Soha II/Mosfet-MAX for brightish headphones

I'm not going to comment on the SOHA II because it wouldn't be proper as a MOT with a competing product.

What I can say is that MOSFET-MAX is a different animal than the MiniMAX.  It's optimized for high-current, with not as much of a voltage swing.  Class A bias is recommended at 125ma on the diamond buffer output MOSFETs.  That's well beyond the Class A bias of any BJT amp, so current is the thing. It sounds great with Grados and inefficient orthodynamics, but the MiniMAX is a better choice (than the MOSFET-MAX) for Sennheisers or 600 ohm Beyers.  Put in nutshell, the Millett Hybrid design, whether MAX or MiniMAX, tends toward the solid state buffer rather than the tubes.  If I had to guestimate, I'd say 70-30 for the MOSFET-MAX ss buffer vs. tubes and about 60-40 for the MiniMAX.

Of course, if you're looking for something really optimized for Grados, especially the RS-1's or HF-2's, keep an eye out for the Dsavitsk/Beezar Torpedo (see Head-Fi thread of the same name).  It's a true tube-transformer amp that's completely finished on a PCB with a custom case.  IMHO, there's no one like Dsavitsk who can design an amp for Grados.  Rather than "laid-back" or "less-forward," I prefer to think of it as Grado-optimized.  (It's pretty good for a whole lot of other headphones, too, but with Grados, it really SINGS.)  We hope to begin sales for short-kits in the next couple of weeks.

Quote:

Originally Posted by Amarphael 

tomb, I appreciate your attention. This Torpedo amp looks interesting and simpler design than the MosfetMax or Soha-II, Which is better for me, I'll definitely look into it. But what makes it optimsed for Grado's espcially? Can you describe it's sound charcter if any? I'm looking for an amp that sounds towrds a neutral response yet with ample 'body' to it.  

Descriptions are difficult for me.

Suffice to say that Dsavitsk has discovered the way to keep Grados thumping with body and detail, but no harshness.  He's often times described that the drive to super-low output impedance (high damping factor) that everyone goes for with solid-state, is actually counter-productive with Grados.  I can't remember the exact numbers, but through his extensive testing with amp designs and builds, he's found that Grados actually like higher output impedances, so much that the output impedance of the better-sounding amps actually have are equal to or greater than the 32 ohms of Grados.  Again like I said, that's totally counter-intuitive with the conventional wisdom of the most famous DIY amps, which strive for very, very low output impedances.  All I can say is that Dsavitsk's amps - in particular the Torpedo - sound better even with the cheapest Grados than most other amps.

That may be far from a typical "Head-Fi" review description that often includes elaborate, melifluous descriptions that are seen in the Headphone and Amplifier sections of Head-Fi, but that's sort of the way I see it.

I'll add a little bit here -- basically, anytime you use output transformers (and don't use a feedback loop), the DCR of the transformer is in line with the phones, leading to a bit of a higher Zout.  Grados, with a very flat impedance curve, do well with a higher Z out.  Thus one of the reasons Grados and "tubes" work well together is actually the Z from the transformers. On the Torpedo, for the Low Z setting, this is about 6 ohms or so.  Additionally, because the rp of the tube is ~3K, and we are lowering that by a factor of over 300 with the OPT, this adds  another 10R and thus a Zout in the 15-18R range.  I think anything between 5 and 20R is pretty good for Grados with a voltage source amp -- higher than that they can start to sound a little too tubey (or maybe tubby). With a current source amp, the sky is the limit -- my own current source amp has a Zout around 400R which I like a lot. One of the cleanest sounding amps I have.

Additionally, because most other dynamic phones have a relatively flat impedance curve, this is also true for them.  Sennheisers may in fact be the odd man out on this, so they may benefit the most from a higher damping factor (though even there, a near 0 Zout is not necessary).