Welcome! Need support, you got it. Or share your ideas and experiences.
Return to Planar Speaker Asylum
205.74.162.114
Any thoughts on pairing NuForce Class D monoblocks with Acoustat Model 3 speakers? My buddy is having a fire sale.
Follow Ups:
Most all Class D Amps do not like driving the Capacitive loads of an Electrostatic Speaker. Nonetheless, I drive my 2+2 with a H2O type ICEpower amp (ICEPower 500A modules with a class A type linear power supply, just like the H2O). The ICEPower ASP / A Spec sheets indicates max Capacitive load of 470nf....I will have to check the 2+2 tonight and see what it has....from 1000ASP spec sheet:Loading
With its low output impedance, the ICEpower1000ASP is designed to be unaffected by loudspeaker loading
characteristics. However, care should be taken with purely capacitive loads.
Traditionally amplifiers have been tested extensively in laboratories with purely capacitive loads. This was done to
test the amplifier's stability and performance but it does not relate to any normal speaker load as even
electrostatic speakers do not present a purely capacitive load to the amplifier but include a resistive part as well.
The maximum purely capacitive load allowed is 470nF.The 200ASC modules can only handle 330nf...
But still my favorite amp for the 2+2 (and all other speakers I have) is the Krell KSA-50 (heavily biased to 72 wpc @ 4 ohms) clone. It just does everything right.I swap my tubeamps in every now and then (Quicksilver 8417 clones with KT120 output tubes), but it just doesn't quite cut it....
Also, don't forget the Sanders amps - I had one (ESL) and it was excellent on Electrostatics...sorry I sold it. The Magtech is even better...
I VOID WARRANTIES
Edits: 02/19/16 02/19/16 02/19/16 02/19/16
Is the Quicksilver clone not cutting it on bass? or on performance on themid and top end?
I don't recall the KSA 50 mids and top end being as good as a low feedback KT120 PP amp - within the class A region I liked the classe DR 9 better, it is 8-10W stock and can take 20-25W upped bias so long as the load is > 4ohms beyond that bias you need to start thinking of fan cooling. You would also want to up the temperature ratings of the cans.
The original ksa-50 used s l o w MJ15003 / MJ15004 output devices with fT of 4mhz, but everyone still loved it. I subbed more recent and faster Toshiba 2SC5200 / 2SA1943 at 30mhz, as well as another pair, so this could explain the better mids? Choosing output devices that are too fast and you have oscillation issues...the group vetted the best sounding output device and I went with it. no regrets.
I did prefer the sound of my KSA clone over a real KSA-250. Builders Bias? Donno.
The KT120 tubeamp does have a bit of feed back (can't remember what I used, but I was true to the original 8417 design. I dropped in (and biased up) the KT120 as the 8417 were expensive and had a bad reputation of self destructing when bias was uneven or pushed. Quite aggravating seeing a NOS $50 - $75+ Sylvania tube red-plate and self destruct in seconds while setting initial bias. But they were great sounding tubes. Extremely easy to drive, so not much front end was required..
I VOID WARRANTIES
The KSA series was somewhat uneven. I liked the 50 very much. The 100 mostly since it was unflappable on the CLS. The 200 did the Apogee Diva well enough but didn't have the 50's detail and sweetness the 250 didn't.sound right on the Apogees but at least could drive them.
I have a couple of friends taking out the feedback loops off their tube amps and preamps with great results. The problem with the feedbackless tube amps is that they go REALLY loose in the bass. The compromise solution is local feedback on the power tubes without going through the coupling capacitor within the loop.. Not perfect bass but still better than without feedback. You can do the KT120 without feedback in triode very well. But it does not work for tetrode/ultralinear, ,for which feedback is necessary.
In biamping, the fedbackless tube amps are incredible on top - so long as you don't ask for too much power. SETs without feedback do this even better. Once you get used to them you develop the Futterman crossover distortion allergy and PP circuits sound broken to you.
"The problem with the feedbackless tube amps is that they go REALLY loose in the bass. "
This is probably due to their using too small iron in the output tranformers.
Tube amps with sufficient output iron do not have loose bass. My Friend's VAC 30/30, for example, runs without feedback (it is actually switchable in 1db steps from zero to 6db) and has fantastic tight but very tuneful and musical bass. Same goes for the better SET amps I have heard.
You need to have it running in triode I think because that gets you a much lower plate resistance and makes the transformer design less problematic.
"SETs without feedback do this even better. Once you get used to them you develop the Futterman crossover distortion allergy and PP circuits sound broken to you. "
That is the truth.
The problem is the whole thing is affected by load what's sounds good at 8 ohm will not work at 2/1 ohm , I have always rebiased and adjusted the bias to the specific low z being driven , you have to , the Z and phase of a loudspeaker plays a big part in amplfier THD and the transition to class-B , If one re-bias any Krell thought to be hard sounding on any specific speaker most would not believe it's the same amplifier ...
The transition to B is very important, if done wrong or happens at the wrong time you can hear the clunk ..:)
Kind of makes you think that designing for Class-B all the time isn't such a bad idea eh? :)
Better efficiency (obviously)....you don't experience the gain non-linearity when both output devices are on....devices are only sourcing current (vice push-pull.)
There are some other trade-offs as well.
Class AB is a (seemingly) excellent compromise between the Class A and B, but it isn't all that it's cracked up to be. :)
Dave.
Crossover distortion anyone?? BTW, negative feedback doesn't fix this fundamental flaw in a Class B amp.
Who said anything about negative feedback? But anyways, your statement is not entirely true.
Crossover distortion frequency content (generally) varies directly with amplitude. Negative feedback being more effective at lower frequencies, it can effectively reduce crossover distortion at lower power levels. Lower power levels, fortunately, is where we operate our power amplifiers most of the time relative to their maximum ratings.
As usual, there are other variables involved here, but it's not correct to say negative feedback doesn't have any effect on this "flaw" of Class B amplifiers....it can, and does.
We might need to define/differentiate "fix" and "eliminate" in this context.....if you want to be anal about it.
However, I won't disagree that crossover distortion is a fundamental aspect of Class-B amplifiers since the output devices are indeed in cutoff half the time. However, careful design can alleviate the effects quite a bit......both measurably and subjectively.
Regardless of how you feel about the "flawed" concept of Class-B audio amplifiers, if we turned off all the power amplifiers that made use of it, there would be a WHOLE lot of quiet audio systems in the world.
Cheers,
Dave.
Davey, you are simply wrong on many levels.
"Crossover distortion frequency content (generally) varies directly with amplitude"
Umm, no it doesn't because it is a distortion that occurs when the signal is passing through zero. This means that this distortion is basically constant with signal level.
"Negative feedback being more effective at lower frequencies, it can effectively reduce crossover distortion at lower power levels"
THis is a nonsensical statement, Davey. First you talk about low frequencies and then low power as if the two are the same thing. Silly.
The TRUTH, Davey, is that crossover distortion is more of a problem at low power than at high power and that is because crossover distortion is constant so as the power goes down the % relative to the signal level acutally INCREASES. This is why Class B is not for any serious high quality amplification and designers learned a long time ago to avoid this as much as possible.
" Class-B amplifiers since the output devices are indeed in cutoff half the time. However, careful design can alleviate the effects quite a bit......both measurably and subjectively."
No, it cannot, it is inherent in Class B operation. This is why people do AB designs or A to a few watts if possible or if they are more ambitious full class A.
"there would be a WHOLE lot of quiet audio systems in the world."
Only if playing loud I guess and most of those would like sound better off anyway I an imagine.
As usual, you're not reading my statements closely enough and then knee-jerking.
Crossover distortion can/does indeed reduce at lower power levels. Crossover distortion is NOT constant.Also, read again what I said about 'frequency content' of crossover distortion, and don't conflate it with the 'level' of crossover distortion. (That's a key aspect.)
And I will say it AGAIN, there are other variables at work as well. Emitter follower output stages (especially) present some interesting issues....Vbias setting/tracking.....etc, etc. Trade-offs are always with us.However, that said, I DID NOT say that crossover distortion was not inherent in Class-B operation. I said that "careful design can alleviate......" Once again you are mischaracterizing my statements to conform to your argument. This seems to be an ingrained trait with you. Oh well.
I will certainly stipulate there are indeed inherent issues with Class-B designs. I never said there weren't. There are inherent issues with ALL amplifier designs.
I was simply replying to your statement that negative feedback can not influence/reduce crossover distortion. It CAN....under the right circumstances.
I know this seems counter-intuitive to you, and I thought it would catch your eye.
But don't take my word for it. Query this premise (CORRECTLY) to some of your amplifier design buddies and see what they say. I think you'll find that this "crossover distortion is a constant" conclusion is not as done-and-dusted as you think.Regardless, I suggest you do some further reading on this topic.
Dave.
Edits: 02/26/16
No, crossover distortion happens as the transistor goes to zero volts and turns off. This means that by definition the signal amplitude is irrelevant. Transistors, need a bit of voltag before they respond and if there is no positive bias, as in class b, the there is a flat spot in the curve around zero volts. This means it amplitude and frequency independent. This can verified at a number of sources. What it also means is that the relative impact is worse at low levels, which is also well documented. Again it is frequency independent. Please show me a legitimate source that says otherwise.
I suggest to perform a test measurement yourself. You'll need a spectrum analyzer, good 'scope, and obviously you'll need a Class-B (or an AB amplifier you can under-bias to Class-B operation) amplifier.
Output a high frequency sine wave at say 40 and 4 watts into a standard load. Let's say you measure 0.01% crossover distortion at 40 watts, then you would expect to measure 0.1% at 4 watts, yes? I guarantee you won't.
The reason is because your "flat spot" doesn't stay flat when you increase the output power level. This alters the relative distortion spectrum of the two different power levels and thus it's NOT frequency independent and not "constant." It's also why negative feedback can have a noticeable effect to reduce (not eliminate!) crossover distortion (better) at lower power levels, because in that case the harmonic content is concentrated more at low frequencies.
If the components used in said amplifier were all perfect....transistors with no non-linearities...no thermal tracking issues...bias solid....etc, etc, then you (and the textbooks) would be correct in your assertion. However, these are REAL-WORLD amplifiers and they exhibit numerous non-linearities and variables that skew them away from "textbook" results.
My primary "legitimate source" is my test bench. I've verified this.
Quit believing everything you read and start proving some of this stuff to yourself with practical experience and testing!
Cheers,
Dave.
If what you claim to measure is true then I would suggest that the amp you measured is not purely Class B.
From Elliott sound products website:
"Crossover Distortion *# : A form of distortion caused by the power output devices in a push-pull amplifier operating in Class-AB. This occurs in valve and solid state designs, and is caused by one device switching off as the other takes over for its half of the waveform. There are some designs that claim to eliminate this distortion by never turning off the power devices, but in reality, only Class-A amplifiers have zero crossover distortion. This is generally measured as a part of the THD of an amplifier, and becomes worse as power is reduced from the maximum."
Note the last part, "and becomes worse as power is reduced from the maximum"
"Solid state amps tend to create high order odd harmonics, so there will be the 3rd harmonic, only a tiny bit less of the 5th harmonic, and the harmonics will extend across the full audio bandwidth. Transistor and MOSFET amps have very high open loop gains, and use feedback to reduce distortion. In all cases, the crossover distortion is caused because the power output devices are non-linear. At the low currents at which the changeover occurs, these non-linearities are worse, as well, the devices usually have a lower gain at these currents.
This has two effects. The open loop gain of the amplifier is reduced because of the lower output device gain, so there is less negative feedback where it is most needed. Secondly, the feedback tries to compensate for the lower gain (and tries to eliminate the crossover distortion), but is limited by the overall speed of the internal circuitry of the amplifier. This results in sharp transitions in the crossover region, and any sharp transition means high order harmonics are produced (however small they might be)."
"I do not believe that we can simply ignore crossover distortion on the basis the "everyone knows how to fix it, and it is not a problem any more". I would suggest that it is still a real problem, only the magnitude has been reduced - the problem is still alive and well."
In a Class B amp it is certainly very alive and very well.
Now, let' look at some amps that have visible crossover distortion in their traces and how they behave with THD+noise vs. power.
"Fig.9 indicates that the distortion is predominantly the subjectively innocuous second harmonic, though there is a suspicious-looking spike in the residual waveform almost coincident with each zero-crossing point. This spike is the result of a picket fence of higher-order harmonics (fig.10), though it's fair to note that these are all relatively low in level.
Read more at http://www.stereophile.com/content/dan-dagostino-momentum-monoblock-power-amplifier-measurements#oPtGg3HVVI7zzb8F.99"
"Perhaps more important, the distortion content (fig.8) has some higher-harmonic content present. This is coincident with the waveform's zero-crossing points, which suggests that the Paladin's output stage is a little under-biased. This harmonic content can also be seen in the spectrum of the amplifier's output while it drove a low-frequency tone at high power into 8 ohms (fig.9). The highest-level harmonic is the relatively benign third, at -64dB (0.06%), but the more problematic fifth, seventh, and ninth harmonics also make appearances, between -80dB and -90dB.
Read more at http://www.stereophile.com/content/portal-audio-paladin-monoblock-power-amplifier-measurements#SzLhG5FuAMhddzFW.99"
"However, when I examined the distortion content, I was surprised to see—superimposed over what appears to be basically a third harmonic (fig.4)—what looks like crossover distortion. (The spikes in this 'scope trace occur at exactly the zero-crossing points of the sinewave.) Yes, the measured level in this graph is very low (0.003%), but crossover distortion tends to be more audible than its level would imply, due to its high-order content and the fact that it tends to dominate at low signal levels. At high powers (fig.5), it tends to drop into insignificance. Intermodulation distortion was very low, the 1kHz difference component with the very demanding 19+20kHz test signal remaining at -90dB (0.003%), even just below the onset of visible clipping (fig.6).
Read more at http://www.stereophile.com/content/herron-audio-m150-monoblock-power-amplifier-measurements-part-2#YIt1qwqub8Fx6QM7.99"
Have a look at the McCormack:
http://www.stereophile.com/content/mccormack-dna-225-power-amplifier-measurements#6KjWBSgkwAzYIBZR.97
Look at the distortion waveform...you can see a slight bump at every zero crossing of the main signal...
All of these are modern AB amps that still have a visible crossover notch. Most of the rest have suppressed it further but the distortion from it is still evident in the FFT spectra.
I think you've lost track of what your assertions were. You stated crossover distortion is constant and negative feedback has no effect on it.
For crossover distortion to be "constant" it would vary (as a percentage) as the inverse of the power output. ie, 2%/2-watts, 0.2%/20-watts, 0.02%/200-watts.
Rod Elliot notes that negative feedback can reduce crossover distortion, but "never" eliminate it. I don't disagree on that.
Rod also notes that crossover distortion is not constant. He believes it "becomes worse as power is reduced." I do disagree on that. However, he's basing that statement on the percentage "increase" relative to output power, vice a strictly constant level in absolute terms.
Regardless, none of those four JA amplifier tests that you've pointed to exhibit performance that demonstrates anything other than some crossover distortion effects from real-world non-Class-A amps. So what.
The tests don't disprove my assertion that crossover distortion is NOT constant, nor prove your assertion that it IS constant. However, it would be interesting to remove/reduce negative feedback in those particular amplifiers and then re-evaluate the crossover distortion level and spectrum. Do you suspect it would shift/change?? :)
Anyways, you won't get an argument from me that Class-A is superior to Class-AB/B....with regard to crossover distortion. The question was how much crossover distortion could be alleviated with negative feedback and/or biasing into Class-AB. As you've noted, even these four Class-AB amplifiers do exhibit some crossover distortion characteristics. Although I don't agree that it's a visible crossover notch in the sine wave reproduction. I would need to look at the original high-resolution plots from JA.
Cheers,
Dave.
Elliot's assertion that it is worse as power is reduced is exactly the point. As a percentage of power it increases, which is exactly what happens when the crossover distortion is constant and the main signal goes down. You do understand what relative means, right??
You would agree with me, I hope for your sake, that if crossover distortion is constant then as power goes down the % distortion of the main signal will go up. The way distortion is reported is always relative to a reference and not absolute. Otherwise, it would be reported in micro or nano volts and not as %.
Interestingly, many Class AB amps show a higher %THD at low power than they do at several watts. This might very well be because of xover distortion, and yes it is often a factor of 10 or more(usually shown on a LOG scale).
I said negative feedback doesn't eliminate crossover distortion and I am correct.
My point of showing those amps is two-food. 1) even many modern Class AB amps exhibit crossover distortion and most others you may not see it visually in the time based waveform but you see the effects in the FFT. 2) all those amps had significant levels of negative feedback and yet had pretty large amount of crossover distortion, which demonstrates that negative feedback won't tackle this problem when the output stage is under biased. A true Class B amp will be even worse.
You can't tell me these effects aren't audible in a highly negative way.
JA even states that it is crossover distortion in these examples.
I understand what relative means....I explained it clearly with %/power examples. You're just repeating what I said with different language. Silly.
You do understand what constant means, right?
None of the example amplifier measurements you've pointed to (or any that you could point to) will demonstrate crossover distortion that is "constant" at varying power levels. It's not possible. As I said previously, only if we are talking about a "textbook" amplifier would the "constant" assertion be correct. I'm talking about real-world amplifier designs with less (sometimes way less) than perfect performance. Also, constant levels of crossover distortion at varying power levels would still yield identical harmonic content in the spectrum of the output. None of those measurements demonstrate that either.
THD obviously includes many other forms of distortion. I've been referring to crossover distortion exclusively in my comments here. Sorry if I wasn't clear on that.
Separating out the various harmonic distortions in integrated testing (if so desired) needs to be an objective of the data presentation.
Neither have I claimed crossover distortion could be eliminated by negative feedback, so I am correct. :)
You might want to re-read my post and remind yourself it was you that brought up the topic of negative feedback. :)
http://www.audioasylum.com/forums/MUG/messages/22/220433.html
"Audible in a highly negative way" is a subjective evaluation. Subjective evaluations are meaningless. We are obviously evaluating many other factors of the amplifier performance during listening tests.
If referring to JUST crossover distortion, much would depend upon the higher-order harmonic content and the levels generated before we could validate and/or possibly agree on your statement.
Our subjective perception of (and the weighting relative to other factors) various harmonic distortion content produced in audio equipment is not an uncontroversial topic. Some very smart folks are polar opposites with their opinions on this.
In this thread, I've highlighted some of the aspects of amplifier real-world performance that don't necessarily follow our academic/textbook understanding. Audio power amplifier conceptual design is relatively simple...it's the realization that's the challenging part. :) More practical testing and objective evaluation experience is the key to your further understanding of this topic.
Dave.
None of the examples I have pointed to are true Class B amps, you know this. I never claimed that they were constant for the amps I gave examples too other than it is clear they have a problem even at low power, an area where Class AB is supposed to "fix" the problem but it clearly doesn't.
My assertions about an amp that runs in truly Class B are accurate. For sure what you have been measuring at home is not a true Class B amp. Since basically no one uses Class B for hifi it is no wonder that a pure example is hard to find. You were the one claiming that it can be useful if designed correctly, not me.
"Also, constant levels of crossover distortion at varying power levels would still yield identical harmonic content in the spectrum of the output. None of those measurements demonstrate that either.
"
No doubt, basically a forrest of harmonics are generated but the sonic IMPACT of those distortions at different at different SPL. At low levels they are more annoying as the sensitivity to such things goes down as the SPL goes up.
"THD obviously includes many other forms of distortion. I've been referring to crossover distortion exclusively in my comments here. Sorry if I wasn't clear on that.
Separating out the various harmonic distortions in integrated testing (if so desired) needs to be an objective of the data presentation.
"
If crossover distortion is present in an amp's measurements then it is likely to be the dominant form of distortion, at least at low power and probably with many amps up until the actually clip.
""Audible in a highly negative way" is a subjective evaluation. Subjective evaluations are meaningless."
What planet are you on, Davey?? Do you "listen" with your oscilloscope or do you "listen" with your ears?? It seems to make no difference to you. Subjective evaluations about sound quality are the ONLY thing that matters in audio. Correlating those preferences with measurements would be nice and can probably be done but it is very tedious and in the end only useful to designers who want to make an amplifier that people actually want to listen to rather than look at it's FFT spectrum.
"We are obviously evaluating many other factors of the amplifier performance during listening tests."
Yes, and they are all subjective because you cannot quantify any of what you hear.
"If referring to JUST crossover distortion, much would depend upon the higher-order harmonic content "
Crossover distortion, by definition, is a break in the transfer function. This means that by definition it makes a whole spectrum of distortion components. This means harmonic components out to the Nth harmonic and anything above about 5th harmonic is pretty unpleasant anyway. This has been demonstrated and known for a long time, Davey...no need for me to reinvent the wheel...it is why people actively try to avoid this type of distortion. Why do I need to prove to you what has been proven?? Surely you already know this and are just trying to play devil's advocate here.
"Our subjective perception of (and the weighting relative to other factors) various harmonic distortion content produced in audio equipment is not an uncontroversial topic. Some very smart folks are polar opposites with their opinions on this.
"
It is uncontorversial to people who have bothered to listen rather than saying theoretically, "oh that level is too low to be heard", which is the typical engineer speak for "We did it the textbook way...how dare you say it sounds wrong". Most of those you are referring to have too much inveted from themselves to look at the underlying sonic issue in their designs. As a scientist I look at cause and effect and not engineering orthodoxy. If the engineering orthodoxy produced a superior sounding product then we likely wouldn't be having this discussion.
If you think that the whole tubes and in particular SET revival is only about nostalgia then you either haven't heard a good one with an appropriate speaker or you are in denial because you have serious cognitive dissonance in what you are hearing with what you "know" are poor measurements.
I won't disagree that higher-order distortions can be more audibly unpleasant. I specifically stated that "much would depend upon the higher-order harmonic content and the levels generated before we could validate and/or possibly agree on your statement." Did you read that part???
"Listening with my oscilloscope or listening with my ears" is just juvenile rhetoric. I expected better from you Brad.
I'm on the planet where your subjective evaluation is meaningless to me......and mine is meaningless to you. Do you disagree with that? If so, then we ARE on different planets. :) You said it yourself, "you cannot quantify any of what you hear." I agree with that.
Subjective evaluation is, by definition, incontrovertible, which means it trumps any form of objective evaluation in the mind of an audiophile. It's unfortunate, but that's where this industry has been for many decades.
Objective evaluation is the only form of evaluation you and I could both perform and be able to correlate our findings.
I hope you're not saying there isn't a nostalgia aspect to tubes and SET. :)
Gleaming chrome chassis with beautiful orange reflections and a $20,000 price tag add to the allure, yes?
Vacuum tube amplifiers are inherently robust. Any dolt can put one of those together and have it sound halfway decent and not blow up. That can not be said for SS amplifier design.
Dave.
" Did you read that part???"
I did and it doesn't change my statement. Crossover distortion produces the whole spectrum of harmonics out to...well out past where most will measure (I guess 20th harmonic is sufficiently out there). As to level, again as the power goes down the impact from ANY crossover distortion is perceptibly worse. At high levels other factors will also have significant impact and perhaps crossover distortion is then a minor part of the total distortion but at low levels, where other distortions are minimized, the constant level of crossover distortion becomes relatively worse. At exactly what level this happens I cannot tell you precisely as it probably also depends on your speaker sensitivity and what your average power draw is from the amp how important it is among other things.
""Listening with my oscilloscope or listening with my ears" is just juvenile rhetoric. I expected better from you Brad.
"
No it is is a (slight) exaggeration of how I think you are allowing your expectation bias about what you think measurements mean to influence what you are (or are not) hearing.
"I'm on the planet where your subjective evaluation is meaningless to me......and mine is meaningless to you. Do you disagree with that? If so, then we ARE on different planets. :) "
It is really sad that you think this way. I have direct and repeated experience that this is not the case. My subjective opinion on audio matters is highly respected in my audio circle of friends and as a reviewer (I get questions and compliments about reviews I wrote 10 years ago). I have also been influenced by some very knowledgable and experienced manufacturers and listeners/dealers. They helped to steer me away from a purely objective mindset about audio as it was clearly not leading to satisfying listening for me. I have had a big influence on a number of the old Apogee crowd who have turned away from big SS and Class D amps to far better sounding electronics. Maybe you haven't found anyone that agrees with YOUR subjective findings and this is why you think this way. That you and I are on different planets is a foregone conclusion in my mind ;-).
"Subjective evaluation is, by definition, incontrovertible, which means it trumps any form of objective evaluation in the mind of an audiophile. It's unfortunate, but that's where this industry has been for many decades.
Objective evaluation is the only form of evaluation you and I could both perform and be able to correlate our findings.
"
Well, there has been research in CORRELATING the objective with the subjective and this is where I spend a lot my time reading and trying to understand what is underlying human perception that could even conceiveably allow us to prefer the sound of soemthing with objectively higher distortion. It was hard for me at first to accept but I have come to terms with it. What I have done though is look for gear that is consistent with the principles that were laid out by Cheever, Hiraga and others. Obey what the ear/brain wants or you don't get great sound just mediocre sound.
"I hope you're not saying there isn't a nostalgia aspect to tubes and SET. :)"
Very little because how many people do you know who were alive in the 1920s and 30s? How many of those are still audiophiles. I was born in the 70s and by then the transistor was king. SETs were nearly dead by the 1940s when PP tube amps (Williamson design particularly) with feedback took over. Didn't sound better but made more power with less size.
"Gleaming chrome chassis with beautiful orange reflections and a $20,000 price tag add to the allure, yes?"
You will pay that and more for any number of SS amps as well with just as much shiny shit (D'Agostino's 20,000 leagues under the sea amp for example). What you get in a good SET and a big part of the cost is big output transformers. Those cost some coin for good ones. That is the reality. ALL amps are overpriced these days (or nearly all).
"Vacuum tube amplifiers are inherently robust. Any dolt can put one of those together and have it sound halfway decent and not blow up. That can not be said for SS amplifier design.
"
Kind of but not really. If you really want to make a good one then it is no easier than an SS amp. Arguably the best sounding SS amps are those from Nelson Pass's First Watt line. Those are stupid simple as well but selecting the correct operating points and the right parts is not so simple. Your big complex SS circuits are the ones that almost invariably fail in the one criteria that matters most...the listening. For sure they are clever and engineering marvels...which don't obey the ear/brain requirements so they are made for pride not listening.
The continued quoting is not necessary. You and I are the only one's reading and I know what I said. :)
I find it really sad that the objectivist and subjectivist camps have become so polarized/separated through the years. In spite of what you may think, I consider myself a member of both. I used to find snide comments directed at me irritating, but now just find them mildly entertaining. Oh well.
I find it really sad that audio equipment "reviewing" has devolved to the level it has. Subjective reviewers (most with little/no technical background) are ubiquitous. A few exceptions exist....JA for example. The problem is not just lack of objective reviewing, but lack of objective thinking. Reviewing should be a combination of both subjective and objective, not one or the other. Oh well.
I find it really sad the "high-end" audio industry has shot itself in the foot for many years and is ultimately doomed. As the baby-boomers like myself die off and take their disposable income with them, I don't see later generations continuing to support the industry....as it currently exists. Oh well.
I hope you're not referring to the older F1/F2 First Watt amplifiers. :) Let me know what you think of the F7 amplifier when you review it. :)
Dave.
I am following so you guys are not alone there.
I keep waiting for you guys to start working out how to make USEFUL measurements and create a listening and reporting protocol for subjective evaluation to make USEFUL comparisons and absolute observations beyond "nice" and "made me feel X emotion". The emotional reaction component is a "nice to know" personal bottom line but is a non-transferable personal effect as others will have different preferences and would have other reactions to the same thing.
I like Art Dudley's reviews because he states his personal biases and preferences outright and largely reviews equipment in that context. While I don't share all his preferences, I "get" what turns him on in equipment performance.
The TAS tradition as Harry Pearson pushed it was useful to me so long as I could filter out what the particular reviewer was after - not that easy since most did not adhere strictly to the Harry Pearson concept so diluted their observations by blending them with personal preferences and emotional impressions.
Most people are phase deaf in music listening since they did not grow up with real voice and instrumental music but only with electronically reproduced music much of which was electronic to begin with. Hopefully they can make use of evaluations made by the likes of HP who reference real acoustic instruments and unampified voice in real spaces via the relatively few recordings that do retain the information (though there is a bit of a chicken and egg problem there).
I really miss listening sessions with Harry and John. :(
Is the high fidelity report still operating ..?
New post about opera.
Read the master's thesis by Daniel Cheever. It might not be perfect, but it definitely goes in the right direction towards having a quantitative evaluation of sound quality.
The Title says it all:
"A NEW METHODOLOGY FOR AUDIO FREQUENCY
POWER AMPLIFIER TESTING BASED ON PSYCHOACOUSTIC
DATA THAT BETTER CORRELATES WITH SOUND QUALITY"
THe main objective of the thesis:
"This thesis is an investigation of the hypothesis that the current accepted measurements that quantify fidelity of an amplifier fail to correlate well with subjective sound quality. Proposed is a more accurate set of measurement protocols."
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.497.2126&rep=rep1&type=pdf
Other good papers are from Geddes and Lee available on their website. THey also come up with the Gedlee metric for evaluating subjective quality based on the distortion of the amp (or whatever I guess).
There is an interesting one from Boyk and Sussmann where they simulate the distortion behavior of tubes, mosfets and bipolar transistors in a number of simple circuit configurations (building blocks of bigger amps). It doesn't so much discuss sound quality but explores distortion and its mechanisms.
Try reading some of my old reviews in Positive feedback and see what you think. I mixed objective observations about WHAT I hear and subjective impressions about how that made me feel about the way it reproduced music.
I even made a speaker measurment in one review (Piega C2 Ltd.)
I was thinking about the SIT-1 and SIT-2 that use only a single transistor per channel...even simpler than a SET!!
I've read your old reviews years ago.
The C2 review illustrates my point well. It's similar to many current equipment reviews/reviewers. Except they don't even bother to tackle some basic measurements as you did."Objective observations of what you hear" ARE subjective evaluations...in my opinion.
How it "made you feel" is an emotional response and you've gone past even subjective evaluation at that point...in my opinion.Well, simple, but not very useful.
I view most of the First Watt amplifiers as proof-of-concept novelties that are not usable with most speaker systems...in my opinion.
The F7 is an interesting evolution though. I would like to check that one out. :)Anyways, do you know Pete Millett? This is an interesting effort:
http://pmillett.com/r120_se_amp.htmlDave.
Edits: 03/09/16 03/09/16
Well I would disagree with you but, hey I am biased about my own critiques you know.
""Objective observations of what you hear" ARE subjective evaluations...in my opinion."
I disagree. You can make objective observations about WHAT you hear that is different with the device under test (DUT). This can be very accurate or not at all, depending on the observer. Perhaps you have heard of organoleptic detection systems for Gas Chromotography (GC)? Do you know what that is? It is a human detector for odors! People can be trained in their sense of smell to accurately detect what molecule is coming off of the GC and at approximately what concentration. It is not subjective and is quite accurate...used heavily in the food and fragrance industries. Now, the sujective part is how you THINK that smell compares to other smells.
Same with audio. You can objectively hear a sharp treble or woolly bass but the subjective part is how you think that affects the realism of the reproduction. Those are two different things, one is objective but it might be inaccurate or imprecise if the listener is not well trained and/or skilled at describing what he/she is hearing. The subjective is how it compares to other gear and/or live unamplified music.
"How it "made you feel" is an emotional response and you've gone past even subjective evaluation at that point...in my opinion."
There is nothing beyond subjective...unless you want to call in the supernatural to this discussion ;-).
The SIT amps make 10 watts. This is useful for a large range of mid to upper 90s db speakers and up. Since the SET revival there are a lot of players in the game now with this kind of sensitivity and some are very very good.
I am not saying that the SITs sound good but they probably have as much chance as any SS amp would, IMO.
You see I am quite convinced now that Push/pull amps, even ones with Class A and no feedback, are fundamentally flawed designs because they do not follow the ear/brains own distortion pattern. If you then make them Class B and add feedback you stray even further from that paradigm.
So far, humans have only developed on kind of amplifier that kind of approximates this ear/brain pattern (Cheever calls it the Aural harmonics) is a well designed SET. BTW, the Aural harmonics and their relative levels are SPL dependent so the sensitivity of the speaker system then plays a role in what kind of pattern SHOULD be generated.
what does this Millet amp put out in terms of power, 5 watts? 10? Not very useful...LOL!
Somehow I just knew you wouldn't agree with my definitions of subjective and objective. :) Do you feel the supernatural is not at work in areas of the audio industry? I believe it is. :)Just thought you might be interested in that R120 amp.....if you hadn't already seen it.
It checks many of your boxes, yes? Not push-pull, no crossover distortion, probably rich with even-order distortion, selectable negative feedback, etc, etc.
That's one of the nicer "DIY" tube-based amplifier designs I've seen in quite a while.Cheers,
Dave.
Edits: 03/10/16
Morri,Still racing your lotus around your circular driveway ... :)
Easy to fool test equipment when bench testing , a class-B amp can be made to flatter on the bench, nothing new or sinister, done it many times before Subjectively, the sound is typically powerful but unrefined with a 2 dim sound stage and very hi-fi sounding.
It's far more difficult to fool the ear ....
Edits: 03/10/16
Yep, just had it out today in fact. Still the fastest around my circular driveway too 😉
Sadly, those 2d, coarse, unrefined sounding amps are still quite common. There are some push/pull amps out there that at least manage to not make a complete dog's breakfast of the sound, but they are mostly Class A triode push pull. Maybe the new Pass XA .8 series is finally equaling the best tubes (Jack Roberts thinks so...he got rid of his beautiful WAVAC for one.). I certainly haven't heard a top SS or even hybrid, although the Ypsilon amps might be an exception) that delivers all the way.
Whatever happened to your friend with the Devialet , did he keep it ...? Does he have the latest firmware ..
You mentioned paladims, are they the big Aragon mono blocks? I had a set and I didn't think they were so hot, sold them to a friend, he stuck them on his Scintilla's and they went up in smoke.
Al
Satie, I compared an xti2000 to my H20 on several different speakers including Acoustat 2+2's, H20 was better in all aspects. jmo to my ears.
Al
I would have expected as much. But it does not fully implement the technology. Not that I would have expected that much more from even those amps that apply the full BCA variant of PWM design, but I was hoping you might have had a chance to experience one, the pro audio reviewers view the BCA amps favorably relative to the linear predecessors.
. .
You are alive Davey,
Hi-bias class-a does bass better than D, so , no feedback alone wont fix the bass..
BTW, you never did answer the speaker question ..
So says you regarding Class-A and Class-D bass.
But that's not what I said anyway. Try reading the post.
Speaker question???
Dave.
You mean this .... ?"Negative feedback being more effective at lower frequencies, it can effectively reduce crossover distortion at lower power levels." -Davey
What amp speaker combination are you currently running..
Edits: 02/26/16
I'm still don't know what speaker questions I missed.
What amp/speaker combination am I currently running? What does that have to do with anything pertaining to this particular topic?? If I'm not using a Class-D amplifier into electrostatics my opinion is not valid? Okay, let me solve that for you. I just hooked my Stax headphones to my Hypex amplifier and it's making nice music. That's what I'm "currently running." Does that properly answer your question???
Hilarious.
Brad and I don't always agree, but at least I know I'm talking to someone who has given thought to his statements.....considered the concepts.....understood the premise. You might try adopting some of his approach.....your current one is pretty pathetic.
Dave.
Hallucinate much , Brad and you never agree , come to think of it , you are on a planet of one there ..
@Satie , it was the xls and I series i tried .....
Actually, you don't have a clue. Brad and I have chatted on/off for quite a few years.....especially years ago on the Apogee forum. For me to say that we disagree 100% would be an incorrect statement.
You have the same statement mischaracterization affliction that Brad does. There's at least two on your planet. :)
Dave.
Projecting your insecurities , its a basic simple question Davey, what are you currently using , since you stepped away from apogees ..
Edits: 02/26/16
Insecurities?
I mentioned my current setup here on the PA just a few weeks back. I haven't changed it since then. If you were paying attention to my posts you would know what I'm currently listening to.
Apogee's haven't been my primary speakers since approximately 1998.
Dave.
LOL, you have issues son ...........
Brad I don't agree on much, but I do take him seriously.Dave.
Edits: 02/26/16 02/26/16
Do you thing the newer "class I" PWM amps from Crown actually eliminate crossover distortin as they claim?
I don't know. Why don't you test one of them and see if you can identify/measure crossover distortion.
Dave.
I don't have a lab at hand and am not on the market to buy one of these costly items (cost of entry is $1k and up from there). The equivalents to my linear MA5002V are $4500 in the MA5000i and $6750 for the IT5000 HD. So not in my ballpark.
The XTI amps don't implement the entire technology and don't make the claims.
You don't actually need a costly setup to evaluate this. A scope and spectrum analyzer will do an adequate job.
You need to evaluate the distortion components and also the "slant" of the crossover distortion at the zero-crossing point.
You really should have some basic test equipment anyways. It comes in very handy for many measurements and practical experiments.
Dave.
I think he meant the price of the amps...but maybe the measurement gear too.
I meant the equipment as well. I don't really need it for anything other than curiosity.
Not the ones i have tried, well i did not bench test one, sonically it was not to my liking, narrow sound stage and bad timbral balance.
It was inexpensive , the end ... :)
Are you sure you tried an "I" (balanced current or BCA) amp? The lowest range is a partial implementation in the XTI series amps. e.g. XTI 1002 ($500) 275@8 ohm
The following are actual implementations
BCA models: K series, CE 4000, CTs 2000 and CTs 3000.
CDI1000 and up ($600) 275@8 ohms
Com Tech CTS series 70V and 100V amps
I Tech, IT HD series (lowest model is the IT 5000 HD > $6k for a 1250W @ 8 ohm)
Macro Tech I series, MA5000i is the lowest model at $4.5k for a 1250W @ 8 ohm amp similar specs to my MA5002VZ now over a decade old and a magnificent performer.
Thanks for the additional info Satie , i will look into trying one of the full "I" versions of the crown....
Only good crown I listened to was the Macro Tecks.
Did you try any of the newer Class I/BCA models?
Agree about the Macro Techs and use one for bass.
Yeah , thanks for clearing it up for us little guys Dave , BTW .... Did you and Graz hug it out yet ..?
'Got any thoughts of your own on the Class-B-exclusive topic?
We have an...understanding. I had no problem with Graeme for a period of time, but then he put himself on my bad side many years ago. Once someone gets there, very few make it back.
Folks who have a commercial interest generally don't take well to probing comments/questions. That happened here with Peter Gunn as well. :)
I have no commercial interests, so am not bothered by this aspect.
Dave.
'Got any thoughts of your own on the Class-B-exclusive topic?⏳
Nope , we're still focused on the obvious and necessary flat FR .... :)
Edits: 02/21/16 02/21/16
Yes, I forgot....speaker impedance magnitude/phase is more important than frequency response.
I need to concentrate on staying focused on the obvious priorities. :)
I know Class-AB amplifiers are popular from a marketing standpoint, but they do have their issues.
Well designed Class-B amplifiers do as well, but they do check a lot of boxes.
Class-D amplifiers, of course, solve all problems.
Dave.
Never said it was , but i do enjoy you chasing the academic obvious and yeah Im sure class-D solves everything....😂😂
Edits: 02/21/16
I never overlook the academic obvious at the expense of a non-academic premise. (A basic concept but it will never fail you.)
Our understanding starts with an academic baseline. If we don't have the chops to understand it that's fine, but we can't ignore it.
The world is full of people who think they know what they're talking about. Especially in the world of audio. :)
Stay objective.
Dave.
Such a pity to waste that much Irony , i have a bridge to repair .... :)
I hope it wasn't an audiophile who designed that bridge. Actually, the fact that it needs repair........ :)
Dave.
Is this how you wore Graz down ,,? :)Your beloved class-D has an output LC bridge ( Filter ) to limit it's bandwidth and to help noise, how is it affected by the loudspeaker impedance, does the sonic' s change when operating between 8 and 1 ohm ... ?
Let me help you out ...:)
Speaker ImpedanceAn important design parameter, which is unfortunately often overlooked, is the impedance of the speaker. Fig. 3 shows the influence of the speaker impedance for a typical Class D filter. It is apparent that this parameter affects both the resonance overshoot and also the phase. As the speaker impedance is an integral part of the filter, this must always be considered in the selection of the filter components. On account of the frequency dependence of the speaker impedance, especially at high frequencies, compensation of the LC output by means of feedback would in fact be necessary. Unfortunately, a sophisticated circuit arrangement would be required here, so that in most designs it is simply omitted at the cost of sound quality.
Another problem with Class D designs is EMC. As previously discussed, a poor layout almost inevitably leads to such problems. But the design of the LC filter is also critical: As can be identified from the Bode plot (Fig. 2), the signal amplitude is attenuated above the cutoff frequency, but it is not fully suppressed. Especially at high power, one runs the risks that the high frequency residual ripple on the speaker input lines is too high from an EMC perspective. Good Class D designs often therefore use higher order passive filters.
Since most speakers are not purely resitive (unless ribbon and then the xover dominates) you can imagine what the phase angles and impedance swings does , maybe not in your world .. :)
Edits: 02/22/16 02/22/16
Actually, Graz wore me down....but that's another topic. :)
You're just copying/pasting information regarding Class-D and regurgitating information from earlier in this thread. Plots generated with textbook loads. That's the kind of thing you give me grief for. :) Is that all you have? Any thoughts of your own on this topic?
I'm interested in data from real-world amplifier(Class-D) / speaker(electrostatic) interactions.
You might have noticed that I queried earlier regarding impedance sweeps of various electrostats above the audio band so we could better understand the Class-D/electrostat interface. Both how the particular switching frequency might interact with real-world electrostats and also how the above-audio-band impedance would interact with the Class-D output filter.
So far, no one has come forward with some actual data. How about measuring/posting some impedance sweeps of your electrostatic system(s) so we can take a look? Said impedance sweeps would tell us a lot about this interface since the above-audio area is the key to understanding this combination....IMO.
The OP's original query/premise is a good one, but I haven't seen anyone really address it objectively.
Morricab's brilliant/incisive reply was "they all sounded like crap." Great, that really helps.
You seem to have inferred Class-D amplifiers are beloved by me. You would have inferred wrong. :) I have multiple amplifiers (and multiple speakers) in my listening room currently. One of them is indeed a Hypex-based setup, but also others (A/AB/B) as well. Two of them even have those glass things that glow orange. :)
Speaking of not answering questions, you ignored my post asking your opinion of Class-B exclusive operation. A little ironic irony there? :)
Dave.
"Morricab's brilliant/incisive reply was "they all sounded like crap." Great, that really helps"
Geez, Davey, what kind of measurements would you like to see that quantifies "sounds like crap". It is clear the interaction with a highly reactive load was reeking havoc with the sound. How to represent that in a way your narrow mind will accept?? I can tell you that I had a Class AB OTL amp that literally oscillated with my Acoustats...does that count as "sounds like crap" to you?? It did to me...by the way these amps were great with a normal less reactive speaker...I imagine a speaker that was a pure resistor would have been even better.
I have an impedance trace of my Acoustat 2200s because it is in the owners manual. It goes from 2 ohms in the bass about 12 ohms and then 2 ohms in the highs. It is the reactive behavior interacting with the feedback loop that is the problem...thus the oscillation of my OTLs, which had around 10db of negative feedback...not even a lot by SS amp standards.
I have no problem with the "crap" subjective evaluation....it just does nothing to help the topic along.Can you measure/plot an impedance trace of your Acoustat 2200's above the audio band? I'd be interested in seeing that.
Dave.
Edits: 02/26/16
Just go on eBay and search for an Acoustat Spectra 2200 manual, it was in there.
As I said, I'm interested in the impedance above the audio band. That's why I asked if you could measure yours.
It's easy to decouple a capacitive load above the audio band to (possibly) provide better operation with Class-D (or other) amplifiers. Did you try that?
Dave.
No and I don't have the speakers anymore, haven't had them for years...
One thing Acoustats do well is beam like crazy at high levels, I also noticed this on the MG2B's as well, powered by various SS amps, for whatever reason the H20 seems to remove the beaming to a huge degree.
Al
Spectras did not really beam at all...that was the beauty of them over the older 1+1, which didn't beam too bad, and 2+2 that beamed pretty badly. The Spectras did successfully improve dispersion electrically that Martin Logans TRIED, only somewhat successfully, to do with curving the panel.
I did not notice my Spectra 2200 or 4400s beaming as the level went up...it stayed pretty constant.
Thats because you weren't running super class-B with tons of feedback, no flat side, zero notch..
That is an unlikely cause since beaming is a dispersion issue and a property of the speaker geometry. In the Spectra 4400 vs the 2+2 the top end is not played by all the panels.
Where you may be right is in there being a phase contribution to beaming - usually off axis, but I don't see how the phase relationship would change that differently between frequency ranges by the amp alone - though severe and badly implemented feedback as is common in SS amps might do so because of propagation delay in semiconductors and the presence of stray reactive components within the feedback loop.
The only class D amps where I am aware of the designer being careful to track and minimize propagation delay and to correct for phase distortion that can result is the Spectron Musician (at least for version III) . In SS amps, Nelson Pass pays attention to this in all his designs (or at least lip service) and particularly with the FirstWatt non-NFB amps. Ayre and Simaudio/Moon also restrict feedback for those reasons in their SS amps. Devialet and NAD in their Masters series class D amps claim to use DSP to compensate for the phase characteristics of their amp sections. The Devialet definitely seems to do so as successfully as do Spectron.
Does the observation of perceived reduction in beaming apply only to the 2+2 or to the MG2B as well?
I recall the 2+2 as beaming about the same at any level as far as the need for "head in a vise" - you find that not to be the case with all amplification?
MG2b's as well
So you answer with more irony,platitudes and conjecture ....
TSK, TSK
Wonderfully informative. Thanks.
Maybe you need some new material? Maybe you could adopt a Paul McGowan Audio Myth approach? :)
Dave.
Davey,Maybe you could actual post up something, anything , instead of the usual drive by trolling, your last time out you left 5 answered questions, anything for the OP, can your Class-D special drive the Acoustat ..?
Did you fix the scinnies..?
Edits: 02/23/16
I missed this one, sorry.
You're using my shtick now! You really DO need some new material!
Dave.
I wasn't aware that there was that much room to adjust bias on these Krells. Do you generally just up the bias for low impedance lads?
Technically speaking you can only have a full class-A amp at 8 ohm and higher , so for most their is no such thing as full class -a( ok talking about DC , SS amps ) when you have a z-min of 4 ohm and lower. How much bias is actually necessary is very much dependent on when the amp slides out of class-a on a particular setup. impedance , sensitivity and typical listening levels play a big role here, if you listen at relatively low levels on a high sensitivity speaker , you will most likely favor an amplifier running high class -a bias into that Zload , with low sensitivity speakers at a moderate to high listening level , not so much..
On a pr of scinnies we lowered the bias considerably on a KSA200 , it takes time, there's no magic number, bias is gradually increased and adjusted for best sonics , at the listening levels required , we ended up with a relatively cool running Krell and speed galore , the dynamic contrast we were able to obtain was fantastic compared to it's original hi bias operation , much more life and energy to the music ...
Good Class-D has this jump factor in spades , SET's too , it is one of the main reasons i keep a close ear to class-d , unfortunately and like SET 's , their ( class-D) sonics are very much impedance dependent and only the Devialets dont seem to have this issue , but they also dont have the high jump factor .
"Technically speaking you can only have a full class-A amp at 8 ohm and higher , so for most their is no such thing as full class -a( o"
You are of course referring to push/pull and not SET :).
The Sumo "Nine" amp was 60 watts Class A into 8 ohms and 120 watts Class A into 4 ohms...or so the manual claimed...they said they were the only amp to do so... :).
I don't quite get your point about "jump factor" what exactly is it that you think is causing this perceptual difference??
Nelson Pass talks allot about the portion of the transistor S curve you want to be working in and for lower signal levels he selects very particular transistors, some of which are custom made for him in small batches, e.g. JFETs.
Look at his white papers. I just don't remember which ones discussed this directly. It is part of what drove the philosophy of the First Watt designs.
You get that jump factor when you operate at the low end of the S curve as you come close to the asymptote at the center of the S. Pass is not in favor of operating there but does. The class B operation of his amps generally starts above that if I understood it correctly.
So what you are saying is that you want to operate in the "more than linear" portion of the output transistors' IV S curve - so you are lowering the bias on low Zload so that portion is where you operate.
At Z <3ohms the DR-9 ran hot on increased bias while it was just warm with even higher bias into 5 ohms. So that was limiting.
I agree about the Devialet, very impressive. Try the Spectron Musician III and on. .
John ,
I was surprised to see the KSA thread still active at DIY , are any of the pinkmouse boards still available ? Do you still have the KSA250 ? They are not biased as high as the KSA 200 which idled at 14 amps , i think the 250's were biased at 7 amps ...
Yes, quite active. On a recent poll of DIY type amps, the KSA-50 and 100 rated very high (top two / three?) in popularity on the DIYAudio site.
You can post a WTB on the swap meet and I bet you will find the pink boards. Although Jan's boards are fine too. Jans may have ability to change from Class A to Class AB (bias switch) so you can somewhat control heat dissipation.
I have another set of pinks all populated and ready to drop into a KSA-300 Chassis that I bought after original owner blew it up. It came to me totally scorched. I sandblasted all parts and anodized it black. I have the 1kva tranny, 200muf caps - everything in the chassis but life keeps getting in the way of finishing. Kinda overkill for the KSA-50, but plan on adding additional output devices later for slightly more power, but will be getting close to current limits on original design...
Yes, the KSA-200 was far better designed (nearly all in class A) than the KSA-250 - the KSA-250 really was the first sliding bias design. Like I said, I liked the KSA-250, just liked the KSA-50 clone better, so I sold it (it went rather quickly). Always have an eye out for a KSA-200 tho...
I VOID WARRANTIES
72watts class A at 4 ohm, thats 144 watts class-A at 8ohm , how is that possible from a Class-A 50...?
Its a KSA-50 clone I built (google pinkmouse diyaudio ksa50 for more info), with three pairs output devices (instead of the two pair as original) and I set the idle bias at 1A per output device.Rail voltage is +/- 36VDC. Emitter resistors are .68 ohm. Voltage across emitter resistors (for idle bias) is 680mv, and 3 pairs output devices.
So 1A idle bias per device, and 3A "total bias" per channel. This gives 72wpc @ 4 ohm, 144 wpc @ 8 ohm.
Also dissipates 432 watts from heatsinks (both channels). Makes for a great sounding amp and decent wintertime space heater.
I VOID WARRANTIES
Edits: 02/19/16
Another pic ....
I VOID WARRANTIES
Ok , searched they are different from Jim's ....
Jan's PCB was the other one - I chose pink's as Jans pcb had some errors and needed traces modified...pink's were perfect.
I VOID WARRANTIES
Pretty stout emitter resistors, :)
OK i know the DIY project , yours is different from Jan's original done some years ago, his used the early metal can type outputs as the original.
Is yours the same as the kit offered by Jims ..?
Edits: 02/19/16 02/19/16
I tried Spectron Musician III to drive my Sound Lab Majestic 945 and I didn't like it. Now I am bi-amping the 945. The Musician is used as bass amp. I don't advise to use any d amp to drive any speakers. D amp is not there yet.
I haven't heard all the Class D amps that exist so I can only comment on my SX1000 Mono Blocs driving my Maggie 3.6R's. I have power out the wazoo and play a wide range of music Classical, Jazz, Easy Listening & 60's Folk with great results.
I can only say these Class D's are great for listening with my Maggies.
Here's my take on Class D. There is a problem with Class D amps and that is the output filter (usually a 2nd order Butterworth) relies on the load impedance. With a Magnepan speaker which has a relatively flat output impedance curve, they work great. However speakers that have drop in impedance with increasing frequency, which is typical of electrostatics, may lead to disappointment. The manufacturer can design feedback to minimize the effect of the impedance load but probably most do not and they shoot for a target impedance. As a result the performance of a Class D with an 8 ohm speaker will be different than with a 4 ohm speaker. You may want to look at Rod Elliott's web page for a further discussion.
Edits: 02/11/16
I rewired a pair of magnepan tympani tweeters. I measured them with an ohmmeter, they gave me a reading of m19 ohms?, is something wrong?
I think you have a typo
Should be either 2 ohm or 3 ohm (2.75)
Yeah, it's a good observation, but I think it would be informative to look at impedance sweeps of a few electrostatic speakers above the audio band.....preferably above 100khz.
Electrostats are capacitors, but not very good ones. I believe other parasitic effects might swamp the capacitive characteristic at the frequencies of interest here.
Dave.
don't do it. I had Class D amps (three different types) and they all sounded like crap on my Acoustats (1+1, Spectra 2200 and Spectra 4400).
morricab, I sent you an e-mail, but I didn't realize you had replied here to my question about Class D.
Have you any experience with the Acoustat amps (TNT-120 in particular) driving Acoustat speakers? I am currently using one on my Model 3s and am pretty please (actually pleased as punch) with the sound, but I always wonder what more can be had. Would tubes be a noticeable improvement, in your opinion?
My next tweak is going to be QRD diffusers on the front wall directly in line with the rear wave of the speakers and bass traps, so we'll see how that goes, but I am always interested in even more of that holographic sound stage.
Great amps for the speaker. That is how I spent my two weeks with the 2+2. + a sub.
used the 200 version of the TNT.
Amp can use some parts upgrades but a solid design and a perfect match. Good tube amps can do better.
I use a H20-S250 built by Henry Ho, it's classD, it's better than my Krell ksa 100 on 2+2's more tuby sound.
I remember him as an early adopter of the technology. Had great reviews in the day.
Class D brought the first real world amps that could drive a 1 ohm Scintie.
I am always on watch for ESL users who like class D amps on their speakers as their resolution and complex load tend to trip up so many otherwise fine amps. Do you perceive any "white powdery grain" at the top octave on the S250?
A guy I know from Holland had the Signature monos from H2O and while they do indeed drive Scintillas (the guy who owned them...and several other Dutch guys I know had or have Scintillas) they didn't sound particularly good doing so. We compared them to a Sphinx Project 16 and a pair of bridged Classe DR3s. The Sphinx was tops, the Classe not too far behind and the H2Os were dragging far behind in 3rd sonically.
A DR3 can drive a 1 ohm load bridged ..?The Class-D amplifiers i have tried had timbral issues at 1 ohm, frankly it was unlistenable on classical music , had plenty of drive thou , i can see someone liking them if they listen to mostly electronic music . As to driving low-z loads all amplifiers sound different when driving loads lower than 8/4 ohm, higher THD levels for sure ...
@OP
What speakers were the NuForce being used on ..? I would try them anyway , there are other factors at play too , one of noise , your system/setup may or maynot be sensitive to it ..
Edits: 02/17/16
Yes, a DR3 can drive 1 ohm bridged...at least I didn't see any smoke coming out of them ;-).
It wasn't just timbral issues that we experienced it was also spatial issues. As you know the Scintillas can throw a very DEEP soundstage...especially when far away from a front wall (in this case more than 3 meters). The H20s sounded pretty flat, both images and soundstage depth...it was really disappointing. I also heard some homebuilt Hypex based amps on Scintillas back in Switzerland and it was also horrible.
I would expect just that, but it is also a question of power vs. SQ, the classD has practically unlimited power if you want it. I would, you wouldn't..
The question is what is the nature of the differences in SQ characteristics between the class A SS classics and the H2O class D's.
Brad, I have a Krell ksa 100, I ran it on my Scinys, It has a good reputation behind it, known as one of Krells best, I'll take the H20 over it in bass and midrange,I also tried a Spinx project 18, sent it back!!.
Al
The Project 18 is an all SS amp (the PJ16 and PJ14 were hybrids)and is not nearly as good as the Project 16. A friend of mine bought an 18 for his Divas and we found it to be a decent SS amp but no more than that...how it would behave at 1 ohm I do not know but it might not be capable enough for such a load.
The Project 16 was an all out assault for a hybrid design. It has 4 power transformers for the output stage and a separate power supply for the tube stage. The power suppply is switchable between 8, 4 and 2 ohms, which i believe means that it changes the voltage and current delivery characteristics depending on what you select (it had a front panel button). They only made a handful of these amps (less than 20).
Another friend had an original KSA100 that he kept out of nostalgia as much as anything and we we try it out on his Scintillas but his LAMM M1.1s were better as was his Musical Fidelity KW750 stereo amp. The best sound we got from his Scintillas was with a tube amp specially designed by Allen Wright of Vacuumstate to drive 1 ohm loads. This was very nice sounding and allowed the best overall development i have heard from those speakers to date...with the possible exception of the PJ16.
I can only tell you what I heard that day in Holland but I can tell you it was unanimous and that the guy who owned the H2Os then promtply sold them.
It would be easier to get to Mars than find a PJ16 ..... Is the PJ16 the best SS /hybrid amp you have ever ( MF amplifiers?) heard , well , since the Lamm is now off your list .. :)
That will drive a Scintilla? Yes. Otherwise, I have heard some others that were quite good as well.
Gryphon ..?
Never tried one out? You?
Yes, very good as is FMA and Spectral ....
Edits: 02/19/16
My own misgivings over the Nuforce 8.5 was at the top end - driving the tweeters. It was not quite right and had a whitish fine grain particularly obvious on triangles vibraphone and cymbals left to ring in the air. Also bothersome on brass instruments.
Is that something you heard on the H2O compared to class A SS?
I don't know the Project Sphinx but for reputation, I like the DR3 (more so the VHC version) allot and had a DR9 for most of a decade. I prefer either of the latter Classe's to the KSA series, but perhaps for the 50.
Satie, To my ears it misses the ambience that the krell has in the top end, the final finishing edge of sound as decay, funny you mentioned triangles , I made a 4channel tape of triangles tings postitioned and marked on wall, the only amps I have that got it right were the H20, and several Carvers. The krell gave me 3 tings back, the Bryston 7bs monos were all in the wrong places. lol .
Al
When you say the Krell gave you 3 tings back, does that mean 3 in the right places or one ting was missing completely? Just curious as it sounds like a good experiment.
I made some recordings of my ex-girlfriend playing Paganini caprices while standing in between my (now gone) Acoustat 1+1 speakers. I set up the microphones and recorded it with my (now gone) TEAC R2R machine. You could easily tell the way she was standing and relative motions when played back through the Acoustats but not so well with other speakers/systems.
My 2+2s circa '03 sounded their best with VTL MB450s:
The Acoustat TNT amps were specifically designed to drive the Model and Spectra's. They sound great doing it. I wouldn't mess with success, but that's me.
Although I've never heard a D driving an Acoustat, one of the criticisms I've read of D is that they can sound sterile. Given the ruthlessly revealing nature of Acoustats relative to what precedes them in the chain, it could potentially be a bad pairing.
The Acoustats are like a microscope. They will tell you exactly what your amplifier sounds like. What amp are you using now?
I don't think Acoustats are that tough a load. They demand quality not necessarily quantity.
If you do try the Nuforce, please report back.
You might want to also post on the Nuprime Audio Circle site. Jason Lim the founder of Nuforce is very active there and might have some good insight. Best of luck!
For a while I drove my Maggie 3.6's with class D. Sounded very good.
I don't see why it won't work. Depends if you like the sound of the class D amp. See if your friend will let you try it out
Alan
I don't see why not. Try and see for yourself.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: