The UULE exercise can be done with any TV with HDMI: you have a source, transport, that passes a digital signal to an integrated playback unit. A high end version of same is a Meridian source unit feeding one of their DSP speaker models -- then it is purely an exercise of subjectively assessing what is being done right, and what is being done wrong ...
The Meridian version failed, and failed badly when I heard it in a showroom ...
The Meridian version failed, and failed badly when I heard it in a showroom ...
I wonder if Meridian are still using the same (common) power supply for all amps in their active setup? A pic of a very old active speaker model revealed they were doing this. To me the primary advantage of active is easing the demands on the amps and on their PSUs which means optimizing the PSU impedance for each drive unit's band of interest.
Yes, I suspect they were making significant key mistakes like that in their topology up to recent times; the model I heard was that prior to their current lineup; the latest iteration got a very good write up in HiFi Critic fairly recently, so they must be learning.
Why the UULE can work is because everything is extremely simple, only a simple full range speaker to worry about, the bass content goes nowhere. The subjective treble quality is remarkably good, I had on some poor quality 1950 recordings of Maria Callas yesterday, and it did them very reasonably.
Why the UULE can work is because everything is extremely simple, only a simple full range speaker to worry about, the bass content goes nowhere. The subjective treble quality is remarkably good, I had on some poor quality 1950 recordings of Maria Callas yesterday, and it did them very reasonably.
its not all reject stuff, for instance a couple of weeks ago I got a 1kg bag of Lavazza coffee there for $20 (normally ~35) as well as some nice vacuum packed pork belly. if you pick carefully you can find good stuff, but there is a lot of budget junk and weird foodstuffs too, like Chinese copies of products where they tried to make a western flavor of their own, but just completely miss the mark, with odd chemical tastes, or strange combinations. and some Atlantic Salmon I got around Xmas was truly horrible (not an easy thing to do) and although sold as prepared, was only partly scaled and had bones all through it
as Frank said they also have a scattering of various other household goods, but you wouldnt catch me dead buying electronics from there. I hardly buy any finished audio hardware these days, all scratch built.
WTH is this preference for electronics that masks bad recordings? while claiming you want realism? you say it like its a good thing. bad recordings should sound bad, they shouldnt sound needlessly bad or over etched, like some audiophile gear does to make it sound more immediately impressive and detailed in the short term, but a system should not turn water into wine. Claims of a system making bad recordings less so, that is also true to source with high quality recordings, like the electronics has some sort of method for discerning the quality and making decisions....are fantasy.
as Frank said they also have a scattering of various other household goods, but you wouldnt catch me dead buying electronics from there. I hardly buy any finished audio hardware these days, all scratch built.
WTH is this preference for electronics that masks bad recordings? while claiming you want realism? you say it like its a good thing. bad recordings should sound bad, they shouldnt sound needlessly bad or over etched, like some audiophile gear does to make it sound more immediately impressive and detailed in the short term, but a system should not turn water into wine. Claims of a system making bad recordings less so, that is also true to source with high quality recordings, like the electronics has some sort of method for discerning the quality and making decisions....are fantasy.
Last edited:
Yes, Aldi does do duds too, but they have 60 days no-questions-asked return policy, so you're able to buy safely.
The thing is, masking does not place -- I've heard plenty of tube systems that do that, the recent hifi show had many of those; and also heard in the context of home high end systems -- and there is another road from those you mentioned. Why only a small number of people know this other way is because it's normally very difficult to achieve - and of course it shouldn't be so!
I only recently realised how important the psychoacoustic factor is, this is key to getting the results - a certain quality level has to be reached, and then your mind does the rest of the work, effortlessly. It separates the musical content, the recording, from the remaining distortion without conscious focusing on doing so. And even on "poor" recordings enough information has been captured for your mind to do this; it really does seem like magic when this happens, recordings one has always dismissed as impossible rubbish spring to life, you can "hear" that there are real musicians in there who feel about what they're playing, just as much as on an excellent recording.
Edit: when I found the best room at the Sydney hifi show, I knew straight away that this was capable: it was playing a track from Gracelands and all the indicators were there, that I was so familiar with. The 1000W Brystons and Dynaudio C3s were firing, so I asked for some meat to go on. A drum solo, perfect, the kick drum punched with tremendous impact and bite, the cymbals shimmered and soared, the sound tore through your body like the real thing does - it was a invigorating, inspiring experience ...
The thing is, masking does not place -- I've heard plenty of tube systems that do that, the recent hifi show had many of those; and also heard in the context of home high end systems -- and there is another road from those you mentioned. Why only a small number of people know this other way is because it's normally very difficult to achieve - and of course it shouldn't be so!
I only recently realised how important the psychoacoustic factor is, this is key to getting the results - a certain quality level has to be reached, and then your mind does the rest of the work, effortlessly. It separates the musical content, the recording, from the remaining distortion without conscious focusing on doing so. And even on "poor" recordings enough information has been captured for your mind to do this; it really does seem like magic when this happens, recordings one has always dismissed as impossible rubbish spring to life, you can "hear" that there are real musicians in there who feel about what they're playing, just as much as on an excellent recording.
Edit: when I found the best room at the Sydney hifi show, I knew straight away that this was capable: it was playing a track from Gracelands and all the indicators were there, that I was so familiar with. The 1000W Brystons and Dynaudio C3s were firing, so I asked for some meat to go on. A drum solo, perfect, the kick drum punched with tremendous impact and bite, the cymbals shimmered and soared, the sound tore through your body like the real thing does - it was a invigorating, inspiring experience ...
Last edited:
I must take issue with this, Frank.
My own approach to designing starts from the output stage. I first determine what I want out of it, how low impedance will I go, will I watch steady state or impulse bursts, etc.
For example, a prefectly good amp rated at 100 WRMS/8 Ohms can be made with two pairs of MJL 3281/1302. As some manufacturers have so adequately demonstrated (e.g. Harman/Kardon, several models, including prestige line Citation), even with two pairs of output devices, you could be able to dump well over 500 Watts into 2 Ohms in bursts.
However, I believe that to be a little too tight, so I will use three pairs, thus letting my amp be able to work even steady state into 3 Ohms and a bit less. But I honestly have no need for more than 3 pairs, as I don't expect the amp to be used in full power steady state regime into 2 Ohms or less. But, chances are it will be capable of delivering over 600 Watts into 2 Ohms in short bursts. Which I am not at all likely of actually ever using in real life.
But, like I said - better to have and not use, than to use and not have.
Obviously, I believe in as good a load tolerance as I can squeeze out of it. I will not get it unless my output stage is very capable, and of course, unless my power supplies can swing it. I like big custom toroids, I like parallelled capacitors and plenty of them.
On the other hand, in part I agree with you, there are units out there with ridiculous numbers of output devices which sometimes serve for show rather than function.
Sorry, dvv, not having a go at any particular design technique, merely those who believe that one must use certain approaches to achieve meritorious sound, that there are no other methods that will do the job. The key phrase I used in my piece was "if they don't perform the task required" -- if the unit does the job well, achieves very high SQ, then there are no issues, all is well.
The dartZeel philosophy is obviously an example where the ideas of a single pair of output devices are followed through religiously, with considerable acclaim -- there are many roads to Rome ...
The dartZeel philosophy is obviously an example where the ideas of a single pair of output devices are followed through religiously, with considerable acclaim -- there are many roads to Rome ...
yet again, just making **** up as you go along. doesnt even seem to be relevant and it sure isnt based on any information whatsoever, youre just spouting nonsense because your views are under dispute. I have 14wpc on tweeters (6Ω), ~125-150W on mid-bass 8Ω (or 250 with less bias, but mostly running lower again for more class A) and when I finish the subs 4Ω, 250-300wpc available for them. digital crossover. 100% DIY, no bling heatsinks I dont need, no preamp, no crossover parts bling (well, apart from having 3x as many dacs and amps as strictly nessecary )
)
Last edited:
Sorry, dvv, not having a go at any particular design technique, merely those who believe that one must use certain approaches to achieve meritorious sound, that there are no other methods that will do the job. The key phrase I used in my piece was "if they don't perform the task required" -- if the unit does the job well, achieves very high SQ, then there are no issues, all is well.
The dartZeel philosophy is obviously an example where the ideas of a single pair of output devices are followed through religiously, with considerable acclaim -- there are many roads to Rome ...
Didn' think you did, Frank, just a point I thought worth mentioning.
Generally, I agree with you - there are many roads leading to Rome indeed.
What are the tasks put before an amp is, of course, a different thing. Some are happy with modest tasks and objectives, others are far more stringent, and I say - to each his own.
My belief is that an should generally act as a true voltage source, with only small deviations from the absolute; I believe most users would be quite happy with that. Obviously, for that to happen, one must in all cases, without fail, have high quality power supplies, or all bets are off. But there again, there are several possible solutions, and it's the designer's job to choose one he believes to be the best overall for his amp and to implement it properly.
I wouldn't be caught dead without a rather stiff power supply, yet Armstrong's (now gone British company) last effort, a 200WRMS amp, used soft (slack) power supplies with aplomb. Nige and I are great fans of that company, and trust me, with good reason. They had an outstanding design team, whoever the gentlemen were, because cheap or expensive, they never failed to deliver as much of the essence of music as possible for the money. Nige can correct me if he feels he should, but Armstrong's 30+30 WRMS receiver from the late 70ies must have been one of the best sounding receivers I have ever heard.
Anyway, no misunderstning, Frank, just a point I made.
yet again, just making **** up as you go along. doesnt even seem to be relevant and it sure isnt based on any information whatsoever, youre just spouting nonsense because your views are under dispute. I have 14wpc on tweeters (6Ω), ~125-150W on mid-bass 8Ω (or 250 with less bias, but mostly running lower again for more class A) and when I finish the subs 4Ω, 250-300wpc available for them. digital crossover. 100% DIY, no bling heatsinks I dont need, no preamp, no crossover parts bling (well, apart from having 3x as many dacs and amps as strictly nessecary
Way to go, Qusp! Bling does nothing or your music.
well, I confess to the use of copper-teflon coupling Caps (I get OEM pricing) on the ESS transconductance stages (a Pass D1 common gate variant by OPC, with no buffer and higher voltage rails to push up gm), they use Zfoil TO220Z power resistors and tight tolerance thin film SMD.
I do have a bit of a parts fetish (particularly high end SMD) and problems with power supply overkill, but even though expensive, the resistors above and anal retentive fet matching bring real improvements in distortion and CMRR. thats my excuse anyway. the amps though, the money has gone towards tight tolerance SMD parts, good ALFET/Semelab fets and Audiopower DPS600 500W regulated power supplies. some number chasing there too, but as i've said elsewhere, the hobby, for me has to be intellectually satisfying too. I dont generally go for flowery magic parts, but do insist on higher performance than is likely to be audible. I'm making this for me, for the long-haul, so if I spend more than is strictly necessary, So be it. I have my eyes open.
output impedances on all the amps are in the very hard to measure category. conrad heatsinks and the floor take care of all the heat. the IV stages put out 50W of heat per stereo channel, so those have got pretty serious heatsinks for a dac. the IV stage and analogue dac/clock supplies are the major discrete SS stages, so I did go to town somewhat in those more delicate areas where it matters to get close to datasheet spec on the ES9012 dacs
I do have a bit of a parts fetish (particularly high end SMD) and problems with power supply overkill, but even though expensive, the resistors above and anal retentive fet matching bring real improvements in distortion and CMRR. thats my excuse anyway. the amps though, the money has gone towards tight tolerance SMD parts, good ALFET/Semelab fets and Audiopower DPS600 500W regulated power supplies. some number chasing there too, but as i've said elsewhere, the hobby, for me has to be intellectually satisfying too. I dont generally go for flowery magic parts, but do insist on higher performance than is likely to be audible. I'm making this for me, for the long-haul, so if I spend more than is strictly necessary, So be it. I have my eyes open.
output impedances on all the amps are in the very hard to measure category. conrad heatsinks and the floor take care of all the heat. the IV stages put out 50W of heat per stereo channel, so those have got pretty serious heatsinks for a dac. the IV stage and analogue dac/clock supplies are the major discrete SS stages, so I did go to town somewhat in those more delicate areas where it matters to get close to datasheet spec on the ES9012 dacs
Last edited:
I have always been reluctant to use copper directly with Teflon because of corrosion problems. e.g.: http://www.residues.com/pdfs/foresite_teflon.pdf , http://nepp.nasa.gov/docuploads/16F...3CFE54D64/RED PLAGUE mtrls protection pub.pdf , Most Mil-Spec Teflon wire has silver plating to resist this corrosion. Judging from the article even the silver plating doesn't always help.
There are other audiophile cap combinations that are similarly questionable. At the prices of these objects I would want stable operation for a very long time. However the corrosion may be adding interesting effects?
Haven't seen anyone who designs anything without his own quirks.
I see nothing wrong or unusual in going to town with what you are working with; as you said, we do this for ourselves and are not limited by economic concerns as the industry is. Besides, it's FUN to spend money on quality parts you want to use.
I see nothing wrong or unusual in going to town with what you are working with; as you said, we do this for ourselves and are not limited by economic concerns as the industry is. Besides, it's FUN to spend money on quality parts you want to use.
no problem with mine over 2 years, I use copper and teflon wire for output, as well as all other hookup right through to the speakers and have since I began DIY, so it just seemed natural and as a bonus he matches them tightly for me. I make no magical claims, it needs caps on the output, there is a single fet and cap in the signal path on each phase of the dac IV stage for ~ -115db THD+N measured performance, so I went to town a bit on these few components and I make no apology for it. the dac is also on rogers teflon substrate, so its rather consistent in that regard.
the caps are expensive for sure, but he does make them inhouse and my experience with his caps over the years are very positive for build quality and consistency.
if you think there is room for pleasant coloration at -115dB THD+N Demain, well hey maybe there is!! haha there is a certain satisfaction from getting exceptionally high performance (not bad even for opamps) from a handful of discrete parts. the performance is due to very high transconductance fairchild fets being selected and then matched for VDS and gm in a replica circuit. The resulting circuit is run at +/-45vdc, where the performance from higher gm stops improving; all verified with an AP2.
the whole thing is a marriage of objective measurement and design, partnered with a few idiosyncrasies for fun.
oh yeah and it sounds brilliant, but it would wouldnt it?
the caps are expensive for sure, but he does make them inhouse and my experience with his caps over the years are very positive for build quality and consistency.
if you think there is room for pleasant coloration at -115dB THD+N Demain, well hey maybe there is!! haha there is a certain satisfaction from getting exceptionally high performance (not bad even for opamps) from a handful of discrete parts. the performance is due to very high transconductance fairchild fets being selected and then matched for VDS and gm in a replica circuit. The resulting circuit is run at +/-45vdc, where the performance from higher gm stops improving; all verified with an AP2.
the whole thing is a marriage of objective measurement and design, partnered with a few idiosyncrasies for fun.
oh yeah and it sounds brilliant, but it would wouldnt it?
Last edited:
An externally hosted image should be here but it was not working when we last tested it.
got it in 1! trawling the net and parts catalogs for parts, constantly trying to learn more with each step, making things attractive where possible as well (i'm a graphic designer and chef by trade in years gone by) is all part of the fun.
at the moment i'm spending a lot of time on research and prototyping common mode and differential filters for my power supplies and the inputs/outputs for a dac/headphone amp i'm working on, using SMD X2Y caps, resistor networks and common mode chokes to maximize rejection, multilayer PCBs with designed in faraday shields etc etc basically EMC to cope with todays wide bandwidth and RFI.
especially since the bias switching and automation for the whole system will include Xbee wifi modules for monitor/control/power sequencing and display, so I need to make sure those frequencies are rejected and cellphone chatter in my headphones (and speakers to a lesser degree) is a pet peeve, that isnt handled well by that many commercial designs either.
we are working on giving the MCU (beaglebone) access to environmental temp (local) as well as weather forcast via web access, so that the amps are in low bias for start up and sleep as well as a summer and winter setting to maximize performance given environmental conditions and non-infinite heatsinks
that would be fine if it was the psychoaccoustics that are usually given credit, instead of talk about the hardware, which is Franks usual position. what would be the point of testing/assessing different hardware systems and tweaks for this quality if its an internal psychological effect?
or are you trying to maximize the psychological effect by making subjectively chosen and tested parts changes and involving youself more in this process Frank? you cant really make claims of both.
Last edited:
Yep... but it looks good for someone usually making claims about hardware and now looking inside his own brain... perhaps it is starting a change
- Status
- Not open for further replies.