This review page is supported in part by the sponsor whose ad is displayed above

Reviewer: Srajan Ebaen
Source: Zanden Audio Model 2000P/5000S; Ancient Audio Lektor Prime
Preamp/Integrated: Wyetech Labs Jade; ModWright LS-36.5 [on review];

Amps: Yamamoto A-08S; Fi 2A3 monos; Melody I2A3; Almarro A318B [on review]; Coda CSX [on review]; Trafomatic Audio The Experience One [on review]; JAS Audio Bravo 2.3 [on review]
Speakers: WLM Grand Viola Signature Monitor; Mark & Daniel Ruby; Mark & Daniel Maximus Monitor [on loan]; DeVore Fidelity Nines

Cables: Crystal Cable Ultra loom; Zanden Audio proprietary I²S cable; Crystal Cable Reference power cords; double cryo'd Acrolink with Furutech UK plug between wall and transformer
Stands: 2 x Grand Prix Audio Monaco Modular 4-tier
Powerline conditioning: 2 x Walker Audio Velocitor S fed from custom 117V AudioSector 1.5KV Plitron step-down transformer with balanced power output option; Furutech RTP-6 on 230V line
Sundry accessories: GPA Formula Carbon/Kevlar shelf for transport; GPA Apex footers underneath stand, DAC and amp; Walker Audio Extreme SST on all connections; Walker Audio Vivid CD cleaner; Walker Audio Reference HDLs; Furutech RD-2 CD demagnetizer; Nanotech Nespa Pro; Acoustic System Resonators
Room size: 16' w x 21' d x 9' h in short-wall setup, with openly adjoining 15' x 35' living room

Review Component Retail: $950/pr in Cherry or Maple; $450/pr nude; $25 for custom jumpers

From zero to hero in 16 ohms.
In essence, that's what the impedance transformers from AntiCables' Paul Speltz promise. That's especially relevant to OTL and low-power SET fanciers whose amps suffer high output impedances that interface poorly with low-impedance speakers. The Zeros are available as raw toroids with spade-terminated secondaries or encased in fancy Maple or Cherry boxes terminated in pure copper binding posts. They offer a 2 x, 3 x and 4 x impedance multiplying function to alter how your amplifier views your speakers. If your speaker is a nominal 4-ohm load with dips to 1.3 ohms, the Zero, depending on how its output taps connect to your speaker inputs, can turn your challenging speaker into a nominal 8, 12 or even 16-ohm load. That raises its undesirable minimum impedance value to either 2.6, 3.9 or 5.2 in our example, making most amplifiers, shy of the expensive load-invariant brutes who don't give a twit, happier campers.

It's obvious that this math works across the board. If your speaker exhibits wild impedance swings -- say from 3.8 to 30 ohms like an electrostat might -- the multiplier action widens the entire swing window. In a x 4 scenario, our example would create a crazy 110-ohm differential. Depending on what load impedance is encountered at what frequency, amplifiers whose high output impedances modulate their frequency response will naturally introduce off-the charts linearity aberrations with such exploded swings. But then, SETs with high output impedance are the most unlikely of all candidates to ever be leashed to electrostats and other speakers that exhibit such deviant behavior in the first place. It's a safe generalization then to say that most amplifiers will prefer a higher load impedance. The Speltz devices are an intuitive, elegant and simple way to accomplish just that.

The operative word is mismatch. The Zeros are not recommended or necessary without a mismatch. They're a cure for a particular ailment, not a maintenance vitamin for healthy amp/speaker marriages. A nominal 8-ohm speaker transformed into a 32-ohm load is not an automatic guarantee for mo betta sound. To a transistor amp which increases its output power into lower impedances (even if not doubling like the brutes), such a transformation could create rather than remedy a mismatch. And while more advantageous impedance matching should primarily benefit bass performance, improvements might materialize elsewhere just the same. Consumer audio has no impedance matching standards to create predictive diagnostics. Experimentation thus remains the sole solution.

The gestation for this assignment was the publication of my Mark & Daniel Maximus Monitor review, in particular the added postscript. There M&D retailer Mike Garner of TweakGeek had reported on "Taming the Maximus with Autoformers". As an 85dB current-hungry 4-ohm load with low-impedance dips, the Maximus certainly challenges less than macho amps in the drive department. Garner had managed to drive the Maximus with a completely counter-intuitive T-amp once the Zeros had leveled the impedance playing field. Spotting this postscript, Paul Speltz wrote in with thanks and a question. Did I want to try these devices for myself? As a believer in personal experience, I naturally accepted and a pair of Maple-encased Zeros hopped on the next Canadian courier flight.

Impedance multiplying also works when bridging amps for higher power. Bridging always compromises impedance stability. For example, my 50wpc AudioSector Patek SEs bridged to 100-watt mono run ferociously hot when strapped to the 4-ohm Maximus. After all, bridging turns the Maximus into a 2-ohm load for the amps and the speakers' low-impedance dips plummet precariously below half an ohm. Applying a 4 x impedance multiplier could be just the ticket. The bridged amps would see a standard 8-ohm load to cruise rather than bruise.
Complete no-brainer applications and in fact often mandatory for best results are OTL amplifiers. Their power increases into higher and reduces into lower impedances, i.e. just the inverse of typical transistor amps. But even lower-power SETs should benefit whenever the issue isn't achievable SPLs but control over sub-optimal loads.

A theoretical issue is the resolution impact from the insertion and connections of an additional and passive signal path device. One operative term for transformers becomes bandwidth. The Zeros are single-winding multi-tapped autoformers where the input enters on a different tap than the outputs exits. As a low-ratio transformer, the Zeros aren't concerned with DC saturation nor complex winding geometries for high-impedance transformation such as SET output transformers must accomplish. DC resistance is a low 0.3 ohms, claimed bandwidth 2Hz to 2MHz. This should make for an essentially transparent device though high amplifier DC offset voltages could mandate a fuse addition (check Paul Speltz's White Paper for details).

Should the Zeros be guilty of a minor veiling effect as something extra in the signal path, a much improved power transfer from amp to speaker should overrule any such mostly theoretical losses. 10 steps forward, one heel back.

One very real benefit is increasing the operative damping factor, i.e. the ratio of amplifier output and speaker input impedance. If you can't lower the former, raise the latter. The outcome is the same. That's a definite boon for NFB triode amps with high output impedances and no bass damping. Paul Speltz proposes a second factor: "I suspect it's simply easier for an amplifier and speaker wire to generate and transfer voltage than current. One watt into 4 ohms requires 2 Volts at ½ amp of current. One watt into 16 ohms requires 4 volts at ¼ amp of current. That's twice the voltage and half the current for the same amount of power. Running high voltage/low current lines are a necessity for both the power utilities (high-power lines) and commercial audio installers (70v paging system) where high voltage/low current is generated, distributed and then transformed to lower voltage/higher current at the destination. The advantages of running low-current lines are well understood. It makes sense to apply them to our hobby as well."

In use, you'll have to decide whether to install the Zeros close to the amp or speakers. The necessary leads (standard speaker cable lengths, two shorter jumper pairs) will be the same either way. I decided on keeping things centralized around the amps to be easier to photograph. To maintain proper polarity, divide the Zero box into two equal halves across its length in your mind. On the output side, you'll have a single post above the dividing line, on the input side, a pair. The polarity of all posts above the line is the same, all the ones below the line are the opposite. Which is which is determined by the +/- markers on the single output terminal. In the above photo for example, the left Zero box has the hot posts away from the viewer, the return posts towards the viewer - one each on the left side, two each on the right. Depending on which possible combination of hot and return posts you select on the amplifier end, impedance multiplies by 1.4, 2, 3 and 4. I used Crystal Cable Reference locking banana jumpers for convenience.

With a transistor amplifier like Red Wine Audio's Signature 30.2, quadrupling impedance for the most extreme mismatch naturally attenuated output voltages. This amplifier delivers less power into 32 ohms than 8. Using John DeVore's benign Nines additionally demonstrated diminished bass amplitude and transient definition even after level compensation. Everything softened up as though overall drive was compromised. x 4 was not on the ticket. On easy material, any potential benefits offered by x 2 = 16-ohm speaker impedance were too subtle for reliable detection. Once hard-hitting bass-heavy fare entered, the 16-ohm connection did offer somewhat better overall separation and slightly improved bass transients. However, I felt those changes were not profound enough to make the added expense and box count terribly attractive. The DeVore Nine, it seems, really aren't demanding enough to have a quality transistor amp like the Signature 30.2 benefit from load impedance alterations.

The 18wpc JAS Audio Bravo 2.3 6C33C SET with user-adjustable negative feedback and 8/4-ohm tabs didn't like the x 4 factor on the DeVores either. Yet x 2 for a nominal 16 ohms was clearly beneficial and an incontestable improvement over the direct speaker connection once things got complex, dense and busy in the bass. Ditto for the 4-watt zero NFB Trafomatic Audio The Experience One SET with JJ 2A3-40s. 16 ohms was preferred MO. While this micro-power champ hovered at the SPL edge without gain boost from an active preamp (it's an amp with a passive pot), the now wholesale improvements in overall grippiness and definition definitely suggested that 16 was once again the winning number, albeit by a rather more significant margin here than with the Chinese 6C33C amp.

Time to tighten the screws. Out with the friendly DeVores, in with the burly Mark & Daniel Maximus Monitors, nominal 4-ohm loads at 85dB. With the Trafomatic holding the defensible ground, backup was drafted by way of Coda's 0.5X active preamp set to max 18dB gain to guarantee satisfactory SPLs. No problems with Dulce Pontes nor the added load of the OmniHarmonizer. And all that with at least 10dB of headroom to spare on the Coda at standard levels.

However, Mercan Dede's synth trickery in the bowels of the bass pit where normal acoustic instruments dare not tread caused predictable compression distortion. Nobody in their right mind would mate a $1,700 4-watt SET to the Maximus Monitors no matter what band aids the after market offers. It was simply surprising to observe how far this puny amp could be pushed before pleading 'uncle'. Protest arose far earlier without the Zeros but even with them, this silly setup was flawed conceptually and beyond reason.

No serious 'phile should advocate to match the 18-watt Bravo to these speakers either. Nor the 30-watt Melody I2A3. Nor necessarily even the 30-watt Red Wine Audio Signature 30.2. Perhaps the 50-watt AudioSector Patek SE even though the speaker manufacturer himself could still consider it borderline. How would the Zeros influence common sense? With soap and water. Off the Bravo's 8-ohm tab, speaker impedance set to 16 ohms, their unbelievable F3 of 36Hz showed its mettle even on this apparently hopelessly underendowed tube amp. Hats off to the mighty 6C33C bottles. Even though far from sensible a match, it nonetheless sounded impressive. And clearly not as a mere freak show spectacle. In fact, it sounded better than the nearly twice the power Melody integrated. The latter's 2A3s were less thrilled with the low-frequency abuse I subjected them to on this load regardless of impedance shifting.

The battery-powered Tripath amp demonstrated the superiority of transistors with increased weight, mass and leading-edge definition in the two lowest octaves. Though it evinced zero issues without the Zeros, it did fancy these speakers as 12-ohm rather than 4-ohm loads. That setting gave the most buxom bass coupled to the best separation, overall sweetness and soundstage depth. My AudioSector 50-watt chip amp mirrored that verdict. Its improvements were a bit smaller than with the Signature 30.2 but still appreciable over running zero-less.

Incidentally, the Zeros can also be reversed, to lower load impedance. Why would you want to do that? Decware amps for example are deliberately optimized into lower impedances. VTL amps assume a 5-ohm impedance. I turned the DeVore Nines into 4 ohms on the Coda monos just for the exercise, without netting any audible justifications for doing so.
Needless to say, all these observations are far from conclusive. Happy averaging of my results merely suggests that most amplifiers not designed as load-invariant brutes, especially tube amps, do prefer a 12 to 16-ohm load over 4 ohms. While the magnitude of improvements differs, the areas of improvements overlap: better bass impact, definition and control; better separation during dense passages; a more dimensionally developed, deeper soundstage; and better overall articulation and fuller tone colors.

Sadly, 16-ohm speakers are rare - more than seems justified. That's where the Zeros come in. Any speaker can now be 16 ohms. If you match your speakers and amplifiers appropriately in the first place of course, these devices should be mostly gilding the lily rather than weigh in as heavy must haves. To press the point, some might even go as far as claiming that anyone who outright needs the Zeros already screwed up. In that vein, none of my experiments above would be considered properly matched. The Zeros simply minimized the penalties. They thus helped accommodate mistakes, a perhaps questionable approach. The 450-watt Coda CX monos for example clearly did not need or want the Zeros' help. Aficionados of micro-power direct-heated triodes meanwhile should know better than forcing a wedding with unsuitable speakers. There are 16-ohm Feastrex, Lowther and PHY-HP drivers after all, 12-ohm Zus and others.

Ultimately, speaker manufacturers should install impedance-matching autoformers in their crossovers if design parameters enforce the use of 4-ohm drivers and even lower impedance sags. Such idealism of course is not for sale yet. Paul Speltz originally envisioned the Zeros as a solution for OTL amps. Customer feedback and personal experiments subsequently suggested that the Zeros can also benefit other amp designs. I cannot disagree. If I had to mate the JAS Audio Bravo 2.3 to the Mark & Daniels for example, I'd do it with the Zeros, hands down, no reasons given why I pursued such a peculiar match to begin with. If I sound less than 100% committed, it's because proper matching of speakers and amplifiers should be done directly, no mediating devices necessary. That it often isn't says as much about poor audiophile habits as it does about modern speaker manufacture which cares little about how hard it makes amplifiers work because "power is cheap". None of these various arguments and counter comments take away from the fact that the Zeros work exactly as advertised. Whether you should like a pair depends squarely on how well you've managed your amplifier/speaker interface and whether you've obeyed good sense and sensibility. With many tube amps being less than sensible where optimal speaker control is concerned, the Zeros can level the playing field in your favor...
Manufacturer's website