Rowland 625 versus Audio Research DS450

I cannot directly respond to your question, not having heard either amp, but it's interesting to see that ARC is finally building a no-global-feedback amp (the Rowland is, too). The analog power supply is another reason for hope.

Regarding Rowland, the Model 2/6/8/9 vintage is generally considered the pinnacle of his designs - they featured battery power supplies and outrageous parts quality that today's much smaller market could not support.

Hi Lougiants - I remember the Model 10's seemed more "lush" and tubelike, in comparison to the 201's. If there is one minor area lacking, the Model 10's did not seem to have as much bass control as the 201's.

I really liked "quickness", low level detail, mid-range and bass control of the 201's. I decided to go with 201's at the time, as I needed a meaty amp to control Thiel's. (And the 201's are nice and small...)

Hindsight being 20/20, I think I probably would have preferred the Model 10's (over the 201's) with a more efficient speaker. Subjectively, the 10's seem to have more "soul" or body than the 201's.

The 201's relatively is a more neutral amp, but nonetheless still musical. It's a matter of preference.

The 201/501's straight into the wall are heavily dependent upon the quality of the A/C for their sound, much more so than a traditional amp using an analog power supply, so it can be misleading to judge performance in a given system. This is the primary reason that Rowland uses power factor correction circuitry in many of his amps, and used to offer a stand-alone power factor correction device, the PC-1, as an option for the 201/501.

Raquel, Icepower amps by definition should be less sensitive to power line than linear supplies since SMPS are both line and load regulated.

In reality all supplies are switching only frequency is different. Linear supplies switch at 120Hz polluting mains with narrow current spikes while 120Hz ripple on the secondary (amp) side requires a lot of caps to clean it up. For that reason Rowland started using SMPS even in preamps (Capri) where efficiency is not important. Also newest Rowland class AB amps (625, 925) use switching mode power supplies. PC-1 is further improvement (energy storage) but main supply of voltage to speakers is SMPS.

SMPS got bad rap from cheap computer supplies but properly executed deliver great performance with very little noise (zero voltage, zero current switching). I'm not sure how Rowland made it at 1MHz since it is difficult to get good efficiency because of limited slew rate switching losses (but easy to filter and get very fast load response).

Rowland posted few articles/answers on switching power supplies:

http://jeffrowland.com/knowledgemanager/categories.php?categoryid=17

Have to agree with Raquel... I also think that the 2/6/8/9 amps would be a tough deal to produce today due to the smaller market. I have heard most of the current amps--exception the 625, and IMHO, while nice; they don't really play in the same arena as the 2/6/8 and 9.