This review page is supported in part by the sponsor whose ad is displayed above
Final findings
Upon their return from the Milpitas day spa, the monos now sported different connectors but misbehaved again. While the relay would click and the power LED light up, they were stuck in what amounted to auto mute - no signal. As Jason explained, the capacitor across pin 1 and 4 in position U13 of the new overheat protection circuit was a 2.2uF value in my amps. This was discovered to be too small. It has since been changed for a 10uF cap as standard production issue. How had one of my loaners given up its ghost in the first place? "The bad amp suffered a burn-out at the output choke. We have added an overheating protection circuit on every unit now." This begs a question. How can an amplifier, under zero signal just sitting there with a load attached, overheat - especially one that's supposedly ultra-efficient? Why would it make any difference whether the source was turned on or not? Does it have to do with shoot-thru currents generated by the overlap of output switches? Insufficient heat sinking? Drifting bias current going into positive feedback? Frequency instability to cause oscillation? As a self-oscillating design with a 1MHz control loop and 500kHz switching frequency, the amp is "running violently all the time [Jason's emphasis]". NuForce suspects a high-frequency pulse or square wave generated during turn-off but couldn't replicate the event in-house.
So back the amps went to California to get fitted with the 10uF cap. While under surgery, they received the latest upgrade involving the removal of a zener diode and one opamp replacement. "We think this offers a slight improvement in sound. We debated about whether to hold back on our discovery. We already have one of the best amps on the market so why bother to upgrade it now? But that's against our principle. One interesting side effect of our circuit board upgrade policy is that used NuForce amps have high resale value because they will never be obsolete!" Okay, time out. For as new as these amps are to the market, there sure has been quite the slew of mods,
improvements and fixes. Somehow I'm getting suspicious that something fundamental about the design still isn't finalized, hence the nearly frantic implementation of changes. But perhaps I'm overreacting because of our site's track record with the NuForce amps.


Commenting on the moons' freak streak of malfunctioning loaners, Jason Lim assured me that "we have more than 600 units of Ref 8s and 9s in more than 20 countries and there have been failures. The failure rate is about 6% including problems during the initial launch of products from March to June. What are these failures and how do we plan to address them to assure customers that they will be able to enjoy the products for years to come?


I agree with your statement that we're still early in our development and that some problems were discovered post launch. We didn't realize the no-load overheating condition until we launched. That's what caused failures for Edgar. In the rush to ship products, some circuit boards have been through minor rework at the factory to create temporary quality issues. There are 3 known cases of "bad" preamps that suffered high DC offset (in one case, the preamp manufacturer acknowledged this to the customer) and this DC killed our NuForce amp. Lately, we discovered twin incidents where if the source device connected to the amp was turned off, the amp overheated and failed. This is rare but we have since added a cautionary note in the instruction manual. Sure, there could be additional issues we haven't discovered yet. That's a fact of life for any new company and new product. We don't make excuses for that but back our products with a 3-year parts and labor warranty so customers are assured if there are any problem with the amps, they will be isolated and fixed. We also back that up with our friendly and timely service. In most cases, we have shipped replacement amps before we even investigated the problems of the original units. Those customers are very happy with the way we handled things."

Jason's candor and the company's commitment to customer service are highly commendable. As a reviewer who has worked in manufacturing for years, I'm sensitive to the realities of product assembly, parts vendors issues and gremlins in first production runs. I don't mean to be unrealistic or hard on these guys. After all, the 3 pairs sent to The Absolute Sound reportedly worked so flawlessly as to net an Editor's Choice Award 2005. Neither the StereoTimes nor Positive Feedback loaners experienced any mishaps according to the new force. For some reason, their planet and our moons weren't aligned. I started to feel bad for Jason Lim who throughout it all was a stand-up guy who made no excuses but simply took care of business. Simultaneously, our readers expect open-book reporting so view all this in the appropriate spirit of even-handedness. Time for the modified Reference 9s. Check out the new connectors. The trend for glitz and jewelry in audio brought with it a reliance on massive connectors that looked appropriately butch to seemingly belong on the tail end of a 200 lbs monster amp or $100,000/pr speaker. It took an Aussie to cry foul, condemn this trend as bad-sounding foolhardiness and introduce the now legendary Eichmann connectors. Their claim to faim? Good riddance to mass and lots of metal. Less is more became the new battle cry. So successful were the sonic reports that WBT was essentially forced to revisit their approach and author the NextGen family of low-mass connectors.
As cable guys know, less dielectric is always better than more. Thin (especially for interconnects) is often far superior to thick. Time for the male-dominated obsession with size and weight to make way for a more enlightened perspective also on connectors. Kudos to NuForce for bucking a silly trend by focusing instead on what really matters - the sound.


In the area of crystalline transparency, the NuForce amps make quite the statement. The soundstage is wide open and visibility extends into all its nooks and crannies even under heavy assault of complex interleaving strands
and massed forces. If the Reference 9 was an orator doing an impromptu reading of a never-seen-before speech, he wouldn't get confused. He'd not trip when the language got dense, verbose, complicated and loaded up with rare words of questionable pronunciation. Bring on the heavy artillery. The NuForce monos keep sorting things out unruffled regardless of who wrote the speech - er, score.


Needless to say, they set up camp on the dry side of the river. They expectedly lack not only the surface texture of tubes but also the extreme degree of image density Peter Daniel's chip amps conjure up. However, the Californians are fuller and more fleshed out than the Bel Canto Tripath eVos were (now superseded by the ICEpower e.Ones). Not only that, they're also more articulate and manifest.This suggests a bit of lackluster restraint and bleaching with the Tripath amps. Compared to this particular Class D competition, the NuForce circuit is already a clear advance. Compared to a truly superior linear transistor amp like my Patek SEs? The Ref 9s come in second in two other areas. Something about their treble is best called subliminally damped when compared to their sparkly demeanor elsewhere in the audible range. It's not hooded nor is it opaque. Yet simultaneously, the terms suave and elegant don't suggest themselves either. Two, these amps foreshorten decay trails compared to the Pateks. This incurs a kind of delayed clipping - not by cutting off the tops of waveforms but by curtailing for how long they flutter on the air before they fade to nothing.


Bass grip, tautness, control and sheer gumption are exceptional as is the lack of grit and glare. Ditto for dynamic acuteness, the sense that the pulse of the music is fully exposed. There's not just the big cardioid one of the main groove but also the subtler pulses buried deeper in the music (like the kind Japanese acupuncturists will read to diagnose the balance or imbalance of internal organs). Images in the soundstage and rhythmic lines alike are thus extremely well articulated, all the while avoiding chiseling, hyping or etching.


In my system and as evidenced by their first-rate showing on the 101dB Zu Definitions, these amps were also dead-quiet in operation. Some on-line reports have mentioned a hissing during no-signal interludes. Kevin Haskins of DIYCable (now Hypex central as well) explained that "I used to have the same problem with the LC Audio amps [that's another Class D topology - Ed.]. It's caused by a difference in the oscillation frequency on the two channels. It's intermittent between modules. Sometimes I wouldn't have a problem and other times it would raise its ugly head."
In conclusion
My NuForce excursion was quite the journey. From the on-line forum buzz and preceding reviews to the over-the-top hype in the firm's advertising copy; from repeat failures in Arroyo Seco and Sydney to actual performance - I've run the gamut from mistrust to admiration, from slight restraint to real-world enthusiasm. On the restraint side of the ledger, there's the strange fart slash paper-tear sound upon turn-on. It's a tacit reminder of the oscillating "digital soul" of these amps. Just sitting there powered up, they generate a constant amount of violent ultrasonic noise. True, that may not be audible to human ears. Still, it's radiation pollution pure and simple - yet another small box that generates high-frequency switching garbage. The reliance on an SMPS rather than conventional power supply -- despite the big-talk claims of superiority -- clashes with the confession of needed overheat protection. Without it, the power supply can take out an output coil under zero signal. Something about that entire scenario reeks questionable. I never had the slightest issue with the eVo4 Tripath amp from Bel Canto which uses a conventional massive power transformer. And I've run it for years.


On the other hand and disregarding the still emerging status of NuForce -- and not having heard Hypex or ICEpower amps in direct comparison -- something's plain as day: these amplifiers patently eclipse the prior eVo amps of comparable power. Considering how highly regarded those were, that in itself is a major compliment. It's also proof of ongoing progress in this field. If I seem a bit reluctant yet to fully join this particular camp, it's because the old-fashioned, nearly primitive by comparison chip amps in-house still do certain things better. That's for essentially the same coin. And the linear amps aren't outclassed in any aspect I can hear (but then I don't have the sort of humdinger loads for which the Ref 9s were reputedly optimized).


Now add to this colorful mix the fact that simply repackaging an ICEpower module does nothing to further the art. The gents in Denmark already did that - at Bang & Olufsen and the University of Stockholm. The team at NuForce, on the other hand, isn't rebadging inventions. They're inventors themselves, with all the risk and glory, heart ache and sputtering advances that entails. In fact, they're already talking about a tube-buffered NuForce amp - though still only conceptually, with nothing outright planned except consideration of the idea. In short, these folks have squarely positioned themselves as advancing agents for the audio arts, doing the pioneering work in-house. Entrepreneurs in the US when everyone farms out in the Far East? Bravo!


Finally, for the cost, output power, styling and size on hand, the NuForce Reference 9s open the doors for more budget-conscious music lovers to partake in high-resolution High Fidelity sound. We must always remember that only the most intensely committed would even enter a soirée with these monos intent on picking them apart. Everyone else will pretty much fall all over themselves over what they offer. So, no Second Coming or Rewriting The Rules yet when judged from the jaded ivory tower seat; but a major thumbs-up from the real world where everyone else cavorts and frolicks.
Manufacturer's website