Put it this way: in recent years every time I experienced "good" sound that sustained its quality and I was in the position to, or was interested enough at the time to investigate further, I've found that the volume control didn't use mechanical wiper mechanisms ...
Like all parts, there's good and bad, AND there's a hell of a lot in getting the implementation right. If you use an electronic control and then add a micro processor to do the twiddling, then you're potentially opening up another can of worms - audio ain't easy!
...Frank
Last edited:
In the spirit of the "measurement" part of this thread, can you offer anything technically, even a hypothesis, why this might be so?Put it this way: in recent years every time I experienced "good" sound that sustained its quality and I was in the position to, or was interested enough at the time to investigate further, I've found that the volume control didn't use mechanical wiper mechanisms ...
I can imagine the mechanical connection acting something like a diode, and/or I've read such a claim. I met a guy 20 years ago, back when Audio Amateur had the "POOGE" articles, who bypassed his receivers' A-B speaker switch and soldered his speaker cables directly to the amplifier's outputs.
There's nothing wrong with micrprocessors that can't be "fixed" with proper coding. The processor should normally be in sleep mode, without its clock or anything else running. Turning a front-panel quadrature switch should generate an interrupt that wakes it up to normal running state, changes the volume appropriately, waits a half second or so, writes the current setting to eeprom, then goes back to sleep. This way there's nothing "running" that could make a signal to be picked up by the audio circuitry.Like all parts, there's good and bad, AND there's a hell of a lot in getting the implementation right. If you use an electronic control and then add a micro processor to do the twiddling, then you're potentially opening up another can of worms - audio ain't easy!
Frank
Always a good idea.
I once was part of a geographically-distributed design team, with the firmware people lacking a command of English (and forget about my Mandarin). I got a prototype in which they had defeated a bipolar supply for the integrated level control IC, and wondered why.
It turned out that they were continuously refreshing the IC
from the uC and apparently using its readback capability to figure out what the level setting was. So they stripped away 6dB of range to be able to use the bus bidirectionally. Now of course there was more than ample memory elsewhere to keep track.I was of course blamed for not knowing how the part worked.
You can't schedule stupidity, as Dilbert says.
Put it this way: in recent years every time I experienced "good" sound that sustained its quality and I was in the position to, or was interested enough at the time to investigate further, I've found that the volume control didn't use mechanical wiper mechanisms ...
Like all parts, there's good and bad, AND there's a hell of a lot in getting the implementation right. If you use an electronic control and then add a micro processor to do the twiddling, then you're potentially opening up another can of worms - audio ain't easy!
Frank
Frank, what you are saying makes sense to me, but is unfortunately totally beyond my real world capabilities. Remember, electronics are not my trade, it's a hobby, and the only part of electronics I can safely say I know well is the eeny weeny itsy bitsy part regarding my filters.
Digitally, I'm a complete dumbell.
If, for example, I needed a digitally controlle input switching system, I'd have to ask somebody to do it for me. The best I can do is use a mechanical switch to turn relays on and off.
My first redeeming value is that I do not try to pretend to know what I don't know well or at all. And I am reminded of what I don't know well enough each and every day, whenever I run into a problem.
My second and last redeeming value is that I am not afraid to ask those who know more than I. The only intellgence of mine involved in that process is that I have met and cultivated a lot of different people, in different fields, so it's rare that I have to search long and wide for an answer.
I have a degree in economics, majoring in foreign trade and keeping track of the systems theory for my own pleasure of knowing. Since the systems theory is in fact social sciences' version of electronics, with many common points, this was helpful to me in my hobby.
As you can see, your suggestions are all well and fine, but way beyond me. And I intensly dislike dabbling in something I am unclear about, I have a need to really know what I am doing. I try to keep my ego in check, so, for eample, when I set about building my own tailor made car from the ground up, it was natural for me to assemble a team of friends, each of which excelled in one field or another. As a result, that car works way better than its factory counterparts and has never once broken down. And when it eventually does, it's likely to be classic wear and tear rather than a malfunction.
However, if you should get down to it, and if you're willing to share, I am certainly willing to give it a go.
Your thoughts mirror my own; I know micro-diodes and such are considered bad form by many in these parts, but everything over the years has sent a clear message to me that subtle levels of non-linearity in areas such as non-"perfect" metal to metal and similar contact are enough to introduce audible distortion, very conceivably by low-level rectification. I have not the access to equipment that can measure such behaviour, but I do know that every time I effectively bypass simple mechanical connections I improve the sound.
Agreed, this is all solvable. I did it using completely discrete logic, as an exercise, and was happy with the results.
Frank
Me too. Computing was my real game, and where I got a buzz from being highly analytical. Cooked my brain doing this, I'm afraid, so I have to do things these days in a very measured, and non-stressful way ...
I find it hard to believe that someone hasn't posted a DIY version of such. The way I've done things is very much in the prototype mode, never as a finished, polished unit; my interest is in understanding what the factors are that influence sound quality, and these days how "low class" you can go in the kit and still deliver impressive sound.
What I find intriguing is that the "big" answers aren't known by anyone, it seems. And what I mean by that is that if you hand any individual a selection of nominally well behaved gear that they will achieve satisfying sound from it, there's no magic lookup cheat sheet that guarantees success every time.
The best that I've achieved is an understanding that exceptional performance is possible, if persistence is applied. And that excellence is achievable at relatively low cost, if one is prepared to do things "one's own way".
I'm impressed!! Sounds just like how I'd do my own audio, if I had the means ...
Happy to give it a go - but fair warning, I run out of energy easily, a result of over-stressing the grey matter ...
Cheers,
Frank
Carbon, cermet and conductive plastic pots all measure very low harmonic distortion on the bench. They may well have low frequency distortion problems due to high temperature coefficients, but swamped by LF speaker distortion.
Far more likely they have problems with (sound) vibration affecting contact pressure and modulating the wiper position. The contact is not a point, so various parts of the contact zone must be at different tap positions.
Far more likely they have problems with (sound) vibration affecting contact pressure and modulating the wiper position. The contact is not a point, so various parts of the contact zone must be at different tap positions.
...
Happy to give it a go - but fair warning, I run out of energy easily, a result of over-stressing the grey matter ...
Cheers,
Frank
Since you have no obligations, you can't really fail, Frank. You can get there or not, in which case I'll just go on using what I already have.
Obviously, I'd prefer you to get there eventually.
If you'd like a few shots of what I did with the car, just drop me a line at dvv@beograd.com .
Well i guess we have to get ole D into music first , then we can discuss sound..
@Gootee ,
Tom whats this mystery amp putting out 1k into 1ohm ...?
Adcom GFA-585.
I wanted to see what it would take to make the Maggies distort. The 585's 400W into 4 Ohms rating enabled it to do that, without lighting up its clipping indicators.
I also learned that my Vandersteen 2Ce speakers have red lights behind the socks, which start to flash when there's an overtemp condition. <grin>
Last edited:
I have a Pr and yes they will drive 1 ohm , but i could never wrap myself around its sonics , favoring to use The Threshold S500 , PS Audio 200cx or the Krell Ksa200..
The adcom does have alot of drive and did see off the A21 in My setup , a friend now has one (a21) powering his 3.6 , we have not done any comparisons in his setup as yet ....
The adcom does have alot of drive and did see off the A21 in My setup , a friend now has one (a21) powering his 3.6 , we have not done any comparisons in his setup as yet ....
Last edited:
I am basically still just dabbling with lower-cost used power amps and speakers, as I get back into consumer audio. I also plan to end up with three or four complete systems that I can give to my grown children. I acquired a small armada of Adcom 545II, 535II, two of each, and Hafler 220, and half a dozen late 70s-80s HK and Luxman, which were all incredibly low-priced. I'm still sampling speakers, theoretically. But after the Magnepans arrived, the sense of urgency for sampling others was diminished greatly.
With all of these amps, I am planning to try building some line-level active crossovers to experiment with bi-amping and tri-amping.
With all of these amps, I am planning to try building some line-level active crossovers to experiment with bi-amping and tri-amping.
Last edited:
I would say just about everyone in audio has an agenda or two, some of which are tied up with ego,
. One of my personal agendas is to get the most out of every recording I have, and everything is sacrified, ultimately, to that goal.If I were in your shoes, how I would approach this situation is as follows: someone reckons conventional volume controls are a problem, so before I do anything else I would confirm whether such is the case for me; in the other words, whether I can hear this as a problem. And how I would do it is like this: get my best setup, find what a good, average volume is for assessing my best test recordings. Then measure what the resistances are for the pot at that setting, and swap out the pot with fixed value, decent quality resistors. Yes, eliminate the volume control! For a short while, anyway ... Obviously, listen like crazy, see what you think; and then reverse the swap, go back to the pot, and listen again. If you hear a clear difference then your path is set, most likely ...
When doing the listening I would do it in a relaxed, not highly focused way ... which way sounds more "right"?
The recordings I use to test with are key, the best I find are those what are considered bad by most people: heavily processed, lots of artifical echo, excessive high frequency content - these are very hard going on systems with problems, give the game away immediately. Personally, a conventional volume control renders these sort of recordings almost impossible to listen to, that's how bad the distortion effect is for me.
Frank
I've come across a few descriptions of what audio working properly should sound like - rare, usually only achieved with very expensive gear and lots of costly twiddling and optimising. This is not necessary, but is a technique that will work at the moment, with the current thinking of most people.
A taste of what it sounds like for some: State of the Art in D.C. | The Absolute Sound
Frank
A taste of what it sounds like for some: State of the Art in D.C. | The Absolute Sound
Frank
I have a Pr and yes they will drive 1 ohm , but i could never wrap myself around its sonics , favoring to use The Threshold S500 , PS Audio 200cx or the Krell Ksa200..
The adcom does have alot of drive and did see off the A21 in My setup , a friend now has one (a21) powering his 3.6 , we have not done any comparisons in his setup as yet ....
Personally, I think that the Adcoms that I have are very, very good, and the 585 is the best of them, by a small but noticeable margin (and the huge power capability increases that margin).
Maybe I should mention again that my goal is the most-faithful reproduction of the source (whether I like "the sound of it" or not).
The Maggies and the Vandersteen 2Ce and some Thiels that I have are all fairly transparent and revealing, when kept within a reasonable range of operating conditions (which turns out to be fairly wide, usually). And I must have been very lucky with my exact room configuration because the accuracy is spellbinding, and the sound is exquisite, and simply gorgeous; everything is much better, maybe twice as good, the way I have it arranged now, when compared to arranging things differently, in the same room.
Anyway, I compared quite a few amplifiers and receivers (maybe 12 or more?), over the last two years. (This might be meaningless, or perhaps laughable, to those who know the truly high-end equipment, which I do not. But anyone who wants a very large quality/cost ratio, and relatively-low cost, might appreciate it.) Truthfully, I loved them all, but, of course, I had only purchased models that I thought would be excellent. However, the HK 560 (circa 1978), a Luxman 1050 (and I think also the 1120), and the Adcom GFA-585, seemed to have some extra magic to them, although the Adcom GFA-545II was also very, very good. (And except for the 585, they were all under $200, I think, and some were much closer to $100. I figured that even if I used only the cabinets and transformers, for DIY projects, the prices would not be bad.)
My son, who is now 22, also loves music. He used my systems, daily, but, by chance, almost never when I was around, because our schedules differed. We didn't talk about the systems, for many months. But one day we did, and without me mentioning which ones I favored, he named the same three or four that I named above. (I'm pretty sure that he "borrowed" one of the Luxmans, when he moved out of here, even though I had already given him a Hafler DH-220.)
I am still looking for a big pair of Magnepan speakers, probably the MG-3.6/R, hopefully within a half day's drive.
I still think that the decoupling and reservoir capacitors are the main key to the sound reproduction's accuracy and imaging, since they are the actual "signal path", and the high-current signal path is probably a little more unforgiving than the small-signal path. And I would guess that not enough attention has been paid to the details of the cap-current signal path, so far.
I have ben dabbling at developing methods for analyzing and designing decoupling capacitor configurations, and also reservoir cap and power/ground rail configurations. But I suspect that's the wrong paradigm. The caps and their impedance at the output devices ARE the amplifier, or at least they bound its capabilities.
Therefore, my next DIY project will probably be a power amplifier (maybe even just a power output stage), centered around Terry Given-style capacitor-array power/ground-plane boards, and the details of the design of their connections to the power output devices could be a major focus area.
Maybe something like this: Large arrays of large capacitors on large unbroken copper planes (maybe 15x15 1000 uF array per rail), an un-etched two-sided PCB with power on one side and ground on the other, one for each rail (or one for both rails, with two half-planes on one side, for power), just one drill hole per cap with copper removed around edge, NO traces, all copper; "lots of current, under pressure, with almost no constrictions"; about 0.5 nH or less inductance seen by power device connections, with less than 10 mOhms total impedance seen at output devices, maybe significantly less.
Output circuits might be on a small PCB mounted parallel to the back side of the cap array board, only a few millimeters away, and over the area where the two power planes almost touch. Power and ground connections will be able to go directly to the planes, with almost-insignificant connection lengths. No other reservoir or decoupling capacitors will be necessary; power supply and decoupling caps "all in one". To really do it right, I might eventually have to mount a whole amp assembly right on each speaker driver's shortest-possible connections.
Last edited:
I'm p*ssed off at the moment, did a decent reply to you, Tom, and it went down the browser gurgler. In essence, I like your thinking here, my DIY chip amp used a lot of these ideas, successfully ...
That "wide range of operating conditions" will improve, that's the main benefit, to be able to go from an almost inaudible whisper to deafening, live music levels with tonality intact -- the amp being fed from healthy power rails is key to making this happen ...
Frank
Been THERE. I feel for you. Now I always right-click, select all, and copy, before posting. And if it's more than a half-hour's work I usually pop open Notepad or Wordpad, to make a safety copy.
Go listen to your reference track, for a while. I just did and I feel better for you already. <grin>
Last edited:
It wasn't even like that, I only wanted to click the Italic button and it was "just out of reach" and somehow the manouvering managed to trigger a refresh ... ya just give up at times ...
Anyway, one of the things I mentioned before I was "rudely interrupted" was to think carefully about separating the decoupling into "dirty" and "clean" areas. The dirty, at some distance from the output circuitry, has the "grunt", capacity to handle the bass needs, is where the big current spikes occur; and the clean area, absolutely on top of the amplifying parts, has the quality devices to handle the high frequency transients. That 10 m.ohms or better is a good target ...
Frank
Tom, sorry to do this, but I was curious about the Adcom amp you mentioned, found the service manual for the GFA-585, and it doesn't appear to have the grunt you mentioned. 400W into 4 ohms, as a continuous rating, is the best the manufacturer offers.
On the power supply side you can see why, only 36,000uF smoothing for each rail; significantly more would be needed for continuous power into heavier loads ...
,
Frank
On the power supply side you can see why, only 36,000uF smoothing for each rail; significantly more would be needed for continuous power into heavier loads ...
,Frank
- Status
- Not open for further replies.