Components Specifications use and the Truth

[FrantzM]

Hi

Specs do not tell all the story. For most here that seems to be the opinion. They do however tel part of the story. I would like the gurus to chime in on Specs, their truthfulness and how they can be used as a guide in choosing components.
While many (most) here at WBF rely on audition for final determination, the reality is that this has become increasingly difficult to audition gears. A good way to interpret and find the truth within even published specs would help.

The idea of the thread came to me in an ongoing debate about FMA amplifiers. A. Wayne pointed something I never really consider: Most so-called Class A SS amps slide (unadvertised) to Class B so that they can "double down" ... Big names, big amps et all most double down amps don't (can't) do it down to the lowest loads in Class A ...

Same for speakers. Most are not as sensitive (I prefer to use the term "inefficient" but microstrip is keeping me straight so .. ) as they say they are . And many 90 dB speakers are actually 87 or even less ...
Another important specs is the input sensitivity of amps, most amps are 26 dB gain devices, many are not some are 32 dB or some even more ... Input impedance of an amp is another needed parameter for matching different technology or brands (say tubes preamps with SS amps, etc) . Same with impedance matching between amps and preamps.. Not as straightforward as it would seem .. as some amps impedance characteristics vary with frequencies (same with preamps). FR of devices are often load dependent .. All this would help determine n advance the matching of equipment. Since I am in a good mood I will use the audiophile overused term "synergy".

Rise time (important in my opinion but not as much as settling time which let me know about the time the component takes to settle back after a pulse, important in my estiamtion inthe way the amp reproduce a short pulse without oscillation, Slew Rate etc

What about power ratings? Peak output versus Continuous?

Specs can help and a way to interpret them would help many of us. What say you , the experts? I know we are asking a lot as books could be written on the subject of specifications. An ongoing discussion can only help. Questions , answers, replies and clarifications are welcome and helpful

 

[dallasjustice]

I am no guru, so I'm not sure whether you want my opinion.

If I knew how to use an oscilloscope, I would probably do so when testing amplifiers. Unfortunately, I can only go by the specs provided and JA's measurements. I've tried a lot of different amps and I keep score of the measurements and compare with what I hear. In my case and with my speakers, output impedance is one of the most relevant specs. Unfortunately, there's no standardization. Is output impedance measured using standard size speaker cable or at the amp's binding posts? Is output impedance the same throughout the frequency spectrum? Does it vary with load or power output? At the end of the day, I can only decide based on listening. Having said that, specs CAN be useful in winnowing down choices.

I don't think output impedance is the ONLY spec that matters. But, it really can help to narrow things down. I have yet to hear an amp that sounds remotely good in the mid-bass with my speakers with an output impedance more than .2 ohm. However, I've heard Hypex class D which has ultra low output impedance and it sounded mediocre in the HF. Some of these things can be inferred from the amp's design; How many output devices? How much feedback?

I was recently looking at a solid state amp where the dealer website claimed a super low output impedance (significantly less than .1 ohm). I then went to the manufacturer's website which claimed a super high output impedance (.4 ohm). The dealer admitted that the manufacturer's website spec WAS correct, not his. The dealer tried to convince me that a .4 ohm output impedance was very low for a solid state amp. That's absurd and I told him so.

 

[mep]

Frantz-I’m not surprised that your thread is a hotbed of inactivity-and not because it deserves to be so. Quite the contrary in my opinion. The problem is that people who claim they love them some measurements in most cases mean they love them some unsubstantiated specifications. I have written numerous times about this topic on this forum and have tried to explain the difference to people who think they are one and the same. They aren’t and people should never give the same credence to specifications as they do to actual measurements.

Unfortunately, I think lots of people are quite comfortable with specifications in the absence of measurements and take them to the bank just like they would if they had measurements to take to the bank.

 

[A.wayne]

Its a non -issue MEP and academic really , as the discussion will automatically lead to independent test results , as I'm not aware of any manufacturer giving out detailed data necessary for such a discussion ...

Regards,

 

[dallasjustice]

Maybe if words like "fraud" were used, there would be more interest. There's no question that some manufacturers misrepresent their specs to sell more gear. That's fraud in my book.

 

[DaveyF]

This old war horse discussion regarding 'measurements' is an interesting one. I remember years back, a S'phile review of the original Krell 300s and the VAC 300b mono amps ( I think these were the two amps in question but)...the cover stated something like: If one of these is right the other must be wrong" or something like that.
The VAC measured poorly, while the Krell being a ss amp measured much better. However, to this day, i think most listeners would prefer the VAC amps. IMO, the best measurements are the ones that your ears tell your brain are good.

 

[jazdoc]

Measurements are there to guide you, not tell you what to do. If you try to pair a 2 watt 45 SET amp with 82 db speakers, you are going to be unhappy, irrespective of how good the amplifier and speakers are.

Specifications are useful for designers and manufacturers to test hypotheses/designs and for quality assurance. Unfortunately, we have yet to discover a single measurement or set of measurements that can tell the designer and consumer how a piece of equipment sounds. If we did, everyone would own the same equipment within a certain budget range!

 

[A.wayne]

This old war horse discussion regarding 'measurements' is an interesting one. I remember years back, a S'phile review of the original Krell 300s and the VAC 300b mono amps ( I think these were the two amps in question but)...the cover stated something like: If one of these is right the other must be wrong" or something like that.
The VAC measured poorly, while the Krell being a ss amp measured much better. However, to this day, i think most listeners would prefer the VAC amps. IMO, the best measurements are the ones that your ears tell your brain are good.

Maybe you prefer the distortion spectra of the VAC vs that of the Krell ....?

 

[A.wayne]

Measurements are there to guide you, not tell you what to do. If you try to pair a 2 watt 45 SET amp with 82 db speakers you are going to be unhappy, irrespective of how good the amplifier and speakers are.

Specifications are useful for designers and manufacturers to test hypotheses/designs and for quality assurance. Unfortunately, we have yet to discover a single measurement of set of measurements that can tell the designer and consumer how a piece of equipment sounds. If we did, everyone would own the same equipment within a certain budget range!

Bingo .........We have a winner ...

 

[microstrip]

This old war horse discussion regarding 'measurements' is an interesting one. I remember years back, a S'phile review of the original Krell 300s and the VAC 300b mono amps ( I think these were the two amps in question but)...the cover stated something like: If one of these is right the other must be wrong" or something like that.
The VAC measured poorly, while the Krell being a ss amp measured much better. However, to this day, i think most listeners would prefer the VAC amps. IMO, the best measurements are the ones that your ears tell your brain are good.

I think it were the Cary 805 monoblocks. Martin Colloms wrote a letter to Stereophile about it. Although it is dated, as tube amplifier design has evolved a lot in 20 years, it is still very interesting. IMHO the key sentence of the paper is If a reviewer starts to value just one aspect of reproduction, the entire balance and framework of the high-end audio industry is endangered. Perhaps measurements can help as a check for this.

http://www.stereophile.com/content/cary-audio-design-cad-805-monoblock-power-amplifier-martin-colloms-594

 

[MylesBAstor]

Bingo .........We have a winner ...

A means, not an end.

 

[bretdago]

This is an interesting discussion... As a designer and manufacturer, my use of specs is a baseline for consistency and a development tool. As an example, my BSC amps are around .06% distortion at 1K rated power into 8 ohms. I am happy with the performance that I get from this product, the ratio of gain, to feedback, to bias, and any number of other factors is working for this design. Now if I increase the class A bias for example passed where I originally calculated it to be, distortion will not in fact get much better, if at all, and eventually it will get a lot worse. Therefore, for argument sake, I have found the perfect balance The smile is because it is a paradox to find the perfect balance. I can re-adjust the circuits, increase feedback, change gain, change current output in a stage and get better distortion than my standard of .06%. In doing so I have changed so much of the circuit that now my performance has changed in a way that may not work for me sonically. Now all being equal, if the circuit is left alone and I have this baseline golden set of specs that I work with for a certain design topology, and I make a slight change to the pre-driver stage as an example, I may really like what that does sonically to the product. In the measurements that may translate to be a decrease in distortion at 1k or maybe a decrease in distortion at 20k or 30K, or in fact it may have actually gotten worse, however I still could prefer the sound. When this occurs, you will stay with the sound you like and continue to fiddle, more subtly now mind you, to improve upon these specs to meet or surpass the original, set by the golden standard. It can be never ending, and for some of us we continue to play with the circuit until all the possibilities are exhausted then we try a different topology…

So for me specs are a guideline for consistency, no matter how many amps I build, if the topology is the same and the golden specs do not deviate more than a few percent plus or minus, the product will perform virtually the same from sample to sample.

Now when I listen to an audio system, I would not be able to say that a particular company’s amplifier had more distortion than mine did. Unless the design topology was more or less directly proportional to mine, the interpretation of the specifications is not apples to apples. So comparing a set of specs between a 100watt tube amp and a 100watt SS Amp will tell you virtually nothing about what you are going to hear unless a spec is wildly out of whack.

Where I think specs help is when there is more data to plug into the equation, i.e. speaker, room size, listening habits etc… Earlier in this thread Frantzm brings up an interesting set of specs. Some amps double power some don’t, mine do, other amps I like in the world don’t, and where is this relevant. As an example, if you have a speaker in a big room that is let’s say 85dB efficient and nominally 4 ohms and dips below 2ohms at like 200hz and you like Pipe organ music at over 110db in your chair then you actually could use an amp with a stiff supply and the ability to double its power. Most amps are not up to this task so here is an example where a spec could help you tailor your buying decision based on environment. In a different case, where things are more similar, a 94db speaker 8 ohms medium room, 85-100db at listening position then specs play a different part. Power becomes less of a concern by most standards and the applicable contender’s increases. Other specs like noise may become a lot more important in this case, you may narrow the field by only accepting amps with better than 95db of noise. In any rate, using specs to filter some of the choices we make helps, but in no way does it define what sound is the most pleasing to you. It does however create a baseline for all to work toward, more to make sure a product group can meet a minimum quality standard.
What would be interesting to most listeners is the difference between design philosophies i.e. things like DC coupled to AC coupled, Transformer outputs to no transformer outputs, 5 watts to 5,0000 watts etc. etc. etc. …

This is not meant to be a trigger for a million different specification discussions, but serve as a guide to one man’s use of specs and hopefully help one person or at the very least sell another amplifier.

 

[JackD201]

Welcome to WBF Bret. I haven't heard your Bully Sound amps but hope to get the chance in the future. Great, practical post by the way. I hope you'll have time to contribute even more.

 

[JoeyGS]

@Bret....Well said and put things in better perspective

 

[A.wayne]

Bret ,

Welcome and congrats on continuing the tradition of building amplifiers for us 1ohm nutters ..

As to distortion , we know that the amount is not as important as its Spectra , that is the harmonic distribution , as to being, 1st,2nd,3rd etc...

@the OP,

When back behind my PC again , I will post up Test pics for discussion in relation to our previous post in the FMA forum ..

Regards

 

[bretdago]

Thank you gentleman for your comments. A.wayne, I would like to see that... There is most definitly a correlation to distortion spectra and what we hear. Being able to measure this consistency along with frequency spectra could change the way we compare products and voice them for market. Could take the fun out of it as well

 

[microstrip]

Great points Bret!
FIY, I particularly appreciated your opinions about using measurements to check consistency, not as predictor of absolute sound quality.

 

[A.wayne]

You can predict sound quality from measurements once you have locked in on the specific DNA of the amplifier you are working with.



Regards ,

 

[microstrip]

You can predict sound quality from measurements once you have locked in on the specific DNA of the amplifier you are working with.



Regards ,

Nice. We have been discussing about the psichoacoustics approach and now we have genetics in the debate.

 

[bretdago]

Nice. We have been discussing about the psichoacoustics approach and now we have genetics in the debate.

I see what you did there and I liked it...