Who Makes the Best Sounding Digital Amplifiers?

[ddude003]

I wonder what the impedance curve looks like for the woofer compared to the ESL panel... I think by vertical bi-amping with your Crowns your getting 1/2 of the ML ESL hybrid right... Makes me think about using Zeros or Tubes for the panels part or some other Power Paradigm amp...

 

[AudiocRaver]

Not a huge tube fan and definitely not on a ML panel, unless distortion is SUPER low, and then... why bother with tubes? Class AB solid state, sure, but I was partly going for high power, low weight. The Crown's quality is surprisingly high on clarity.

 

[dlaloum]

I find that, when an amp can easily drive a speaker, you can compare it to any other amp that can easily drive the speaker, and they all sound alike!

Where they start to sound different (assuming that they are basically good designs to start off with obviously) - is where the amps are used to drive speakers that are at the edges of, or outside of, an amps performance envelope....

Then different amps will misbehave differently - and they will sound different.

The Crown XLS's will drive almost anything well... their performance envelope is extremely wide... most of the other series within the Crown range, don't have this ability (check out the 2 ohm specs on the different models.... the XLS's are quite different to most of the other series).

I noted that "clarity" too on my Gallo speakers - and I suspect that it is due to the absence of distortion artifacts caused by the amp trying to drive a 1.6 ohm load.... ie: the Crown is perfectly happy driving that load where the Integra AVR was absolutely not happy about it. (sounded veiled... muddy... not obviously distorting / clipping, but not sounding great)

Lots of speakers with exotic drivers (ESL's the CDT tweeters on the Gallo's) - have these issues.

Having said all that - the Crowns are not champions of Signal/Noise ratio's / SINAD - and if not properly gain matched with the preamp... most especially with high sensitivity / efficiency speakers... quite a few people have reported Hiss... (at low but apparently audible levels).

I have not experienced this on my 88db SPL/wm Gallo's

On the topic of Tubes / Distortion - ESL's (and the Gallo's) achieve part of their magic, by being typically an order of magnitude better in terms of distortion, than standard "dynamic" speakers. Yes there are exceptions in the realm of dynamic speakers... but as a genre - the "standard" speaker typically has 10x the distortion of an ESL.

This is one of the reasons for their "transparency".... - and they will typically show up whatever is being fed into them - for better or worse.

If your tube amp, is providing lush, romantic, 2nd Harmonic heavy, tunes - then that is what the speaker will reproduce.

Nothing wrong with that preference.

My personal preference is for a Neutral system - what comes off the recording is what comes out the speaker - with the only proviso being my personal preference, as applied to RoomEQ (Dirac) - so the room voicing is adjusted to my preference. - But it is a conscious choice... and something that is tunable - it is not imposed on the system, with no control being possible other than major system alteration (component replacement).

 

[dlaloum]

I wonder what the impedance curve looks like for the woofer compared to the ESL panel... I think by vertical bi-amping with your Crowns your getting 1/2 of the ML ESL hybrid right... Makes me think about using Zeros or Tubes for the panels part or some other Power Paradigm amp...

Just looked up the Martin-Logan ESL-9

According to the specs - it drops to 0.8 ohm @ 20kHz...
Havn't managed to track down the impedance of the woofer... but seems likely to be around 3ohm.

So it is an even worse torture test for an amp than my Gallo speakers... but in exactly the same way!
Sensitivity is 90db SPL/wm (so very slightly louder than my Gallo's)

The difficult part of driving these will be the ESL Pannel, the woofer should be relatively (! - it is 3 ohm but...!) easy - however the woofer is going to need more outright power than the ESL will most likely.

The part of the BiAmping that is probably right is the upper - for the bass, giving it something with more outright Watts would possibly give it a boost.

To give an idea... :

The Crown XLS1502 is rated at 300W@8ohm, 525W@4ohm and 775W@2ohm

So it can drive Voltage to:
8 ohm 300W = 139V
4 ohm 525W = 130V
2 ohm 775W = 111V

The amp will Clip (broadcast wideband distortion!) - when asked to provide a 111V peak into 2 ohm

So its "Real" power capability (in simultaneous wide frequency range terms) is whatever it can do up to 111V.

111V at 8 ohm is 193W - this will determine the max clean SPL peak the speaker attached can output if it has a 2 ohm trough anywhere in its frequency range.

Sadly, I have no data on what the amp would do into 1 ohm (or the 0.8 ohm of the ESL-9)... But I would not be surprised if the V out capacity were half the 2 ohm value - and the effective power output would be similarly "knobbled" - which would result in substantially lower peak (clean) SPL than for the 2 ohm calculations.

Which may or may not be an issue.

My own XLS2500... 440W@8, 1200W@2 - barely idles in my system, I feel I could get by with an XLS1500 quite easily.

But I have yet to hear an amp in my system that sounds as good - and I put that down to the ability to cleanly drive the low impedance loads.

ddude003 - if you want to point me to the amps you propose - we may be able to find specifications, and calculate out the limitations.... Many very very good amps, will crack when faced with a 0.8 ohm load!

 

[ddude003]

As an owner of ML ESLs and having discussed these issues at length here on AV NIRVANA I'll just quote Ralph Karsten... "The Power Paradigm assumes that amplifiers produce power and speakers are power-driven. Current produced by a power amplifier is not ignored and is considered in the amplifier's power response. Under this model, the ideal amplifier will make the same power into all loads, 4, 8 and 16 ohms. The typical amplifier, in this case, is a vacuum tube amplifier which usually makes its power into these loads via taps of its output transformer. There are a small number of transistor amplifiers that are designed with this behavior in mind also. Ideally amplifiers under this paradigm have little or no feedback. The Voltage Paradigm refers to such amplifiers as 'current source' amplifiers but the term is not accurate; in reality they are what they are called - power amplifiers."

"Loudspeakers that operate under Power Paradigm rules are speakers that expect constant power, regardless of their impedance. Examples include nearly all horns (currently the Avantgarde Trio is the only known exception), ESLs, magnetic planers, a good number of bass reflex and acoustic suspension designs. Horns, ESLs and magnetic planers do not get their impedance curve from system resonance and so benefit from a constant power characteristic and indeed, many of these speaker technologies are well-known as good matches with Power Paradigm amplifier designs."

I would only add that ESL panels have a frequency to impedance curve (map) that is counter intuitive to Voltage Paradigm amplifiers...

@dlaloum - I was not proposing any thing... However, Sanders Magtech and McIntosh (with autoformers) come to mind... No need to search, Sonnie has Magtechs and I have McIntosh...

 

[dlaloum]

As an owner of ML ESLs and having discussed these issues at length here on AV NIRVANA I'll just quote Ralph Karsten... "The Power Paradigm assumes that amplifiers produce power and speakers are power-driven. Current produced by a power amplifier is not ignored and is considered in the amplifier's power response. Under this model, the ideal amplifier will make the same power into all loads, 4, 8 and 16 ohms. The typical amplifier, in this case, is a vacuum tube amplifier which usually makes its power into these loads via taps of its output transformer. There are a small number of transistor amplifiers that are designed with this behavior in mind also. Ideally amplifiers under this paradigm have little or no feedback. The Voltage Paradigm refers to such amplifiers as 'current source' amplifiers but the term is not accurate; in reality they are what they are called - power amplifiers."

"Loudspeakers that operate under Power Paradigm rules are speakers that expect constant power, regardless of their impedance. Examples include nearly all horns (currently the Avantgarde Trio is the only known exception), ESLs, magnetic planers, a good number of bass reflex and acoustic suspension designs. Horns, ESLs and magnetic planers do not get their impedance curve from system resonance and so benefit from a constant power characteristic and indeed, many of these speaker technologies are well-known as good matches with Power Paradigm amplifier designs."

I would only add that ESL panels have a frequency to impedance curve (map) that is counter intuitive to Voltage Paradigm amplifiers...

@dlaloum - I was not proposing any thing... However, Sanders Magtech and McIntosh (with autoformers) come to mind... No need to search, Sonnie has Magtechs and I have McIntosh...

Thanks for that - I had not encountered the "Power Paradigm" although I have heard the term "current source" for years...

The Sanders Magtech has always been on my short list of "amplifiers of interest"! - But also in a price bracket where I have to be seriously convinced of the benefits!

The Crown XLS' unexpectedly provided me with an excellent solution... (unexpectedly, as it was an experiment in listening to class D, based on dirt cheap used amps.... which have since become my main amps!).

There is little or no opportunity (here in Australia) to listen to amps in ones home, to identify whether there would be a sonic improvement and whether it would be sufficient to warrant the price.

 

[ddude003]

Without knowing how the frequency to impedance curve (map) of your Gallo's it wold be difficult to predict if there would be much sonic improvement... The problem I was addressing with ESL panels was driving more power into higher frequencies... ESL panels can really screech when driven by Voltage Paradigm amps... And not necessarily from clipping...

 

[dlaloum]

The Impedance map of the Gallos can be found in this article:

Anthony Gallo Acoustics Reference 3.5 Speakers - HomeTheaterHifi.com

In 2007, I reviewed the Gallo Reference 3.1 speakers along with Gallo’s Reference AV center, A’Diva rears and the Reference SA subwoofer amplifier. I liked them so much I bought the entire system. Recently, Gallo has released the new Reference 3.5 speaker, replacing the venerable Reference 3.1.

hometheaterhifi.com

The CDT tweeter has a very "electrostatic like" impedance profile.... and the speakers sound very electrostatic like. (which is why I chose them!)

They replaced a set of Quad 989's (front) and Quad 63's (surround) - as a more WAF friendly alternative... while still meeting my sound quality criteria.

At the time, I thought that the Gallo's could not match the Quad's on complex massed orchestra pieces - where the sound seemed to get confused/mixed.... but since switching them from the Quad 606 amps to the Crowns, there has been an additional level of clarity, which makes me wonder whether the Crown/Gallo combo is now a match for my Quad606/989 combo of 10+ years ago... needless to say, impossible to tell! (and WAF will continue to mitigate against a return to large panels... so it's a bit of a moot point!)

 

[ddude003]

If I am reading those charts correctly it seems your Gallos dip into the 2.5 to 3 ohm range at ~20kHz... Not so bad...

If those Crown amps are class D and work for your Gallos, you might be interest in these New Class D Mono Amps using GaN Fets in a unique Patented Circlotronic output stage by Ralph Karsten... I believe they are Voltage Paradigm amps and your not running ESL panels...

 

[dlaloum]

I am having trouble finding the measurements/review that pointed to the 3.1/3.2 version of the speakers having 1.63 ohm tweeter.... (mine is a 3.2)

The final version of the family (3.5) is around the 2.5 ohm range and easier to drive apparently (no surprise!)

It gets difficult tracking down reviews from 12+ years ago...

I think you are reading those charts fine....

I have been keeping an eye on the Benchmark AHB2... another feedforward/current dumping design, as well as the new Purifi High Current 7040 modules... read a bit about the GaN Fets....

But first I would have to be dissatisfied with current performance.... and right now, it is hitting the spot - no itch there!

 

[AudiocRaver]

I find that, when an amp can easily drive a speaker, you can compare it to any other amp that can easily drive the speaker, and they all sound alike!

Where they start to sound different (assuming that they are basically good designs to start off with obviously) - is where the amps are used to drive speakers that are at the edges of, or outside of, an amps performance envelope....

Then different amps will misbehave differently - and they will sound different.

Agreed, and... not being argumentative, just a different perspective...

Let's look at a couple of the terms, like "basically good designs," and "misbehave." Many amps that are able to drive a reactive load like an ESL panel and sound good ("basically good designs?"), can be affected somewhat by that reactive load ("misbehaving?") depending, I believe on that amp's feedback structure. My NAD receiver's frequency response has a little ripple, around 0.5 db max variation from the flat, at high frequencies when driving the ML Classic 9's (as I recall, it has been awhile since I looked at those measurements). The HF response with the Crowns rises gradually from around 6 kHz to a peak of +3 dB at 15 kHz (again from memory). There is no indication of instability with either amp at different power levels looking at impulse and step responses. And they sound slightly different, as you might expect. I consider neither a bad design, nor as having been misapplied, and think of neither as misbehaving (nothing about the specs of either amp suggests it is not an appropriate choice for use with reactive loads). With either amp, I run Dirac Live and the differences are flattened out and then they do sound the same (pretty much I think, I am still working on that last point). I am guessing that the output of most AVR's and amps is affected at least a little by that load. Think of it as creative application of an amp with a minor undesired effect easily compensated for, goal achieved.

My point is that the design limits and misbehavior thresholds of amplifiers are not always as clear or straightforward as one might hope. My reasons for selecting the Crown XLS series to try out with the ML Classic 9's were:

• light weight
• moderate cost
• high power
• acceptably low distortion
• lots of power into low Z loads (with power into 2 ohms specified)
• remains stable with low Z loads
• no other spec points raised any red flags

Maybe not an obvious choice, but I have been happy with the result, and so far I see the Crown XLS series as a consideration for HiFi use.

The Crown XLS's will drive almost anything well... their performance envelope is extremely wide... most of the other series within the Crown range, don't have this ability (check out the 2 ohm specs on the different models.... the XLS's are quite different to most of the other series).

So true, I spent a lot of time looking at amp and AVR specs.

I noted that "clarity" too on my Gallo speakers - and I suspect that it is due to the absence of distortion artifacts caused by the amp trying to drive a 1.6 ohm load.... ie: the Crown is perfectly happy driving that load where the Integra AVR was absolutely not happy about it. (sounded veiled... muddy... not obviously distorting / clipping, but not sounding great)

Lots of speakers with exotic drivers (ESL's the CDT tweeters on the Gallo's) - have these issues.

Having said all that - the Crowns are not champions of Signal/Noise ratio's / SINAD - and if not properly gain matched with the preamp... most especially with high sensitivity / efficiency speakers... quite a few people have reported Hiss... (at low but apparently audible levels).

I have not experienced this on my 88db SPL/wm Gallo's

On the topic of Tubes / Distortion - ESL's (and the Gallo's) achieve part of their magic, by being typically an order of magnitude better in terms of distortion, than standard "dynamic" speakers. Yes there are exceptions in the realm of dynamic speakers... but as a genre - the "standard" speaker typically has 10x the distortion of an ESL.

This is one of the reasons for their "transparency".... - and they will typically show up whatever is being fed into them - for better or worse.

If your tube amp, is providing lush, romantic, 2nd Harmonic heavy, tunes - then that is what the speaker will reproduce.

Nothing wrong with that preference.

My personal preference is for a Neutral system - what comes off the recording is what comes out the speaker - with the only proviso being my personal preference, as applied to RoomEQ (Dirac) - so the room voicing is adjusted to my preference. - But it is a conscious choice... and something that is tunable - it is not imposed on the system, with no control being possible other than major system alteration (component replacement).

Good points all.

Just looked up the Martin-Logan ESL-9

According to the specs - it drops to 0.8 ohm @ 20kHz...
Havn't managed to track down the impedance of the woofer... but seems likely to be around 3ohm.

So it is an even worse torture test for an amp than my Gallo speakers... but in exactly the same way!
Sensitivity is 90db SPL/wm (so very slightly louder than my Gallo's)

The difficult part of driving these will be the ESL Pannel, the woofer should be relatively (! - it is 3 ohm but...!) easy - however the woofer is going to need more outright power than the ESL will most likely.

The part of the BiAmping that is probably right is the upper - for the bass, giving it something with more outright Watts would possibly give it a boost.

To give an idea... :

The Crown XLS1502 is rated at 300W@8ohm, 525W@4ohm and 775W@2ohm

So it can drive Voltage to:
8 ohm 300W = 139V
4 ohm 525W = 130V
2 ohm 775W = 111V

The amp will Clip (broadcast wideband distortion!) - when asked to provide a 111V peak into 2 ohm

So its "Real" power capability (in simultaneous wide frequency range terms) is whatever it can do up to 111V.

111V at 8 ohm is 193W - this will determine the max clean SPL peak the speaker attached can output if it has a 2 ohm trough anywhere in its frequency range.

Sadly, I have no data on what the amp would do into 1 ohm (or the 0.8 ohm of the ESL-9)... But I would not be surprised if the V out capacity were half the 2 ohm value - and the effective power output would be similarly "knobbled" - which would result in substantially lower peak (clean) SPL than for the 2 ohm calculations.

Which may or may not be an issue.

My own XLS2500... 440W@8, 1200W@2 - barely idles in my system, I feel I could get by with an XLS1500 quite easily.

But I have yet to hear an amp in my system that sounds as good - and I put that down to the ability to cleanly drive the low impedance loads.

More good points, especially the last, the ability to drive those low impedance loads, being the key success factor. Add to that the fact that the HF content, above 10 kHz or so, in most program material is relatively low anyway.

 

[dlaloum]

More good points, especially the last, the ability to drive those low impedance loads, being the key success factor. Add to that the fact that the HF content, above 10 kHz or so, in most program material is relatively low anyway.

My Quad 606 (unconditionally stable into any load, but current constrained at 2 ohm...) - did not sound as good as the crowns... I assume (in the absence of specs, and based on known specs for the related 520F 100W pro version of the 405... which drops back to 100W @ 2ohm) - that it can probably put out around 140W into 2 ohm....

In theory given the typical demand spread over the frequency spectrum - that might be enough... - given you typically have a 10db to 15db difference between the lows and highs.... Still it was consistently less effective than the Crown.

I probably could put some effort into determining what an average music/movie profile is.... to calculate the power requirements for the 2 ohm part of its impedance curve (say 10k to 20k) - and then compare to SPL reference level (75db) +20db headroom.... That would provide a benchmark...