|
|
Post by ryanmh1 on Sept 23, 2017 14:57:14 GMT -5
I will leave it up to Dan or Lonnie to present any further test results, but the assertion is that the amplifier tested to produce our published specifications did in fact meet them at the time they were generated. Well if it is so easy why don´t you just publish that exact same measurement? Otherwise we are comparing apples with - whatever.  Exactly it. And what is also less than clear is where the Emotiva amps transition to Class G/H operation. Based on Emotiva's and Stereophile's measurement's there's a bump at around 50W to 60W. Is this where the supply rails start their tracking? Possible. Could there be a problem that starts right in that range? Possible. Still, I *suspect* the issue was RF noise being injected somewhere after the input filter due to a power supply failure . This would use up the slew rate really quick, and cause what at least looks for all the world like slew limiting. The Emotiva posted charts don't disprove much of anything in the Stereophile tests. What is really needed is the 20kHz power sweep. That would reveal whether there is actually a problem or not. I suspect not, but at this point, it's still unknown. |
|
|
|
Post by leonski on Sept 24, 2017 1:20:28 GMT -5
G/H are power supply nomenclature. ALL EMO Amps are A/B output. Some might be considered High Bias, but still transition to straight A/B at some higher power level.
There IS not transition to G/H or whatever.
The low voltage rail responds to UP TO a certain power. The Higher voltage rail than takes over. When dealing with Real Music, this change should be buried in the noise and of no consequence.
Carver used several similar schemes for decades, starting with his Carver Branded m400t which had a PS that 'tracked' the input signal and supplied power on that basis. I think it also had 2 or 3 rails.
My example of that amp was quiet (a mechanical putt-putt noise was normal) and ran very cool. And was 200x2 It ran my low sensitivity Magnepan MG-1 speakers right to redline.
I'm beginning to believe that amp manufacturers are supplying information which the end users are Ill-Equipped to evaluate. If EMO had just made some kind of generic PS claims would we be having this chat?
Throw in some test data of equally obscure interpretation and you are on a real Sled Ride down the hill.
I'm still mulling over what ryanmh1 says are the differences between G and H power supplies.
I'm looking forward to whatever Stereophile has to say about a possible retest and their reaction to the EMO test data.
|
|
|
|
Post by ryanmh1 on Sept 25, 2017 10:54:04 GMT -5
A/AB/B/G/H are not power supply nomenclatures. They are output stage nomenclatures. A class A, AB, or B amplifier uses a fixed voltage rail. A class G uses two or more fixed voltage rails, and switches between them. A class H uses a modulated rail that tracks the input signal. I've seen speculation here and there that what Emotiva calls Class H is actually Class G. Below a certain level, all Class G amplifiers operate without switching the rail voltage. Many Class H amps also do this before they start modulating the rail voltage. The "power supply" can be either a linear supply with caps or a SMPS. Either type can be used in conjunction with a variety of output stage topologies. Emo has taken a topology usually used in PA gear and claims to have tamed it for high end use, and perhaps they have. Hopefully Audioholics will get a test unit at some point and do a full bandwidth power test and a THD sweep.
|
|
|
|
Post by leonski on Sept 25, 2017 11:53:48 GMT -5
Wrong, Ryan, Sorry G and H are PS topologies. A / B / C (unused) and D are all amp topologies, though 'd' amps come in Bridge or Halfbridge configurations. You can stick a switcher or linear of 'conventional' or G / H in front of ANY of the real amp classes. www.electronicdesign.com/analog/understanding-amplifier-operating-classesThe linked site calls G / H 'variations' on a traditional A/B amp. The output stage and to what extent it's biased determines 'class' of operation. |
|
|
|
Post by hdkeith on Sept 25, 2017 12:37:57 GMT -5
So I am just a guy that likes home theater and music and was in the market to move to separates and was looking at the XPA-9 Gen 3 since it would allow me to drive my 7.1.2 system without any need to use the amps in my AVR. I admit I am a specs and review guy becasue some of the stuff goes over my head so I have to rely on people with more experience than myself to help guide my decision. I had stumbled across the Stereophile review early in my search and it raised concerns with me. I now read this thread and am more baffled that before. I see valid points all around, but as someone else mentioned, while Emotiva pulled a warehouse unit and ran tests they did NOT run the same tests as Stereophile ran. I see no issue with pushing an amp to 1/3 its rated power to stress it, hell I see no reason not to push an amp to its rated power to stress it. While I will likely never get anywhere near that demanding use in my use it assures me something is built well and engineered to meet its specs.
In my simple take I look at it as A/B has been around for a long time and is a proven straight forward way to design an amp. All these more advanced designs are basically cost cutting form either parts, weight, etc. They seem to add a bit more complexity and thus complicate the design and reliability of the product. I am sure I will get flamed for that simple approach, but when I look at buying an amp it is a long term investment, vs. AVR that I toss every 3 years. I want to like this XPA-9 Gen 3 and wish I had the luxury of test gear and to sample ever amp I am considering in my own home, but I can;t so I am leaning toward going with a more proven A/B even if it weighs more and generates more heat.
|
|
|
|
Post by wilburthegoose on Sept 25, 2017 13:39:12 GMT -5
hdkeith - AMEN, my friend. I like to ask 3 questions (in order of importance to me): 1. Does it sound good? 2. Will it last? 3. Does it energy-efficient (green)?
|
|
|
|
Post by leonski on Sept 25, 2017 14:37:42 GMT -5
WTG?
1: Only you can tell. If EMO continued with the 'house sound', no reason it shouldn't sound fine when NOT clipped.
2: That's the question, isn't it? Between Higher MTBF parts, when available, premium circuit board materials, certain design factors AND maybe follow the lead of Bryston who does a 24 hour factory burn in of each amp?
Good warranty protection gives a feel-good aspect and what sounds like very good customer service, too.
3: A/B amps are only going to be so 'efficient', no matter what. Gains of a few points with a SMPS might help, but the REAL help is a Lot Less Weight to ship around.
And for 'HDKEITH'? 1/3 power is worst case for an A/B amp from the heat standpoint. I don't know WHY, but that's the way it works. Only the end user can decide the validity of such tests. Me? Not that it matters, but I think
such tests are good. You tend to accelerate lifetime that way and may find problems, like vastly insufficient heatsink. I'm personally willing to PAY for big heatsinks.
And Yes, the new SMPS does add complexity to the amp AND save weight. Time will tell.
|
|
KeithL
Administrator 
Posts: 10,261
|
Post by KeithL on Sept 25, 2017 15:45:40 GMT -5
Just to settle this - at least for now. In an amplifier, Class A, Class B, Class A/B, Class C, and Class D are reasonably well defined 0 and pertain to the operating conditions of the output stage. In a Class A amplifier, both output devices conduct current throughout the entire signal cycle (100% overlap) In a Class B amplifier, each output device only conducts current for half of the entire signal cycle (0% overlap) In a Class A/B amplifier, there is some overlap between the two output devices In a Class D amplifier, the output device or devices act as switches, so the signal is always either fully on or fully off In a Class C amplifier, which are generally only used in radio equipment, each output device conducts for less than half of the signal cycle There is no single authoritative description of what Class G and Class H actually mean; and you will find that different vendors define them slightly differently. (Various sources have historically described amplifiers with two or more sets of rails, or a single set of rails that is varied or modulated, as either Class G or Class H, and both as subsets of Class A/B.) He have chosen to refer to Class H as "a power supply topology" because it refers to the manner in which the output stage is connected to the power supply rails (so that seems to convey the situation the most clearly). If you look at the operating conditions in the output stage on our XPA Gen3, you will find that it complies with all of the characteristics which define "Class A/B". (We use two distinct sets of power supply rails, and a set of active devices which transitions smoothly between them, rather than a "hard switch".) A/AB/B/G/H are not power supply nomenclatures. They are output stage nomenclatures. A class A, AB, or B amplifier uses a fixed voltage rail. A class G uses two or more fixed voltage rails, and switches between them. A class H uses a modulated rail that tracks the input signal. I've seen speculation here and there that what Emotiva calls Class H is actually Class G. Below a certain level, all Class G amplifiers operate without switching the rail voltage. Many Class H amps also do this before they start modulating the rail voltage. The "power supply" can be either a linear supply with caps or a SMPS. Either type can be used in conjunction with a variety of output stage topologies. Emo has taken a topology usually used in PA gear and claims to have tamed it for high end use, and perhaps they have. Hopefully Audioholics will get a test unit at some point and do a full bandwidth power test and a THD sweep. |
|
|
|
Post by amped on Sept 25, 2017 15:50:22 GMT -5
Wow! 7 Pages on this topic...The bottom line was generated somewhere in this post. They asked does it sound good (to you), will it last and is it green. Well to me "is it green" is soooo negligible it's not worth mentioning. As for the other two I will say as an owner of some 17 odd Emotiva pieces over the years and currently with 4 XPA-2 Gen 2s (well now 2), A Sherbourne 7030 and an ERC-3 (not to mention all the cables etc. I can say they do not break. I haven't had a failure on anything yet. As for sound? Well Emotiva is a great value BUT you get what you pay for (actually a little more than that)...So if you are on a budget and want a lot of bang for the buck then have at it! But if you're looking for Krell, Classe, Bryston sound and "specs" out of your $900 amp you may be disappointed. To break down THD to where one thinks there is an audible difference in .00034 and .0034 is a tad funny. It's kinda like arguing over whether or not your 1984 Dodge Omni Turbo had 146 or 152 HP.
|
|
|
|
Post by rbk123 on Sept 25, 2017 16:01:24 GMT -5
I want to like this XPA-9 Gen 3 and wish I had the luxury of test gear and to sample ever amp I am considering in my own home, but I can;t The Emo may be the only one on your list where you actually have the luxury to test it in your house; why wouldn't you take advantage of that? Are you afraid you'll like the sound but others will tell you you shouldn't like it and so you'll think you chose poorly? Reminds me of wine tasting - people care more about if they're aligned with certain "experts" than if they actually prefer the wine. |
|
|
|
Post by ryanmh1 on Sept 25, 2017 16:25:50 GMT -5
Wrong, Ryan, Sorry G and H are PS topologies. A / B / C (unused) and D are all amp topologies, though 'd' amps come in Bridge or Halfbridge configurations. You can stick a switcher or linear of 'conventional' or G / H in front of ANY of the real amp classes. www.electronicdesign.com/analog/understanding-amplifier-operating-classesThe linked site calls G / H 'variations' on a traditional A/B amp. The output stage and to what extent it's biased determines 'class' of operation. Sort of. I'm not going to do a detailed technical rundown. This isn't the place for it. Your power supply is either linear or SMPS in 99.9% of commercial products.. Period. It is not a "G" or an "H". Your OUTPUT stage either runs on fixed voltage rails, or rails that are not fixed. With fixed rails, you get above that fixed voltage, you clip the amp. To avoid clipping, you have to increase that voltage. If you just jack it way up there, which is how a traditional high powered class AB amp works. Unfortunately, this is rather inefficient. So, years and years ago, the PA companies started using Class G or H amplifiers that would switch to a higher power supply rail. Exactly how this is implemented differs, and what is "G" or "H" is muddy, but the basic idea is the same across the board: When the amp needs more voltage than the low fixed rail can supply, switch over to a higher one. As you state, all are a variation on a traditional Class B or AB amplifier. Below, say, 60 watts, the amplifier just cruises along on a low power 30V rail. When the amp needs more than 60 watts, it ratchets up the voltage of the output stage. Whether Class G or H, this is what Emotiva is doing in the new series of amps. Hence, it really is primarily an amplifier output stage topology, since the amplifier is where those power supply rails actually get used. Rod Elliott has a good overview, as he often does: sound.whsites.net/articles/class-g.htmThe difficulty with Class G and Class H has long been handling the transition to the higher rail without introducing distortion from switching to the higher rail. I would have to dig through the Emotiva XPR schematic--supposedly "similar" to the G3 XPA--to be be able to see how they did or did not do this. It's pretty blurry, though, and hard to make out and ultimately, completely off topic. I am sure Emotiva will eventually figure out if the problem was a production or design issue, or a single unit failure. Sometimes, it's surprising. I recall one manufacturer having a problem with distortion being far higher than it should have been. Not horrible, like .02% instead of .002% second harmonic. All of the amps went out the door like this. I always wondered why that amp had so much second harmonic--it was just bizarre. Years later they figured out that a switch on the production line from a part from one manufacturer to a part from another manufacturer, supposedly identical, was the cause of the issue. Swap out the part, and distortion drops by an order of magnitude. So, moral is, you never know... And I have no clue what could be wrong with the sample XPR-5. I know a reasonable amount about amplifier design (enough that I go amplifier shopping by looking at schematics instead of faceplates, at least), but I know nothing at all about SMPS design, and next to nothing about rail switchers, and I suspect the issue somehow tracks back to one of those. |
|
|
|
Post by leonski on Sept 25, 2017 18:10:34 GMT -5
My original Carver Cube used some kind of rail PS. But it was quite a bit different, not having a huge transformer but it did make a nice putt-putt sound. Something to do with a Choke (single winding transformer, in effect) and an SCR switching on and off. You are so 'splitting a hair' with this. You are distinguising amp classes based on PS, as opposed to condition of bias, which is how most do it, with the exception of 'd' which is either 'on' or 'off', but the transistor STILL must have bias. For EMO to use as much prior art as possible, they Didn't 'reinvent' the amplifier, but rather adapted other amps sections (prior art) to a newer, more easily shipped PS with some interesting features. Most 'switchers', for example, have only OK dynamic power. By having a higher voltage rail, some of that problem might be 'fixed'. An extreme example might be NAD which can have 6db dynamic range. And what would appear to be highly UNregulated PS. The OTHER way to clip an amp is to exceed the output devices SOA. You could design a PS which would exceed the transistors ability to deal with it. For a MOSFET, I think that is Vds. I was never a probe guy, so I'm reaching here. en.wikipedia.org/wiki/Safe_operating_areaYou might be able to get away with wacky high PS voltage using something like an IGBT which is frequently a 600 volt device. But I'm not sure how linear for audio. Or RdOn, to influence damping factor. But I will agree with one thing without reservation. The mediation of switching rail voltages can cause problems and getting a modulating PS to work introduces yet another possible problematic feedback loop. Amp shopping by looking at schematics? Ridiculous. So many factors. 2 manufacturers could make near-identical amps and one is junk, the other a treasure. Parts ain't parts. Assembly procedures and other measureablees might also be different. One may choose to increase feedback (probalby a simple resistor choice) or to go for a lower power output / high bias design. Don't forget the HeatSink! How do you judge how something sounds based on the schematic? Or do all amps sound alike? How do you evaluate the ability to drive reactive loads? And what do you do with the OUTLAW M2200. They call it an A/B/G amp with......well, read the FAQ. www.outlawaudio.com/support/faq_200.htmlI shop for amps on several levels. Sound is first. For my panels, a brief look at specs shows 8/4 power ratio and may give some idea of current capacity. Pass amps with high numbers of output devices and a high bias design along with a huge PS are all design goals. Company reputation for reliability also plays into it. Are they accessible? USA based? |
|
|
|
Post by leonski on Sept 25, 2017 18:27:08 GMT -5
I want to like this XPA-9 Gen 3 and wish I had the luxury of test gear and to sample ever amp I am considering in my own home, but I can;t The Emo may be the only one on your list where you actually have the luxury to test it in your house; why wouldn't you take advantage of that? Are you afraid you'll like the sound but others will tell you you shouldn't like it and so you'll think you chose poorly? Reminds me of wine tasting - people care more about if they're aligned with certain "experts" than if they actually prefer the wine. 2-Buck-Chuck for me! |
|
|
|
Post by hdkeith on Sept 25, 2017 20:49:45 GMT -5
I want to like this XPA-9 Gen 3 and wish I had the luxury of test gear and to sample ever amp I am considering in my own home, but I can;t The Emo may be the only one on your list where you actually have the luxury to test it in your house; why wouldn't you take advantage of that? Are you afraid you'll like the sound but others will tell you you shouldn't like it and so you'll think you chose poorly? Reminds me of wine tasting - people care more about if they're aligned with certain "experts" than if they actually prefer the wine. The problem will be can I be sure in my short evaluation to stress it and be sure and without being able to a/b some amps will make it harder. Sure this amp is not in the league of amps many times the cost, but still I want to be happy with my buy long term. I may hang out a little and see how this thread runs it’s course to see if they find any real issues with the amp. I would not want to be shipping 55lb amp back and forth too much. |
|
|
|
Post by rbk123 on Sept 26, 2017 8:36:34 GMT -5
You will never be able to match what they did. Your source, your speakers and your volume will never line up with their test which is why your testing it out and stressing it in whatever way you think you'll listen to it is infinitely more valuable than what they did.
The Emo is more than in the same league as far more expensive amps - those amps will have the edge in build construction and aesthetics, but the sonic gains will be extremely small. Amps just aren't that complicated and the differences between very good amps is very very small. Obviously you have preconceived in your mind that cost equates to performance and Stereophile knows much more than you. Getting the Emo will cause you forever doubts and buyers remorse because until the light bulb comes on for you, you'll always be thinking "what if".
|
|
|
|
Post by hdkeith on Sept 26, 2017 10:25:24 GMT -5
You will never be able to match what they did. Your source, your speakers and your volume will never line up with their test which is why your testing it out and stressing it in whatever way you think you'll listen to it is infinitely more valuable than what they did. The Emo is more than in the same league as far more expensive amps - those amps will have the edge in build construction and aesthetics, but the sonic gains will be extremely small. Amps just aren't that complicated and the differences between very good amps is very very small. Obviously you have preconceived in your mind that cost equates to performance and Stereophile knows much more than you. Getting the Emo will cause you forever doubts and buyers remorse because until the light bulb comes on for you, you'll always be thinking "what if". I don't say cost equals better quality. It is a potential indicator, but not an absolute. There are plenty of newer ways to achieve better results while being at a lower cost point. My concern is clearly a review by a person with good credentials has pointed out a concern and even if it is something I may not to be able to easily hear of recognize why would I take that risk if there are other options in the same quality cost ballpark that are not known to exhibit those concerns (although they may have more issues of their own). I am still on the fence the XPA-9 G3 is appealing to me. |
|
|
|
Post by amped on Sept 26, 2017 10:39:44 GMT -5
You will never be able to match what they did. Your source, your speakers and your volume will never line up with their test which is why your testing it out and stressing it in whatever way you think you'll listen to it is infinitely more valuable than what they did. The Emo is more than in the same league as far more expensive amps - those amps will have the edge in build construction and aesthetics, but the sonic gains will be extremely small. Amps just aren't that complicated and the differences between very good amps is very very small. Obviously you have preconceived in your mind that cost equates to performance and Stereophile knows much more than you. Getting the Emo will cause you forever doubts and buyers remorse because until the light bulb comes on for you, you'll always be thinking "what if". Part 1: "The Emo is more than in the same league as far more expensive amps..." That is a VERY true statement! A great value and priced far below what a brick and mortar would charge for their products!
Part 2: "as far more expensive amps - those amps will have the edge in build construction and aesthetics, but the sonic gains will be extremely small..." This is the statement that caught my attention.. That statement is simply not even close to being associated with reality. With all do respect the only comparison you can make is shear volume (db/SPL) from one amp to another. Differences as in Soundstage, note decay, noise floor and about a 100 others help determine the "sound" of an amp. And trust me as an owner of Emotiva amps for years and an owner of Bryston Amps as well as Carver there is a HUGE difference in musicality between them and a very audible difference in a very musical way as you climb up the amplifier ladder.
Part 3: " Amps just aren't that complicated and the differences between very good amps is very very small..." Ouch! Tell that to John Curl, Bob Carver, Stuart Taylor, Tim Paravicini and Lonnie (EMO) just to mention a few. These guys are on a constant quest to "better" their products for better sound reproduction and not always for new sales but it is their passion. And the more expensive they get the better build, electronics, design and typically a better representation of the recording (not always but mostly) So...amps are that complicated and differences are not very, very small.
|
|
|
|
Post by rbk123 on Sept 26, 2017 11:35:49 GMT -5
Part 2 - I've owned many high end amps as well, including Bryston, Parasound, Conrad Johnson, Acurus, Hafler, Carver, etc.. Did I hear differences - yes. Was Emo the best - no. I call BS on your "HUGE difference" - the never ending "night and day" cliche. I stand by my statement the differences are small and subtle. People need to convince themselves this crap is true so they can justify paying the huge mark-up.
Part 3 - all of those individuals will tell you off the record amps aren't complicated, but that's not going to sell now is it? You need to separate Marketing from reality. Once you get to a certain quality level, subsequent differences are going to be very small. Present but small.
|
|
|
|
Post by leonski on Sept 26, 2017 13:25:29 GMT -5
Don't forget the Speaker! It would appear that amps driving fairly easy loads and well below clipping sound much alike. Unless flawed in some fashion.
It is when you start throwing a REAL speaker at amps, even with the same power, that differences appear. RYANMH1 mentioned the Power Cube. A wonderful way to stress an amp and
determine By Measure how well it does, at least electrically, with stressful loads.
Most HT speakers, even if lower sensitivity, do not qualify as 'bad load'. My 4ohm panels of maybe 84db sensitivity do not qualify.
|
|
|
|
Post by tchaik on Sept 26, 2017 13:27:25 GMT -5
Part 2 - I've owned many high end amps as well, including Bryston, Parasound, Conrad Johnson, Acurus, Hafler, Carver, etc.. Did I hear differences - yes. Was Emo the best - no. I call BS on your "HUGE difference" - the never ending "night and day" cliche. I stand by my statement the differences are small and subtle. People need to convince themselves this crap is true so they can justify paying the huge mark-up. Part 3 - all of those individuals will tell you off the record amps aren't complicated, but that's not going to sell now is it? You need to separate Marketing from reality. Once you get to a certain quality level, subsequent differences are going to be very small. Present but small. I've owned the expensive krell, audio research, spectral as well as the mid priced hafler, adcom, rotel amps and others. i agree that the differences between the mid-priced units and the very expensive high end units were small but audible on the best recordings (differences that did not warrant the cost unless you had money to burn). biggest differences came with changing speakers, cd players, preamp, turntable and cartridges. so... having said that, my system (with the xmc-1 driving the XPR-2 and XPR-5) is still the best sounding system i have ever owned. tchaik................. |
|
