Audio Research Vs115

Jake,

I re-checked the net to view the bench tests for the Thiel CS2.3 speakers. What I saw in fact was bench test results for the CS2.4s, which as I said had an amazingly flat 4 ohm impedance plot from 50 or 75 Hz through the audible acoustic spectrum.

Do you know whether the CS2.3 was configured differently? I ask because you posted that the 2.3s had modulating impedance plots? Do you have any idea what the SF Cremonas impedance and phase angle plots look like? Couldn't seem to find any hard data on the net.

The reason I am so curious is because I'm trying to gather some real life anecdotal information about how our amps perform when driving speakers with goofy impedance and phase angle plots.

Btw, how do you even measure the output performance of the amp and speaker combo? I have a digital meter that I use to bias the amp's tubes. But I have no idea how to check actual output with the amp hooked up to the speakers. Or how to check output impedance (with no load) off the 4 and 8 ohm taps. Can you advise?? I kinda would like not to damage my gear just to satisfy my curiosity.

Thanks and have a great Memorial Day weekend.

Bruce

Bifwynne - The CS2.4 has a somewhat smoother impedance plot than the CS2.3. Here are the measurements done on the CS2.3 in a Stereophile review: http://www.stereophile.com/content/thiel-cs23-loudspeaker-measurements. You can see the impedance dips to 2 ohms in the lower midrange and goes through big swings in the bass region. The phase angle varies all over the place, too.

I've seen impedance plots for the SF Cremonas but I have the newer Cremona M and I'm not sure if the data would be similar. The Cremona M is supposed to have "improved" drivers and a different crossover network.

To measure frequency response into an actual speaker load, I use a Tektronix audio signal generator driving one of the auxiliary inputs of my ARC preamp and I attach a Tektronix oscilloscope across the speaker terminals. I bought both Tektronix units on eBay quite cheap. I dial back the signal generator output levels so I see about 1V at the speakers and everything is happy.

Measuring the output impedance of the VS115, or any other amp, is a bit more involved. The easiest way is to get an impedance analyzer that works at audio frequencies. The analyzer is connected directly across the speaker terminals of the amp and it injects a low level signal back into the amp to make its measurements. I have an RF impedance analyzer but it doesn't go down to audio, so I've never done this myself.

Hope this info is helpful.

Thanks Jake. I think the impedance/phased angle plots are demanding as reported in Stereophile mainly because of the 2 ohm dip and general 4-5 ohm range. But other than that, the impedance is quite flat between 4 and 5 ohms across the freq. band. The low bass peaks are typical. Nevertheless, it does demonstrate the VS-115's impressive performance capability. How many amps out there can push power into such low impedance loads??

Bifwynne - I should mention that a network analyzer (sometimes called a "vector network analyzer, or VNA) can also be used to measure the frequency response of an amp into any load. The same instrument will measure the output impedance of the amp as well. My home lab includes a VNA for RF but unfortunately it does not go down to audio frequencies.

Out of curiosity, what speakers are you using with your VS115?

Paradigm Sig. 8 v3. I've had a devil of a time getting reliable and timely info about the specs, but finally got through. The S8s have a 21-22 max ohm peak between 1K and 2K Hz; and a 4 ohm min. in the bass region. I've been in private contact with one of our tech members and he did some math calcs that indicate the VS-115 should be able to drive the S8s with a relatively flat output response because the amp's output impedance is low for a tube amp. Soundstage and Stereophile bench tested the VS-115 and Ref 150 and they share similar electrical attributes re: NF (about 12-13), DF (about 7 or 8) and output impedance (approx. 1.1 ohms off 8 ohm tap; about half that off the 4 ohm tap). Based on his calcs, FR may be off +/- .5 db at the peak, which ain't so bad.