The thing I don't understand is that it takes almost as much effort to mess about with Aldi TV sets as to, say, build your own version of a Meridian DSP speaker from low cost off-the-shelf parts, or experiment with the design concepts of the Grimm LS1, or build a phased array column speaker. Basically something radical you've never heard in your own home before, and you wonder just what it could sound like. This stuff needn't be expensive.
If I'd heard all the possible variations of speaker configuration and the capabilities of DSP and found that they were all the same and boring, then perhaps I couldn't argue with people's desire to tinker in tiny little audio cul-de-sacs, but I don't believe that many people have that much experience of what's possible. It seems premature to declare that vinyl, valves and two-way ported passive speakers are the ultimate, and that new cables (or turning off all RF devices in the house and using a supermarket TV as our amp/speakers) are our only passport to the audio future.
If I'd heard all the possible variations of speaker configuration and the capabilities of DSP and found that they were all the same and boring, then perhaps I couldn't argue with people's desire to tinker in tiny little audio cul-de-sacs, but I don't believe that many people have that much experience of what's possible. It seems premature to declare that vinyl, valves and two-way ported passive speakers are the ultimate, and that new cables (or turning off all RF devices in the house and using a supermarket TV as our amp/speakers) are our only passport to the audio future.
We're really pretty close to being on the same page in terms of how we approach getting a system to work properly, those last couple of posts demonstrated that; where we differ strongly is in how to then test the results of that effort.
I don't have access to the test equipment to take the measurements, hence have to use the next best thing, my ears. And reading what the outcomes are for people who avidly use such, without then also closely subjectively evaluating, I don't feel anything significantly important has been missed.
What I listen for, the only thing that matters, is low level distortion. This is extremely easy to pick up using ears only, so every change is based on whether that area improves or not. An AP set up in exactly the right way would also most likely pick it up, I'm just using a freebie to do it instead!
The psychoacoustics come into it because, obviously, some recordings have more intrinsic distortion than others, the key is to absolutely minimise the disturbing distortion generated by your own system during playback, meaning the ear/brain then only has to deal with the abberations of the recording - personally, I've found that's good enough to make the playback work convincingly on "poor" recordings ...
Last edited:
The TV experiment demonstrates even more obviously that one of the killer problems these days is interference effects - this kills the quality stone dead, it goes from "I want to keep listening all day" to "5 minutes of this, I've had it, turn it off now!". The easy bit is demonstrating the effect; the hard bit is then eliminating the problem occurring!
The HT fiddling was a lot to do with attenuating this intrusion, not 100% perfect, but mighty good considering the intrinsic quality of the base unit.
So the future to me is working out everything that's necessary to do to kill these interference issues stone dead; the ideal is have the system such that then a really nasty RF and mains noise generator can be placed electrically adjacent to the audio components without having any audible effect. Once you've got that under control, then aim to be able to wind up the volume to any level you wish without audible deterioration of sound.
Last edited:
First, are you sure that it's not in your mind? If I had a notion that RF was affecting my hi fi or brain directly, I might find it difficult to suppress my 'expectation bias'. Your neighbours' mobile phones, wi fi and microwave ovens may account for a high level of RF you can't control. You may literally end up wearing a tinfoil hat.
Second, do you know that it is an absence of RF devices that cures the problem and not simply turning off the SMPS wall-wart that powers your wi-fi router?
sorry meant to reply to this, maybe I personalized the compensation comment, given it was in a post that was part of a back and forth.
also, I dont have an AP2 (I wish), the IV stage is not my design, its OPCs like a number of the PCBs (but in all cases built from PCB), its beyond my capability to measure except with long FFT, so cable routing may well have lost some of that -~115db performance that he measured. but the same combination of DAC->IV without the zfoil power resistors gets there. measuring anything at that level is hard and I simply dont have the coin to justify it. i'm building an ADC and linestage + mic-pre that should get me pretty darn close combined with software, without the convenience of dedicated hardware.
as ive said elsewhere, only a few of the bits are fully my design and PCB design is a relatively new addition that came pretty naturally given my pre-press and 2D and 3D vector design experience. the system integration design is mine and everything i'm building now is my design, but I dont want to give the illusion that its all my design. I need to thank Ian for his Fifo and Acko for his DAC PCBs
Last edited:
Yes, because I've done variations on testing for this for years now, my wife can pick it easily - "the violins are sounding scratchy again!!" - I have friends who have tuned into it, and who use a strict procedure for shutting down problem items.
If you have neighbours who are literally adjacent to you, on the other side of an adjoining wall, then, yes, you most likely will have difficulties testing in a "clean" environment. Fortunately, I have many 10's of yards between me and others for all the periods while I've been serious about this, making it easy to test for. Depending on how strong, and the nature of the interference level, just moving the culprit down to the other end of the house may be sufficient.
There is a very easy, more "technical" test - put on a recording that has a tone or instruments that is prone to sounding unpleasantly scratchy at times, listen to the sound as close to one of the tweeters as you feel comfortable doing, compare clean and 'dirty' situations ...
The interference effects are cumulative, every device adds just a bit more 'dirt' to the sound; as you remove each offender the sound gets cleaner, smoother, easier to listen to.
If you have neighbours who are literally adjacent to you, on the other side of an adjoining wall, then, yes, you most likely will have difficulties testing in a "clean" environment. Fortunately, I have many 10's of yards between me and others for all the periods while I've been serious about this, making it easy to test for. Depending on how strong, and the nature of the interference level, just moving the culprit down to the other end of the house may be sufficient.
There is a very easy, more "technical" test - put on a recording that has a tone or instruments that is prone to sounding unpleasantly scratchy at times, listen to the sound as close to one of the tweeters as you feel comfortable doing, compare clean and 'dirty' situations ...
The interference effects are cumulative, every device adds just a bit more 'dirt' to the sound; as you remove each offender the sound gets cleaner, smoother, easier to listen to.
So can you demonstrate the RF is affecting the electronics, and not affecting your ears/brain directly? Have you had a string quartet play a few numbers in the house just to make sure they didn't sound scratchy when your wi fi router was switched on?
Oh dear, watch out for exploding heads on the street from mobile phone overload!!
Well, when I add circuitry and measures that only affect the ability of the audio system to resist interference the sound improves. Without turning off the offender. Seems pretty straightforward to me ...
Edit: live music is doomed! Everyone has mobile phones, which are one of the worst of the worst -- the sound for people in the best seats in the concert hall is now kaput!!
Well, when I add circuitry and measures that only affect the ability of the audio system to resist interference the sound improves. Without turning off the offender. Seems pretty straightforward to me ...
Edit: live music is doomed! Everyone has mobile phones, which are one of the worst of the worst -- the sound for people in the best seats in the concert hall is now kaput!!
Last edited:
The only time RF affects my brain is when I have been working inside a TV transmitter building. These give us headaches. The field strength from these is above safety limits and you do not hang around. The public never goes near these field levels
Of interest, on the long running http://www.diyaudio.com/forums/vendors-bazaar/190434-hypex-ncore.html thread, just now talking about 'improving' the sound of the 'best' class D amplifiers, designed by a pretty straight up and down objectivist, by adding conditioners and power cord fiddles. These are classic ways of trying to attenuate interference, with very audible results from the sound of it ...
coppertop, RF from GSM and various other spill, being rectified by bipolar junctions in opamp input/output stages and various other devices, then folded down as DC error, modulating the voltage across PCB traces and planes etc is a very real problem that will only get worse with the advent of power over wifi among other advances.
Made in a recording studio with your radio in mind, you're the audience, it's a live performance.
For hi-fi, you might record it with microphones in your room, and then play it somewhere else, so you can fuss over quality of reproduction.
That cumulative RF interfearence IS a problem is hardly news. However, I would remind one and all that beisde looking at the phenomenon as such, you also need to remember the relative POWER of cell phones, wireless phones, TV and radio waves, etc.
A nice home grown test is to activate your phone (cell or wireles) and bring it near to your audio. If nothing untoward happens, your problem is small. If any audible interference does occur, then you do have a pronounced problem.
Especially nice is leaning the phone right on to your MC input, or your tuner's front end (puting it on the case right above the front end). Again, if nothing untoward happens, you're all right.
More prone to this sort of thing are the wild bandwidth amplifiers, espeially those of yore, such as say Sansui or Kenwood/Trio models from the late 70ies, which were allowed to do 500 kHz unhindered. Mind you, I also advocate wide bandwidth, do make it go above 300 kHz for good phase response, but do install a band limiting filter at its input, as people like reVox and Sony usually do. Sony even advertises internal and external slew rate, i.e. without and with the band limiting input filter.
Lastly, this is one of the reasons why one must pay a great deal of attention to as good decoupling as possible of the PSU lines. Typically, three unit, obe larger cap of say 100 uF, a smaller around 4,7-1 uF and lastly a 100 nF ceramic cap. In respect to the price of the whole shebang, this isn't really a great expenditure, and it certainly cannot hurt.
A nice home grown test is to activate your phone (cell or wireles) and bring it near to your audio. If nothing untoward happens, your problem is small. If any audible interference does occur, then you do have a pronounced problem.
Especially nice is leaning the phone right on to your MC input, or your tuner's front end (puting it on the case right above the front end). Again, if nothing untoward happens, you're all right.
More prone to this sort of thing are the wild bandwidth amplifiers, espeially those of yore, such as say Sansui or Kenwood/Trio models from the late 70ies, which were allowed to do 500 kHz unhindered. Mind you, I also advocate wide bandwidth, do make it go above 300 kHz for good phase response, but do install a band limiting filter at its input, as people like reVox and Sony usually do. Sony even advertises internal and external slew rate, i.e. without and with the band limiting input filter.
Lastly, this is one of the reasons why one must pay a great deal of attention to as good decoupling as possible of the PSU lines. Typically, three unit, obe larger cap of say 100 uF, a smaller around 4,7-1 uF and lastly a 100 nF ceramic cap. In respect to the price of the whole shebang, this isn't really a great expenditure, and it certainly cannot hurt.
Simple . Great turntables . Simple audio chain . Good amp . Sometimes 10 kHz bandwidth .
I even get good sound on my Sony SW ( ST 3950 ) . No way is it 2.5 kHz as some say .
Naim even claimed long range AM ( Australia ? ) was best ever sound . There was even stereo I was told with a +/- 20 Hz dead band to give a centre reference ( double SSB stereo = mono ) . 10 kHz top end . On researching this there is no limitation placed on HF ( especially long wave ) . The 4.5 kHz stated is typical . Distortion would be about like SE valve amps . Noise 50 dB on a good day .
Most of that music sounds great on a Garrard 401 ( Lenco 75 ) . Goldring 1042 or better . NAD 3020 as pre amp ( not bad ) . DIY power amp . Speaker to your taste . CD players , yet to hear one I like .
Last edited:
I renewed my acquaintance with Quad ESL 63 yesterday . Smaller than I remembered . I took a 200 mile round trip to Nottingham to try out my SE valve amp . So good to hear a speaker that tells the truth . The truth was very acceptable and plenty loud enough ( too loud to talk ) . The great advantage being that no feedback amps are fine with the Quads as is 8 watts ! Bass is very good and as I remembered . As my friend commented when correct the Quads do not sound dull . The amps below 1 % distortion up to 5 watts and mostly second harmonic . Just below clipping the harmonics are in suspension bridge reduction . The frequency response 15 Hz to 47 kHz - 3 dB and 20 Hz to 20 kHz -0.7 db ( related to 1 kHz and 8R ) . The amp is optimized for 6R .
On CD the total system distortion will be less than 1 % typical . At times 0.1 % Hiss and hum inaudible . It didn't make me a total convert to valves . I would run a very discrete subwoofer from 15 to 100 Hz with transistors . I like the hard incisive quality of a big paper cone speaker and transistor amp . Something like Holliman perhaps ?
On CD the total system distortion will be less than 1 % typical . At times 0.1 % Hiss and hum inaudible . It didn't make me a total convert to valves . I would run a very discrete subwoofer from 15 to 100 Hz with transistors . I like the hard incisive quality of a big paper cone speaker and transistor amp . Something like Holliman perhaps ?
agree completely and thats how I design too, except for the obligatory and somewhat arbitrary decoupling scheme, which will not touch (well not much) higher power common mode GSM location chirps. the above is not news to me, rather what i'm working to improve on. its possible to improve rejection of common mode by up to 30-40dB over the above with combination common mode/differential filters for example x2y, as I mentioned
Last edited:
quite soon is planned wireless solution to wired standards, it works ~60ghz
people will be beamed not only by wifi at home
Those Jamaicans you know are into "sound Systems" not HiFi , different genre to those into HiFi , the ones into HiFi are as serious as any ...
I bought into serious decoupling with Otala's texts. He used, as the last before the output trannies, a seris resitor of 1 Ohm and xxx nF capacitor to the ground, naturally on all power amp supply lines.
At first, I didn't think too much of it, but later on, it tickled my imaginaton. So I did some experimenting. To cut a long story short, I found that virtually all power amps could benefit from it, some more, some less. Since it really gets rid of residual inductance from the PSU (non reguilated, just apssive caps), I was not really surprised to note that low quality amps, with low quality power supplies, benefitted the most.
The trick is in the tuning. I'm sure there are ways to calculate it fairly precisely, but that's boring. Much more fun listening for it, start with a very low value like 220 nF and work your way up. Anyway, there aren't that many steps up to 680 nF which is the largest value I ever used. You know, try 220 nF, then the first next value and so forth for as long as you can hear any difference, even if minimal. Once you cannot, go back a step, and that's it. A wonderful excuse to tell your good wife why you simply must run the system at a power level hgher than usual.
And odd things can happen. With Toshiba 2SC5200/2SA1943 trannies, rated at 150W, once I couls ghear some "birdies", literally like chirping birds. Very low level, but it was unmistakably there. On a whim, I took the Toshes out and replaced them with Motorola/ON Semi MJL 3281/1302. Nominally, a very similar transistor, both being based on Toshiba's 2SC3281/2SA1302. Just replaced them one on one.
Chirping gone. But that's only half of it. I now had a smoother (so to speak) midrange and better treble, more life-like. Since everything was the same in both cases, I was a bit caught out with that. So I checked and double schecked, got in some well hearing friends, and we all agreed, the difference was small, but it definitely WAS there.
From their Data Sheets, I note Toshiba says collector output capacitrance is "200 pF TYPICAL", while Motorola says "600 pF MAXIMUM". So I picked up both lots and went to see some friends with an excellent lab nearby. About 10 pairs in each lot. The AVERAGE value for Toshiba was 320 pF and for Motorola 390 pF. That's close enough on two output pairs that I seriously doubt that was the reason for any differences.
What we must never forget is that while two nominally same trannies from two sources may well satisfy all Data Sheet requirements, but they will always be products of two different technologies, quite simply no two manufacturers have the same technological process, so it stands to reason that their products will be at least slightly different - whether that's enough to cause audible differences we cannot know any other way but to try them both.
While I suspect the effect will vary in different circuits and in its extent, I submit that these small, subtle differences WILL produce some differences in sound. As such, I feel this is a variable which also needs to be taken into account during designing.
- Status
- Not open for further replies.