MM Phono Input impedance change 47k to 100k ohms

Thanks to both Al and Mofimadness, really appreciate the response and will stick with 47k ohms. Incidently the Grace F9Ruby seems to mate very well with the Herbies Most Excellent mat

FWIW, I have a Grace Ruby which I load at 100K, and I love that cartridge. However, I would not say that 47K is not also appropriate, only that I perceive no downside to using 100K. Now I have to go read Al's treatise on why he prefers 47K.

OK. Now I know that Al actually uses 100K, too, and with a SS phono stage, whereas mine is tube-based. Actually, I never knew that there was any written statement from Grace to the effect that 100K is preferred; I am using 100K because that's the load resistor that In installed in my phono stage input.

As to Elunkenheimer's question, I cannot imagine how the circuit could be harmed, if you were to change the 47K load resistor for a 100K one. If you know how to solder, get some high quality 100K resistors and go for it. Most phono stages I have ever seen will use a 47K resistor between hot and ground to set the input impedance. Since the grid of the tube or the gate of a transistor have a very high impedance, in the megohm range, 47K will become the dominant value at all audio frequencies. Likewise if you change to 100K. The issue is a bit more complicated if you have a true balanced input; let us know. Beware of dangerous voltages; always unplug your device, and leave it off power for at least a full day before doing any soldering. (A full day does not make it perfectly safe, however.)

Mofi, thanks for the excellent and valuable empirical information.

Lew, thanks for your inputs as well. I think that you are correct that in MOST cases changing the 47K resistor to 100K would not have any adverse effects on phono stage performance, especially if the input stage is tube or FET-based. But it goes against my technical instincts to incorporate a design modification, albeit a seemingly minor one, without having a full understanding of what the existing design is.

And in particular, I was envisioning the possibility that the active stage at the input may utilize bipolar transistors, and therefore most likely have a much lower input impedance than a tube or FET-based circuit. Conceivably, also, some of the passive components associated with RIAA equalization could be located at that point. I seem to recall having seen some Bryston schematics, for example, to which both of those points would apply.

Finally, given that the cartridge is a MM I would assume that its inductance is somewhere in the vicinity of 100 to 500 mH. That means its impedance at ultrasonic frequencies at which it may output significant energy may be upwards of several 10's of Kohms, perhaps even approaching 100K. That would clearly be significant in relation to the phono stage input impedances we are talking about, implying the possibility (if the input impedance of the active stage is not high enough to be insignificant) that the resistor change could result in adverse effects on noise levels, frequency response, intermodulation distortion, etc.

It all adds up to caution being called for, as I see it. Especially since, given Mofi's experiences and my theoretical speculation (in the other thread), 47K stands a good chance of working just as well or perhaps even better than 100K.

Best regards,
-- Al

Al, you, Lew, and many others are way above my knowledge of electronics. But as I understand this one can load down but not up.

Therefore, changing the hard-wired input from 47K to 100K still allows it to be loaded down to 47K (with parallel inputs or Y adaptors) as well as 1K, 470, 40 ohms, etc., to find the best match for a given cartridge. But one cannot upload from 47K to 100K or any higher value.

Indeed, a few older preamps provided switched loading and included 100K as an option.

I'm a strong believer there are no best/single answers in this hobby, too many variables. So it usually comes down to trying any option for ourselves in our own system. It seems changing the phono input to 100K simple gives us one more option.