Audio Research 100.2 Power Amplifier Review

[Blumlein 88]

I wonder like Ray and mitchco about the caps. I heard one of these with rather efficient speakers when it was current. There wasn't any hum audible, and at the level in your measures I think it would have been.

Now the distortion products I think are the norm for these. The owner's manual says 3 db of feedback.

 

[spacevector]

I read some place that it has less feedback so it can deliver less phase shift.

Is this a test your analyzer can run in conjunction with the Frequency Response test? So when showing the amplitude of the frequency response, also show the phase? It may be interesting to see how different design choices - Class A, A/B, D, etc. and amount of feedback used affect the phase response.

 

[SEKLEM]

"FEEDBACK IS EVIL!!!!!!!!!!!!!"

-Dan

I know that's not what he said. He's careful not to be critical of anything in that interview. Probably a safe bet to not absolutely deride anything, lest it come back and bite them in the rear.

 

[gvl]

A very nice enclosure for a Neurochrome Modulus build.

 

[DonH56]

A number of amplifier's "back then" were slow to decay. It got to be a selling point for how long the music played after you turned off the amplifier -- people equated it to large power reserves and all that jazz. ARC used to tout their capacitance reserves in Joules (sounds impressive, scientific'y and all that, but few consumers actually had any understanding of it). So, a feature, rather than a lack of bleeder resistors, improperly shutting down bias circuits, and limited output protection. More than one 'phile lost a woofer or two due to nasty DC surges as the amps powered off. Rare, but very vexing when it happened.

 

[SIY]

According to this interview, Dan D'Agostino was admitedly very measurement centric during his final Krell years. Later with his own brand he relaxed on metrics and focused on subjective.

FWIW, I visited Dan a few months ago, and the place is stuffed with Audio Precision gear.

 

[Milt]

A number of amplifier's "back then" were slow to decay. It got to be a selling point for how long the music played after you turned off the amplifier -- people equated it to large power reserves and all that jazz. ARC used to tout their capacitance reserves in Joules (sounds impressive, scientific'y and all that, but few consumers actually had any understanding of it). So, a feature, rather than a lack of bleeder resistors, improperly shutting down bias circuits, and limited output protection. More than one 'phile lost a woofer or two due to nasty DC surges as the amps powered off. Rare, but very vexing when it happened.

I had a Rapaport AMP-1 back in the late 70's early 80's.
Rare, esoteric & a really quirky design. 100wpc full class A.
Music would play for quite awhile after the amp was turned off.
And yes Don, I experienced the dreaded DC surge that welded the VC on one of my Harbeth drivers.

Other than nostalgia I don't get the vintage gear thing from the proponents standpoint that gear was far better made in days of yore.
Especially with speakers.
I agree with you that the folks who like this stuff are after a certain sound that has nothing to do with accuracy.
Trust me I'm not looking down on folks with this POV at all, different strokes for different folks, right?

 

[Doodski]

By the late '70s ARC was seriously considering getting out of the tube business. Their 'high end' street cred was mostly based upon the many SP3 preamplifier versions, which because of their accessability were taking the Dyna mod route--i.e. cottage designers like Paoli were modding the preamps, adding that special magic only they could provide.

Perhaps in order to stop this sort of thing, ARC came out with their 'Analog Module' scheme, which as near as I remember were ICs or opamps potted in thick goo, so a third party couldn't 'fix' what they didn't like. The upside was that all repairs had to go through ARC dealers. The downside was that after a few years you couldn't get Analog Modules anymore, making your preamp or amp pretty useless if it broke. Of course this sort of thing wasn't endemic to ARC... recall Hitachi MOSFETS (Acoustat) and Sony V-FETS, et al.

View attachment 44502

I've replaced a few of those AM-1 modules. After a thorough rebuild one crosses their fingers and winds up the variac hoping the modules don't blow.

 

[Matias]

FWIW, I visited Dan a few months ago, and the place is stuffed with Audio Precision gear.

Probably for checking and testing, but not as an ultimate goal of reducing noise and distortion from what I learned from that interview.

 

[Dialectic]

I think the conventional audiophile concern about this amp would not be the horrendous measured performance but, rather, the inevitable MOSFET mist.

 

[Blumlein 88]

I think the conventional audiophile concern about this amp would not be the horrendous measured performance but, rather, the inevitable MOSFET mist.

Yes, the Mosfet Mist. In this case there may be some mist with that noise level.

 

[SIY]

Probably for checking and testing, but not as an ultimate goal of reducing noise and distortion from what I learned from that interview.

Maybe. Or maybe he’s thinking about what his intended target customers want to hear. There’s AP gear in the production area, but over on the design and engineering bench, there’s a very nice APx525...

 

[AudioTodd]

Almost not surprised considering they left the distortion spec blank. Then instead of listing SNR they listed a made up spec,
"Hum & noise 150 microvolts RMS (105dB below rated output IHF A-weighted)"

Ha! Those specs would be just as useful if they had simply said “Power switch turns it on.”

 

[AudioTodd]

Someone really needs to write a history of hifi.

What? You don’t trust Robert Harley’s stuff?

 

[Frank Dernie]

Maybe. Or maybe he’s thinking about what his intended target customers want to hear. There’s AP gear in the production area, but over on the design and engineering bench, there’s a very nice APx525...

I thought the same.
As an experienced engineer I imagine he knows how to design and measure his work, as a successful marketing man he knows what to say in interviews...

 

[Addicted to music]

In fairness to AR they could hardly demand that Toshiba keep manufacturing an obsolete part just so that people could repair 20 year old amplifiers. (Toshiba 2SA1303 and 2SC3281)
Where they are annoying is in carefully sandpapering the manufacture's markings off all the low power semiconductors and marking them with painted on colour bands. If you need to replace anything in the voltage gain stage you need to order a colour coded part from AR. They won't tell you what it actually is. Some if these are almost certainly hand selected parts, matched to a specific parameter value, not just random parts off the production line. Even so, it seems a bit mean.

I’ve seen sandpaper and rubbing the inked print off to hide the real OEM on many manufacturers... yeah a real PITA, many still do the practice to this day. I’m not familiar with the AR100.2 but if it uses the Toshiba 2SA1302 and 2SC3281 then that’s a misinformation as the OP says it’s mosfets, these Toshiba devices are not Mosfets

 

[Addicted to music]

Seems you can still get them:


5 Pairs of 2SA1303 + 2SC3284 Audio Power Transistor https://www.amazon.com/dp/B00LQSXOP8/

If I were to purchase semis or replacements I would use a trusted supplier such as Digikey E14 or RS I would never never use eBay or amazon....

 

[Darkweb]

According to this interview, Dan D'Agostino was admitedly very measurement centric during his final Krell years. Later with his own brand he relaxed on metrics and focused on subjective.

Great interview, thanks for that. Cool to see how Dan’s thinking evolved.

 

[restorer-john]

(Toshiba 2SA1303 and 2SC3281)

The 2SA-1303 is a Sanken MT100 (comp 2SC-3284) device and the 2SC-3281 is a Toshiba (2-21F1A- the big flat one that isn't a TO3P that I can never remember the package name because it's a stupid number) (comp 2SA-1302). Sanken snuck a numbered complimentary pair in a Toshiba run of numbers (EIAJ) to confuse everybody who thought the next number up was a higher voltage part...

BTW, all those devices are still available in quantity (various grades up to Y) from my trusted supplier in Japan who has been supplying the OEM Japanese industry for nearly 40 years. If you ever need any so-called "unobtainium" original (not fake) semis, let me know...

 

[DonH56]

I don't think it uses (used) MOSFET outputs; they were in the lower-level stages. From the ARC database ( https://www.arcdb.ws/100.2/100.2.html ):

The Model 100.2 stereo power amplifier sets new standards in musicality and value against all other competitors at or near its price. Rated conservatively at 100 WPC into 8 ohms, the 100.2 produces 200 WPC into 4 ohms and a hefty 360 WPC into 2 ohms -- giving it sufficient drive to handle a wide variety of today’s leading loudspeakers. Both single-ended and balanced inputs are provided for easy connection to any high-quality preamplifier. A.C. power is supplied via a three-pin IEC connector and 14-gauge detachable power cord.​

​

Inside, the 100.2 reflects the latest design strategies from Audio Research engineers. High-impedance (i.e. easy to drive) JFET input transistors feed a single gain stage (for more sonic purity) using selected oversized MOSFET devices in a complementary configuration providing 27.5 dB of gain with tremendous dynamic headroom. This single gain stage is isolated from the power supply via Audio Research's constant current source technology, and directly feeds the MET (Multiple Emitter Transistor) output stage, which is also direct-coupled. A unique (patent pending) limited feedback loop of 6 dB preserves phase linearity, excellent damping characteristics and bass impact and control. The 100.2 is remarkable not only for its tube-like lucidity, but also for its ability to maintain this level of musical composure under demanding dynamic conditions.​

​

The 100.2 also features a fully transparent protection circuit that mutes the output stage whenever the amplifier senses D.C. offset at the input or a shorted connection at the output -- providing an extra measure of protection to loudspeakers. Generous heat sinking provides sufficient dissipation to allow in-cabinet installation, even while the amplifier's overall dimensions are reasonably compact.​

​

The new sonic standard set by the 100.2 must be experienced directly to be appreciated. It possesses a level of midrange transparency until now only achievable by vacuum-tube amplifiers, and delineates a dynamic soundstage that makes conventional solid-state amplifiers seem wooden and diffuse by comparison. Bass speed and harmonic richness are compelling, in large measure because the 100.2 is so liquid and grain-free in its signature. The 100.2 genuinely pushes performance beyond traditional solid-state expectations into new musical territory.​

Edit: But John would know for sure...