Shelter 901: Optimal Impedance

I've owned a Shelter 90X and currently use a Shelter 9000 and have found 38 ohms to work best for me (my phono pre uses transformers). The Shelters do seem to share some family characteristics. One easy way to determine optimum loading is to decrease the resistance (increase the load) until the soundstage collapses then back off one setting. Some like to run their MCs wide open. Maybe their high freq. hearing is shot or maybe they like to listen to surface noise. Anyway, play around with it and see where you like to run your MC. That's one of the nice things about this hobby. You can set up your system to sound the way you like.

The internal impedance of the Shelter 901 is 20 ohms and the rule of thumb I use is to start with 10 times the internal impedance, and adjust from there. I am using my 901 at about 180 ohms and am happy with it. The specs say 100 ohms but I found that it stifled the high frequencies a bit. When you go higher, it opens up the high end, but with some bright cartridges like the Denon 103R, it gets too bright. When you go lower, it deadens the high end and sounds like it is choking the highs. Some call it a muddy and flat sound. I personally haven't found a problem with going with a higher impedance on the Shelter 901, so 470 would probably sound fine. It is always better to go higher than lower.

The best answer to your question is another question, viz., "Which MC phono stage are you using?"

As several have pointed out, the optimal impedance for an MC is system dependent. To be more specific, the type of MC gain has the greatest influence. With the exception of John Tracy, every poster so far has assumed an active MC phono stage. John's much lower number is based on using step up transformers, which is very different.

I owned a 901 and the optimal impedance varied considerably depending on which of four SUT's I used. When I switched to an active MC phono stage the optimal impedance changed even more (higher).

Unless you're using the same MC gain stage, no one else's specific number would be applicable to your system. This is especially true with SUT's, which are highly variable, but it's even true with active stages.

So, what are you using?