X for laughs, maybe 99.99% through the resistor. THAT is where the signal first appears. Any distortion will be amplified. Any noise will be amplified (by 1000-3000 times, remember?)
So yea, it matters.
But back to the question - which sound best? Its obviously murky.
If you doubt there are differences, here’s a link to a guy who plotted the linearity. The website is not reader-friendly, but you can get the graphs and even the raw data.
BTW i have no idea whet that electrical engineer guy needed to be shown. Test equipment engineers sweat this stuff every day for the same reasons the OP asked his question. Its well known by really experienced, open minded (not turn-the-crank variety) engineers. Sadly, those are rare.
data and graphs on resistors here
G
