[James Romeyn]

My Ncore and all new cables are pretty thoroughly burned in now after about a week.  Speakers are 88.0 dB actual sensitivity.  Room is 3200cf, w-w carpet, light treatments @ usual suspect points.       

This is a unique comparison because in both cases the loads are identical except for impedance load and about 2' of wire.  A potential down side is I performed the test in mono only, but results would only magnify in stereo.   

The mono source was not simple summed mono L+R but rather the Center Channel output of a pure analog Trinaural processor.  (I'll test the Trinaural later for transparency to the source.  The Trinaural should have a simple "stereo/bypass" selector but it does not, so I must improvise one.)

HERE'S the speaker load, my own unique evolution of Toole and LeJeune's earlier work.  If you decide to perform this test yourself with two stand mount monitors setup in the described array, employ either the L ch or R ch alone for mono source, don't sum the L+R channels together because of the resulting phase anomalies.     

If you assemble two stand mount monitors as per the above instructions, toward the goal of replicating the test, 8 Ohm nominal monitors is ideal.  Two 8 Ohm nominal monitors will switch from 4 Ohm nominal (parallel) to 16 Ohm nominal (series).  The title specs are as indicated, the minimum impedance, far more critical to results rather than nominal impedance. 

The 4 Ohm load does indeed draw 4x as much current, and your cables must meet this spec, which mine easily meet.  The sum total cable difference was about 2' extra cable in the 4 Ohm load.   

Note: compensate for the amp playing +6 dB louder in 4 Ohms vs. 16 Ohms.  My preamp has 1.0 dB steps in the analog mode employed.   

Two 4 Ohm monitors setup as directed would switch between 2 Ohm and 8 Ohm.  In the 2 Ohm mode I suspect the minimum impedance will be too low. You might risk Ncore damage, I don't know.  If you do the test with two parallel 4 Ohm nominal speakers, do not exceed moderate levels.  Too low SPL may skew test results.  In my case I played a well recorded Telarc orchestral piece at good levels. 

Sorry, I kept the subs on, active crossed at 70 Hz.  Playing full range would only increase the differences though, certainly it would not decrease the differences.     

Sorry for such a long drum roll...Well, I'm hear to tell you, the 10.6 Ohm minimum load absolutely annihilated the 2.65 Ohm load.  The orchestra went from drumming along like the leader took a break, to having drank a couple large cups of my morning brew, not the strongest on earth, but it's up there.  Stage about doubled in size (yes, this array even in mono makes a huge stage, believe me), greater instrument separation, detail, smoothness, timing, it all improved and not a wee bit.  Dynamics were completely different.  Much more "interesting" and alluring, yet if anything, less fatigue because it was just so much more interesting to hear.  Really, they sounded like they played flat and with different instruments in 2.65 Ohms.  The biggest difference was likely "sound density", a quality in which this speaker array may be supreme.  The low impedance load by comparison was "thin" in density, not spectral balance, which was virtually identical.  The high impedance load, by comparison, sounded like the effect of switching to an absolutely killer big, particular anatomical part big, ultra transparent tube amp, in the very best sense of the word.  Highly dense, but still transparent (even more transparent than the low impedance).         

Sorry, I don't know about my speaker's phase angle, but suspect it's not too difficult.  IIRC every test amp drove them well, especially Ralph Karston's sweet 30W stereo OTL.       

With the 80 Hz active high-pass filter, this speaker has only one minimum impedance dip centered at 200 Hz, stretching from about 100-600 Hz.  Above 600 Hz the load is flat nominal.

I'm not saying Ncore is generally ill-suited to lower impedance loads because, well, every application is unique as anyone with experience can and will tell you.  This was one test. 

But I'll never bother again with the 2.65 Ohm minimum load except to prove the point to a visitor.

Powering the 2.65 Ohm minimum load Ncore runs out of current before voltage (specs indicate about 500Wrms). 

Powering the 10.6 Ohm minimum load Ncore runs out of voltage before current and always has an easier load hence sweeter distortion curves (I presume).  Ncore makes a maximum of 100W above 600 Hz with a slight power reserve (about 2 dB, 160W) at the impedance dip of 10.6 Ohms centered @ 200 Hz.

I wonder if there is any difference between peak power into the two different loads.  I'd bet nothing or insignificant.  But there's probably big differences between the peak and average power capacity: Into 2.65 Ohms the peak and average power are likely very close.  Into 10.6 Ohms I bet there is a big  difference between peak and average power.  Music is dynamic and this might explain why 10.6 Ohm sounds louder even though it's not.  I seem to remember months ago Mike Galusha posting that, into 4 Ohms, Ncore's peak and continuous power capacity are pretty much the same.

       

[poseidonsvoice]

James,

Your link doesn't work.

Best,
Anand.

[James Romeyn]

Thank you sir, I think it's fixed now!

[James Romeyn]

Thanks to member Russell Dawkins who passed along Bruno's Hypex advice: Hypex output impedance equals series resistance of 3' of 4 AWG, which my calculator specifies is .0015 (fifteen ten-thousandths) Ohm.

Math employing these two calculators (http://www.bcae1.com/images/swfs/speakerwireselectorassistant.swf and http://home.hiwaay.net/~rgs/awgcalculator.html) showed Ncore needs about 3x lower series resistance than what I had employed to drive 2.65 Ohm.

To experiment on the center channel only (Trinaural system) I triple-paralleled 4' of wire for series resistance of about 5 AWG, then re-compared 2.65 Ohm load vs. 10.6 Ohm.  When I'm wrong, I'm really wrong.  Pure sound quality even at quiet playback levels is better driving the lower impedance now, exactly the opposite of before.  With peak voltage quadrupled, maximum level is not quite up to AudioKinesis, but close enough for my use. 

I don't need that much wire to get the L and R ch up to correct spec.  But it would be much easier to terminate single conductor 4 AWG vs. tripling the dual-12 AWG THHN wire (stiff and difficult to strip, Stan Warren recipe) as per the center channel.  I considered employing single-conductor 4 AWG power cable for speaker cable.

Before buying 4 AWG I paralleled ten x 14 AWG generic zip cord speaker cable to equal about 4 AWG.  Strangely, A-Bing this cable vs. tripled Stan Warren recipe, the former was smoother, easier to listen to, and more inviting throughout. 

I just ordered 10' of 4 AWG single conductor power cable (multi-strand estimate 30 AWG) to experiment.

If you or anyone you know has negative Ncore opinion, use the links above with confirmed specification of the speaker's minimum load impedance (the nominal figure is useless because maximum current draws at the minimum impedance, not the nominal figure).  I'd not be surprised if all or most Ncore setups would audibly benefit with lower series resistance in the speaker cable. 

I know mine did.

For 4' long speaker cables, how many amps need less series resistance than that of dual 12 AWG (equivalent to 9 AWG)?     
       

[cab]

Glad to see you got that sorted out. The ncore again shows that it is not your typical amp.

[jonbee]

Hmm... interesting stuff. My speakers are 4 ohm minimum, 5 nominal. My 8 ft. cables are 7 ga., so I'm probably ok. It sounds great, in any case, but I'll be cautious about going to smaller cables in future.

[James Romeyn]

Hmm... interesting stuff. My speakers are 4 ohm minimum, 5 nominal. My 8 ft. cables are 7 ga., so I'm probably ok. It sounds great, in any case, but I'll be cautious about going to smaller cables in future.

What make/model cable?  I'm curious because I used Stan Warren's DIY recipe for so many years I have no idea the gauge of modern speaker wires (past experience is that Audience cable sounds great but gauge is very thin hence likely inappropriate for Ncore).  The Ncore arc welder needs a lot of copper. 

Did you check series resistance of your cable in the calculator?  I suspect you might improve performance with thicker than 7 AWG for 8' length. 

[Regnad]

I am using Triode Wire Labs 7+ cables, is there a way I can measure the "paralleled" gauge total?  Thanks.

[James Romeyn]

Use the calculators in the links I posted above today.  Current draw is greatest at the minimum impedance, so disregard the nominal figure.  2x _ AWG = 3 AWG lower ( _ AWG - 3).     

I can't stress how important speaker wire gauge (external, not the 2" internal lengths) appears to be with Ncore.  Frankly I'm surprised Bruno did not mention it in assembly instructions. 

For Ncore, right now, if I had to choose between speaker cable X, highly regarded, cost no object, equivalent to 12 AWG vs. welding cable all-copper 4 AWG, I'd choose the latter over the former. 

I now rate Ncore performance approximately equal to any amp I've heard, which include Aesthetix monos, about $16k/pr IIRC (heard powering Vandy 5 Carbon, yummy).

I highly suspect persons with negative Ncore reaction heard more the speaker wire resistance than the amp. 

[Regnad]

Amps right next to the speakers or active speaker systems would seem to be a good way to go.

[James Romeyn]

Yeah, my monos are directly behind the speakers, needing only about 4' wire to the top monitor, another 1.5' to the bottom ambiance/mode cancelling monitor (described here: http://jamesromeyn.com/home-audio-gear/ramca-romeyn-ambiancemode-cancelling-array/)

If I had ultra high end powered monitors such as the studio variety that go up to almost $100k/pr, I'd sure want to replace any non-Ncore power amps.

[cab]

I now rate Ncore performance approximately equal to any amp I've heard, which include Aesthetix monos, about $16k/pr IIRC (heard powering Vandy 5 Carbon, yummy).

I highly suspect persons with negative Ncore reaction heard more the speaker wire resistance than the amp.

But what about that "smear" we have been told about, or the lack of 3d imaging? Or the recessed highs?

[jonbee]

What make/model cable?

Acoustic Zen Satori biwire - 7 ga. total OCC copper. Very nice sounding- smooth, with great bass. Other than their huge size and low WAF I'm happy with them.

[jonbee]

But what about that "smear" we have been told about, or the lack of 3d imaging? Or the recessed highs?

On mine the highs are laid back, but astonishingly clean. The top end detail is definely not MIA at all. Smear and lack of 3d imaging are the opposite of what I (and my listening panel) heard. Best focus and nearly the best 3D I've ever heard. The only area in which the Ncore is not the best I've heard is in the presentation of deep ambient info- very, very good, but not quite as good as the very best I've heard.
In other systems, of course, this may not be true.
Matchups rule.

[*Scotty*]

Checking the NC 400 data sheet, the graph of the amplifiers output impedance shows it to be less than or equal to 700 microhms up approximately 2kHz and less than 1 milliohm up to 17kHz. This about half of what Jim has calculated to NCore output impedance to be at 0.0015 ohms. I am not sure how relevant this fact is to what Jim heard, but it is a change to the underlying assumption about how big a wire gauge might be necessary to drive load impedance loads.
Scotty

[Rclark]

I stumbled across "autoformers" when looking at Anticables.. this guy seems to say what James is saying, that increasing the impedance increases performance.

Is this something you would use James? Fellas?

http://www.zeroimpedance.com/Zero-New.html

http://www.zeroimpedance.com/faq.html

[*Scotty*]

Rclark, Jim changed the design and guage of the speaker wire he was using on the low impedance load he was driving and equaled the performance of the higher impedance speaker arrangement. The "autoformer" seems to be a solution looking for a problem with regards to the NCore amps. Zybar removed the autoformers from his system when he switched over to the NCore amps.
Scotty

[Rclark]

  Ahh, so Zybar must have tried autoformers on Ncores and not liked them. This guy's FAQ's on the autoformers makes them sound like something to try, but apparently it already has been done.

 What is the difference between the speaker guage approach and an autoformer?

 

[*Scotty*]

To my knowledge they were never considered necessary by Zybar as an interface between his Vandersteen 5As and the NCore 400s.
Jim went to a different design speaker wire of much larger gauge and was satisfied with the result.
Here is link to the Wiki on autoformers.
http://en.wikipedia.org/wiki/Autoformer
In the case of the Speltz products, they are used as an impedance matching device. The drawbacks are Hysteresis distortion at low signal levels and core saturation at low frequencies which causes an unwanted rise in mainly third-harmonic distortion in the bass frequencies and a roll-off in the bass depending on the design of the autoformer. Autoformers are subject to the same laws of physics as transformers and have similar limitations in audio applications.
Here is a link to how transformers distort which also applies to autoformers. This is a discussion line level transformers, when the power levels are ramped up to those necessary to drive a loudspeaker the effects are much more pronounced.
http://www.soundonsound.com/sos/feb10/articles/analoguewarmth.htm#22
If you want to fill in the blanks in your knowledge about audio technology including the digital domain this is a very worthwhile book to have on your reference shelf. Handbook for Sound Engineers, Third Edition
http://www.amazon.com/Handbook-Sound-Engineers-Third-Edition/dp/0240807588
Scotty

[Rclark]

 Excellent. Thanks for all that Scotty. I already know they sound great, but the born audiophile in me is already thinking of the next step, which in this case, looks to be unnecessary. I'll order that book too.