Crown XLS2502 Stereo Amplifier Review

[March Audio]

I agree.

Do you think less than a quarter of the rated headline power figure is a reasonable IMD test for all amplifiers or just the Purifi you posted above? Or should we just call the Purifi a 100W amplifier and be done with it? You can't have it both ways.

View attachment 44283

Consider all amplifiers rated IMD back in the dark ages with a 4:1 60Hz/7KHz at full rated power...

Thats just an example of the format to look at relevant in band distortion as opposed to irrelevant out of band distortion. As I said earlier you can choose the power levels, but 1/4 rated power is a far more relevant to real world conditions at 20kHz - even then still way over the top.

 

[March Audio]

So, apparently life should be enjoyed through a low pass filter, custom designed to hide all the stuff we don't want to know about, huh?

Well you havent been complaining about CD brick walled at 22 kHz for the past 40 years. You havent been complaining about the studio microphones which rarely get past 30kHz, or the studio analogue reel to reel tape decks which rarely got passed 35kHz, or indeed your speakers that rarely get past 30kHz. Or what about the out of band noise thats in DSD recordings?

Then of course the final low pass filter, your ears, which you are lucky if you get past 15kHz*

So why all the fuss about amplifiers?

We have been around this multiple times. The point is that THD (or THD + Noise) is not a useful metric when the harmonics (or noise) are at ultrasonic frequencies. The FTC test is useless for judging the "quality" of an amplifier. Its outdated technically, but also its purpose should not be conflated. Its just a very simplistic method of trying to assist technically naive consumers to compare power output. It does not demonstrate audible quality and, as can be easily demonstrated, has little relevance at higher frequencies.

I will add to this, as I have said many times previously, it is good to look into the ultrasonic region to see if anything odd is going on. It should not be ignored. However using simplistic metrics such as THD in this region is very misleading.


*middle aged duffers such as us.

 

[pjug]

Could amplifier rating standards not be configured to show what amps with lower continuous (like my Classé 70wpc class AB power amp) will provide as "clean" power peaks" that actually serve music requirements rather than requiring a continuous power rating for the peak demands?

It's been done before. NAD still uses IHF dynamic power spec. I hope @restorer-john has an opinion on this.
http://www.bassboy.com.au/getreel/site/classicamps/files/articles/ihf/article.htm

 

[restorer-john]

The FTC rules were cast in a time where such amplifier designs were not viable - the lack of fast switching devices being the major limitation

Is that true? I guess someone should have told the Sony engineers who in April of 1977, after starting development in early 1975, released the TAN-88 to the world, using, wait for it, a fast switching (500KHz) PWM amplifier using VFETs and for good measure, a nice SMPS supply to run the thing. A year later (March/April 1978), they released a minor update to the unit and it continued to be sold until 1984.

So, 7 years in the marketplace and it was not viable? I don't think so. They sold a absolute ton of them, and apart from some premature SMPS failures, many are still in use 43 years on. Historical revisionism is not required here.

Bandwidth, THD and adherence to FTC (and DIN) standards wasn't a problem for Sony either. Why? Because they were honest about the harmonic content (above 20KHz) and gave the unit a realistic THD of <0.5% at full power from 20-20KHz. Even so, it was better than spec.

There's a great thread on Class D further back by DonH where I posted the schematic of the power amp if you are interested.

 

[restorer-john]

It's been done before. NAD still uses IHF dynamic power spec. I hope [@restorer-john[/USER] has an opinion on this.

I was very excited back in the early 80s when NAD released their "Powertracker" 2200 power amplifier using commutating HV rails and what appeared to be sensational peak power capabilities.

A 'soft' set of rails that dipped heavily under sustained loads, but could produce what seemed like enormous short term power. I figured the design was interesting, likely quite valid and deserved attention. Listened to the unit when it first came out on various JBLs up to the L250s and some other speaker brands I have forgotten. I bought one, but soon discovered (along with everyone else) they were failure prone and poorly made.

Just like the big Hitachi class G (HMA-8300), the various Kenwoods, a nice looking Proton and even some big Pioneer integrateds of the early 80s, all of which had jumped on the high dynamic power bandwagon, they were unreliable and just didn't sound "right". They simply ran out of steam when really pushed and things got ugly, fast. Compared to big traditional amplifiers, they came up wanting and when they failed, they took out a lot of expensive (at the time) silicon right back to the front end. Charred phenolic and magic smoke was a common by-product.

NAD stuck with their newly renamed "power envelope" for decades and of all the topologies I have in my storerooms for donor parts, theirs and the Kenwood commutating rails make up the lion's share of the amplifier boneyard.

I still believe a well regulated normal power amplifier with stupendous short term capability if realized well, could be another game changer. Rather like the 1980s Proton D-1200 except built properly...

 

[pjug]

I was very excited back in the early 80s when NAD released their "Powertracker" 2200 power amplifier using commutating HV rails and what appeared to be sensational peak power capabilities.

A 'soft' set of rails that dipped heavily under sustained loads, but could produce what seemed like enormous short term power. I figured the design was interesting, likely quite valid and deserved attention. Listened to the unit when it first came out on various JBLs up to the L250s and some other speaker brands I have forgotten. I bought one, but soon discovered (along with everyone else) they were failure prone and poorly made.

Just like the big Hitachi class G (HMA-8300), the various Kenwoods, a nice looking Proton and even some big Pioneer integrateds of the early 80s, all of which had jumped on the high dynamic power bandwagon, they were unreliable and just didn't sound "right". They simply ran out of steam when really pushed and things got ugly, fast. Compared to big traditional amplifiers, they came up wanting and when they failed, they took out a lot of expensive (at the time) silicon right back to the front end. Charred phenolic and magic smoke was a common by-product.

NAD stuck with their newly renamed "power envelope" for decades and of all the topologies I have in my storerooms for donor parts, theirs and the Kenwood commutating rails make up the lion's share of the amplifier boneyard.

I still believe a well regulated normal power amplifier with stupendous short term capability if realized well, could be another game changer. Rather like the 1980s Proton D-1200 except built properly...

Thanks for that. The newer NAD AB amps (e.g. C270, C275BEE) don't do the rail switching though, do they? Don't they just have higher voltage rails that allow higher power output for short bursts, but can't be sustained? Do I misunderstand?

Edit: NAD does still use rail switching (part of their PowerDrive which is still used in the BEE amps). I didn't know.
https://www.stereophile.com/content/nad-c-372-integrated-amplifier-what-powerdrive
also, sorry to take things off track

 

[restorer-john]

Thanks for that. The newer NAD AB amps (e.g. C270, C275BEE) don't do the rail switching though, do they? Don't they just have higher voltage rails that allow higher power output for short bursts, but can't be sustained? Do I misunderstand?

Edit: NAD does still use rail switching (part of their PowerDrive which is still used in the BEE amps). I didn't know.
https://www.stereophile.com/content/nad-c-372-integrated-amplifier-what-powerdrive
also, sorry to take things off track

Perhaps they just renamed it again. PowerTracker, Power Envelope, PowerDrive. Made some minor tweaks and called the low voltage rail a "high current" rail, gave it decent regulation (hopefully) and called the HV rail "high voltage". Foot in each camp and the punters should be happy I guess.

It's all been done before, bigger, better and cheaper in real terms, but don't let that discourage you, apart from the cheap plastic front panels on some current NAD gear, their overall quality is leaps and bounds above what it was. I almost bought a 320bee and matching CD player with JM-lab Focal speakers the other day (2nd hand for cheap) to play with, until I touched the crappy front panels.

 

[levimax]

But it now makes no sense to go looking for harmonics of in-band signals above 20kHz. Why? Because the amplifier has deliberately filled that space with all the shaped noise that allowed it to get such good in-band performance. This is a technique unavailable to the designer of a conventional (ie class A/AB) amplifier. There going looking above 20kHz can tell you something interesting about in-band operation. But once the designer is able to exploit noise shaping, looking for harmonics of in-band signal out of band is simply meaningless. What you see out there tells you very different things about the in-band performance.

Bottom line. Limiting the bandwidth of the measurements (not the amplifier implementation) of harmonic distortion products is done simply because not doing so is mathematically invalid.

It was just a matter of luck that making measurements of harmonic products above 20kHz was useful with conventional amplifiers. And we got used to the idea. It provided insights into in-band operation, and thus was useful. But demanding that all amplifiers behave this way is mathematically indefensible. There is bandwidth to be exploited to make the operation where it matters better, so we use it.

Thank you for clarifying the noise shaping and issues with measuring Class D amps which makes sense. My question is how would you measure the effects of the low pass filter used on Class D amps on the "in band" signal? My understanding is that any filter causes phase shift, group delay, and ringing. Would issues with the low pass filter show up on SINAD or would additional test have to be run?

 

[amirm]

Bandwidth, THD and adherence to FTC (and DIN) standards wasn't a problem for Sony either. Why? Because they were honest about the harmonic content (above 20KHz) and gave the unit a realistic THD of <0.5% at full power from 20-20KHz. Even so, it was better than spec.

Specs say "THD" not "THD+N." Do you know if they did not subtract noise?

FTC guidelines also just state THD, not THD+N.

 

[restorer-john]

Specs say "THD" not "THD+N." Do you know if they did not subtract noise?

No I don't know for sure, but the residual is quoted noise at less than 100uV (A wtd) which would compromise the S/N to circa 111dB (unwtd), based on 160W@8R (35.78V RMS)They claimed better than 110dB S/N, so my guess is they included the A WTD noise in the THD so it really was THD+N, like most of the older gear at the time. Albeit exactly what out of band content there was, in that figure, we don't know. We know A weighting takes 10dB off by 20KHz and drops like a stone after that.

As far as I can tell, the separate THD (not including N) numbers seemed to start to be advertised in the early/mid 80s. I'm not remotely an expert on early harmonic distortion analyzers and would appreciate someone with experience in that era to weigh in.

The 100uV (0.1mV) is a pretty high residual, and I can confirm residual noise is audible on those amplifiers close to the tweeters, but considering when they were made and the relative ground they were breaking, I think they were pretty impressive. They certainly had no perceptible hum, so that end of the spectrum is already in front of a similar 160W/ch traditional design. They also were optimized for the best response for a typical 8R load, not 4R.

FTC guidelines also just state THD, not THD+N.

My understanding of the FTC rule was/is a total harmonic distortion (to the 10th harmonic of a 20KHz, so a 200KHz bandwidth?) figure (which can include total noise) is specified from 250mW to rated power across the quoted bandwidth (usually 20Hz-20KHz) with a specific load.

Pity we can't send an APX-555 back in time to some of those engineers. It'd be amazing to see what they came up with or refined.

@amirm you were at Sony when, the early 90s? We know little about that, I for one would be interested.

 

[RayDunzl]

Pity we can't send an APX-555 back in time to some of those engineers. It'd be amazing to see what they came up with or refined.

Probably a bunch of Class D amps...

 

[restorer-john]

Probably a bunch of Class D amps...

Smart ass.

 

[Francis Vaughan]

My question is how would you measure the effects of the low pass filter used on Class D amps on the "in band" signal? My understanding is that any filter causes phase shift, group delay, and ringing. Would issues with the low pass filter show up on SINAD or would additional test have to be run?

All filters involve phase shift. That is how they work. Not all filters ring. SINAD is independent of filter characteristics.
Filters are fully charaterised with a few parameters, and from these one can derive the result of any issues that might occur in the passband. The frequency response you see for these filters derives from the same parameters, and there is a duality between the shape of the frequency response and the effects you worry about. In general, if you see a very sharp filter it is likely to have untoward effects in the passband. Ripple in the response in the passband is evidence of aggressive filtering, and very rare to see.
Anyway, the bottom line is actually that the likelihood of any chosen filter having anything more than a barely detectable effect is low. The reviews of DACs here usually measure the effects of the reconstruction filters in the DACs, and these are much more aggressive than filters found in class D amplifiers.

 

[Francis Vaughan]

Is that true? I guess someone should have told the Sony engineers who in April of 1977, after starting development in early 1975, released the TAN-88 to the world,

To some extent, this rather makes my point. Sony had to develop new devices to make this amp possible. Before they got to work, there were no output devices that could switch fast enough. In the context of consumer level amplifiers this is, at best, an outlier.

I had vaguely remembered this amp existed, it is nice to be reminded.

A couple of comments. The specs claim the distortion measures are to DIN-45500, not the FTC rule. This is a bit odd, as DIN-45000 isn't a distortion measuring spec, but rather is a spec defining whether an amplifier is HiFi or not. There was a DIN distortion measuring spec, but that isn't it.

Having a look at the schematic, and doing a bit of digging, one sees that the output filter on this amp is a 6th order Bessel with a cut off of about 70kHz, and is 40db down at 250kHz. Which is the sort of filter one would hope for. The Sony guys were never dumb.

The amp uses a fixed clock, which was common until reasonably recently. The advent of self clocking class D is a major contributor to the phenomenal performance we now see, and is intimately tied to the noise shaping that gets them there. Which is a big reason why the Sony only managed 0.5% THD. For the time, impressive.

I had a look for the definitive standards on distortion measuring. The DIN standards are obsolete and seemingly unobtainable. Even for money, no-one will sell me one. However the AES standard for testing digital audio equipment AES17-2015 is available to AES members. The critical point of here is that this standard explicitly demands that a low pass filter with greater than 60db attenuation at 24kHz is placed before the analyser. Thus harmonic distortion of any test signal above 12kHz is not measured.

 

[Severian]

I have their spiritual cousins, the BIC RTR EV-15s. Compression horn, midrange radiator, and 15" woofer. 95dB @ 2.83V/1 watt, 1 meter, 24Hz-20kHz +/- 3dB and by my quick and dirty measurements that seems legit. They look like ghetto blasters, but they don't sound like it... no exaggerated bass, it's a pretty flat response. MSRP is $800/pair but street price is usually more like $400/pair with free shipping to store.

With the grilles on, you don't even have to look at the fugly red woofers. Not my photo... mine are stashed in the garage so the wife doesn't murder me. But they can look somewhat respectable: https://www.dropbox.com/s/e82oejiyerw1zk0/BIC-Eviction-Living-Room.jpg?dl=0

The tweeter is not the last word in surgical audiophile detail, but it is not harsh either. They are pretty tonally correct out of the box and obviously there is an abundance of headroom to play with if one wants to shape the response with EQ/DSP. They easily play louder than needed with my pedestrian HK 3390 amp, and really shine with a high-output amp like the Crown XLS.

Check Google, these things have the beginnings of a small cult following. At $400/pair they are a strong contender for the best bargain in our hobby. 99.9% of audiophiles will turn their noses up at these without a second thought based on price and appearance, but throw in a $100 MiniDSP and some modest amplification and you've got a system with flat smooth response and effectively unlimited headroom down to the 20hz range for maybe $800 all told, perhaps even less. What a time to be alive.

I own these as well and agree with everything you've said. I currently use them in the living room powered by a middling home theater receiver and they sound fantastic at SPLs that I could have never hit with less efficient speakers. I've also used them as PA speakers powered by a Crown amp and they didn't break a sweat. Truly the best deal in audio.

Further, I currently use said Crown amp to power 85dB efficient speakers in a nearfield setup and it sure sounds hi-fi to me. The power really makes the speakers come alive.

 

[Pillars]

I use one of these full range and it performs decently. Absolute gobbs of power seem to overcome any shortcoming.

 

[restorer-john]

To some extent, this rather makes my point. Sony had to develop new devices to make this amp possible. Before they got to work, there were no output devices that could switch fast enough. In the context of consumer level amplifiers this is, at best, an outlier.

I had vaguely remembered this amp existed, it is nice to be reminded.

A couple of comments. The specs claim the distortion measures are to DIN-45500, not the FTC rule. This is a bit odd, as DIN-45000 isn't a distortion measuring spec, but rather is a spec defining whether an amplifier is HiFi or not. There was a DIN distortion measuring spec, but that isn't it.

It's a bit chicken and egg with those faster VFETs. They'd already used their previously developed J18/K60 for a few years (from '74) in their upper range integrated amplifiers TA-4650/5650/8650. Did they get the slower devices as a byproduct of their research to make a switching amplifier or did the prospect of a commercially viable switching amplifier drive the manufacture of the J28/K82 they got for the 88/88B? Or did the VFET itself come to being due to Sony's search for a better horizontal output transistor for color TV? Who knows. Maybe Nelson Pass does- he teamed up with Sony to build a statement VFET amp using Sony NOS parts a few years ago. I've haven't seen enough early TAN-88s to see device date codes prior to 1976. They are a rare beast now.

The DIN spec always seemed to be added underneath the headline FTC specification. You can take the top Sony TAN-88 figure as FTC rated- it was early in the compliance and they hadn't got to the highest standards of English they achieved later on.

From what I've always seen, in hundreds of spec sheets, the so-called "DIN" power spec was less taxing and was stated as power at 1% THD@1KHz into a specific load. It appears that it still remains a popular specification to this day, with all the Hypex and Purifi modules specifying the same way. Makes sense if it was a EU default spec I guess.

I have always taken "DIN" specs as the poor cousins of compliant FTC rated gear. After all, 1% as a reasonable THD number to hang your hat on was thrown out the door by Harold Leak's Point One range of valve gear, in the mid 1940s...

Here's a few examples of gear I could think of where the 1%@1KHz was the DIN rating.

Pioneer:

Marantz:

Yamaha:

Rotel:

In the case of the TAN-88, by having an FTC and "DIN" spec that were identical, it shows that they were pretty much on the money with power and THD (conservative but not excessively).

 

[restorer-john]

I had a look for the definitive standards on distortion measuring. The DIN standards are obsolete and seemingly unobtainable. Even for money, no-one will sell me one. However the AES standard for testing digital audio equipment AES17-2015 is available to AES members. The critical point of here is that this standard explicitly demands that a low pass filter with greater than 60db attenuation at 24kHz is placed before the analyser. Thus harmonic distortion of any test signal above 12kHz is not measured.

This does speak to the wider issue of what standards are currently in force and how they:
a) specifically differ to past standards (relaxed in what ways?)
b) differ around the world in different countries and jurisdictions
c) whether companies are exploiting this in international borderless trading

The playing field is not level anymore, that's for sure. And you are right, the "golden age" of truth in specifications is sadly, long departed (again). For the past 25+ years, we've seen a return to the bad old days of rubber specifications and an apparent lack of interest by regulatory bodies in holding dubious claims to account.

Realistically, change comes from either within the industry (unlikely) or from consumer advocate groups and/or advertising standards authorities exerting pressure and I certainly don't see the ACCC (in this country) taking any interest as they are more interested in pursuing the "big fish" like Volkswagen, Harvey Norman or some other high profile organisation that provides media attention and helps to justify their existence.

 

[anmpr1]

...the "golden age" of truth in specifications is sadly, long departed (again).

Realistically, change comes from either within the industry (unlikely) or from consumer advocate groups...

Audio 'journalism' ought to be doing this, however in most cases the audio press is simply an arm of the manufacturers (and possibly the few dealers that remain). I consider ASR a consumer advocate outfit, but with few exceptions, none of the big 'on-line reviews' fill that roll. Possibly local groups grounded in reality, such as the Boston Audio Society, if you are lucky enough to live nearby and are able to participate. Then, the few manufacturers that for their own part sell value oriented gear backed by top rate engineering. But those can't always be counted on--witness for example McIntosh, which for most of its existence was out of the 'goofball loop', but has now embraced things like 'tri-wiring' speakers and 'double-speed' CD players.

 

[Bruce Morgen]

How did you know ? AU-717
But I prefer my AU-9900. (1975)

What, no

(introduced in 1979)?