Constant input impedance by using a stepped ladder attenuator with and a 3u Au ELMA rotary switch is a decent solution. Depending on resistor type chosen, it can be low cost or high cost.
Attached is a low cost version I made a few years ago. Takes about one hour to assemble.
Sigurd
Attached is a low cost version I made a few years ago. Takes about one hour to assemble.
Sigurd
PMA said:
Attachments
I am using the same balanced attenuator concept. A 5k TX2575 in series with HOT+ and COLD- and then a switched resistor between the nude resistors.
I am using the red PRP resistor in the Seiden switch. Not the perfect solution, but I do want to try some different components.
Sigurd
I am using the red PRP resistor in the Seiden switch. Not the perfect solution, but I do want to try some different components.
Sigurd
audiowolf said:
For balanced signals, we have used common mode cancellation attenuation with no connection to ground whatsoever. This entails a build out resistor, usually about 5K in series with each signal, followed by a switched variable resistance between the 'far' end of the 2 build out resistors. This is not a constant impedance method, but it does work fine. One could with more resources, use a 3 deck switch and create a constant impedance network, and even at max output open the cross connection for ultra low insertion loss. This system would require that the signal series resistors be changed to increase in value as the common-mode value decreases.....It might be noisier than the non- constant impedance version and have other trade-offs.
Woof!
Attachments
I have used them in my classA poweramp and compared to a TX2352 version a collegue built, and I find that the red PRP:s have less detail and also a softer sound. Bass is not as firm and contoured. Note, that I use a trim pot to adjust the DC offset, and for the PRP version a standard cermet trim pot was used but the the TX2352 version of my collegue, a bulk metal foil trim pot was used.
Sigurd
Sigurd
PH104 said:
Attachments
PMA said:
Thanks for the heads-up on this...It was a long shot of an idea I got from surfing around some of the other threads.....Opto's have been around a long time in pro-audio compressors, but the idea in those boxes IS non-linearity, and aging???? Well in those applications, everything is on pots and knob twiddled to death by 'producer/engineers'...
Methinks the reports on the sound of various resistors both brand and model, and by general materials and construction is highly constructive. I am glad to see some comments from recent experiences, on new or currently available resistors.
In some of my tube stuff I have been using TO220 high wattage parts from Caddock, MP800 series in lower ohmage ranges needing higher watts, they are a non-inductive thick film type, and sound OK compared to large metal films or wire-wounds with a ton of epoxy on the wire coils For lower wattage and higher ohmage, the nude Vishay are an old standby, but I am always interested in finding out something new.
Fred Forssell, who builds potted discrete op-amps and complete components for the pro-recording world, [ http://www.forsselltech.com ] is now using all SMD resistors in tantalum of some sort, and reports the best sound he has ever had from them...Others report that Tants may have higher noise floor....perhaps dithering the signal somehow???? With leaded tants not being made any more, and some of us still needing wires on the parts once in while, I have imagined soldering copper wire to larger SMD parts to create a new class of wire leaded parts..sort of homemade I guess...
Other favs from years past for me are the Holcos, if they are non magnetic end caps versions, and of course some of the Resistas have done OK too. I have not found a magic part that no-one knows about....and have generally found that the better parts that others have chosen for high grade audio are generally good, and the basic sonic signatures are pretty well know, so that sometimes one can choose to use one or another to tune a circuit as needed...to add a bit of quickness, or to soften it slightly.
I recently bought some tantalum film resistors that are no longer made from one of the DIY exotic parts distributors, Geeze, I don't quickly recall the brand or model, to build up a set of line-level gain make up amps for use in a Stereo to Middle-Side de-compiler/ re-compiler, using a set of Fred Forssells 993 discrete op-amps. I have the proto nest of parts made with sockets for the discrete op-amp, so I will be able to go back and forth between the piles of parts to see the specific sonic signature of the tants VS the cheap NTE generic metals, which also have [ yuk] magnetic end caps....no PC board, just hardwired parts in air in both cases.
For general proto work, just to get things working, I admit to using the cheap and locally stocked NTE blue 'flameproof' metal films. I can keep a LOT of these parts around to find the right values that work in a circuit, then rebuild with better -read "expensive"- parts once the specifics are known. I have several other jobs to do ahead of this so it might be a month or 3 before I get around to the evaluation....Its a good thing that work gets in the way of the fun stuff as it pays the bills, but a bad thing that it displaces functional learning experiments.
In general, I prefer any resistor that has no magnetic aspect to it, when compared to one that has steel or other magnetic parts...I don't need to pick them up with an electromagnet, and the magnetic aspect seems to really screw with the transparency, getting worse with increased current. It would be useful to create sort of a table or chart of different resistor models, with the general perceived sound of each of them. I don't know many who can try them all, but collectively there should be some good experience to share.
In case someone wants to know, the Stereo MS box de-compiler/re-compiler is designed to convert an existing stereo signal into the equivalent of the output of an M-S mic array. You get the middle information in one channel, and the side and width information on the other..Its sort of funny because both the L and R signal are in that side channel signal....You can then adjust the relative levels and then re-assemble back into stereo with a conventional M-S decoder...You can adjust the sound stage width and depth this way. You can also export one or both of the component channels to EQ or some other process if you need too, like a compressor, de-esser, or whatever.... In case somebody mixed a center vocal too hot, or you need to de-ess a snare on the side without messing with the sound of the stuff in the middle you just convert the stereo into M-S and do the process, then convert back into stereo. It is an old re-mastering engineers trick, something they didn't teach at most recording technology schools, and one of the tools that the real pros would use to fix up a crappy stereo mix and amaze the clients. I have some tapes coming in a few months from now for archive transfer. They are all poorly balanced for speaker listening....mixed in headcans I would guess, and definitely not bi-naural--another different problem to deal with once in a while....
Anyway, I am building the device that will allow me to readjust the sound staging on those tapes. In pro sound and recording, a lot of stuff goes on that would make audiophiles cringe, but a few really do care how the recordings sound, whether they be minimalist live acoustic, or studio multi-channel. Sadly, most get caught up in using 'plug-ins' in their DAW software to huge excess and tend to view the end result as simply "product" for "consumers" I still mix the hard way, in analog in real time, and use DSD instead of tape. Sort of an "old meets new" way of doing things. Alas, the recording, restoration, and transfer part of of the business is only a tiny fraction of what we do, but I have the MOST fun doing it!
Woof again!
Sigurd, your attenuators look very nice indeed! Nice to see good work, and this thread IS all about the really good stuff.
And yes, Tek scope front ends are great place to learn about wide band attenuators. I remember with no great fondness a time about 30 years ago, calibrating a vertical plug in for one of my 7000 series Teks that I had purchased used for cheap....Unfortunately, someone had messed with the attenuator bank, and EVERTHING there interacts with everything else. Took me a day of cussing and tuning to get it right, but I never forgot what I leaned hands on about the trimming process...
I live near Tek and still know many of the engineers who designed a lot of that great stuff back in the Tektronix hey-days of the 60s and 70s. Sadly a lot of them are well past retirement and dying off, taking a HUGE mass of some of the very best analog engineering knowledge with them...
Would you mind sharing?...Which PRPs are you using? They have 9 models in metal film thru hole, one model in metal oxide, and 2 in wire wound. ,
http://www.prpinc.com/
The "audio specific" parts are here...http://www.prpinc.com/pdf/Audio_PR9372_Series.pdf
Are the PRPs that you use these 9372s? Just wondering, but who knows why they are 'softer'....Hey, sometimes you REALLY NEED a touch of softness, and it is good to know what to choose if you do. Otherwise the Vishay TX2352s seem to be pretty neutral for me in the lower wattage ratings, and at a tolerable price from Texas Components. I should order up some of those PRPs and give them a listen. I was not joking about the need for a touch of softness once in a while. It is indeed a serious part of component tuning.
Again, good looking attenuators!
Woof!
And yes, Tek scope front ends are great place to learn about wide band attenuators. I remember with no great fondness a time about 30 years ago, calibrating a vertical plug in for one of my 7000 series Teks that I had purchased used for cheap....Unfortunately, someone had messed with the attenuator bank, and EVERTHING there interacts with everything else. Took me a day of cussing and tuning to get it right, but I never forgot what I leaned hands on about the trimming process...
I live near Tek and still know many of the engineers who designed a lot of that great stuff back in the Tektronix hey-days of the 60s and 70s. Sadly a lot of them are well past retirement and dying off, taking a HUGE mass of some of the very best analog engineering knowledge with them...
Would you mind sharing?...Which PRPs are you using? They have 9 models in metal film thru hole, one model in metal oxide, and 2 in wire wound. ,
http://www.prpinc.com/
The "audio specific" parts are here...http://www.prpinc.com/pdf/Audio_PR9372_Series.pdf
Are the PRPs that you use these 9372s? Just wondering, but who knows why they are 'softer'....Hey, sometimes you REALLY NEED a touch of softness, and it is good to know what to choose if you do. Otherwise the Vishay TX2352s seem to be pretty neutral for me in the lower wattage ratings, and at a tolerable price from Texas Components. I should order up some of those PRPs and give them a listen. I was not joking about the need for a touch of softness once in a while. It is indeed a serious part of component tuning.
Again, good looking attenuators!
Woof!
Bob Cordell said:
I have had very good results with Cirrus Logic's CS3318 .
These are 8 channel (there is also a 2 channel version) analog level controls so linear that it is futile to try to measure any non-linearity.
In a sense, they are the best of two worlds: stepped attenuators, but integrated with an output buffer. Dynamic range 127dB, THD -112dB, step sizes down to 0.25dB mean that it just has a 'feel' like a continuous pot.
My amps and preams that use this part have invariably been well received, but that is anecdotal. No controlled tests, I'm afraid.
Jan Didden
Thanks!
Yes, it is the PRP 9372 that I use. They are actually co-designed by Charles Hansen of Ayre, who uses them extensively as far as I have seen. I think that CH mentioned that he uses onl one 9372 type that has a given TCR, and that only the 9372 with this special TCR is the best one. Not sure which TCR that is. I use the ones with TCR = 50 when possible.
The PRP:s can be bought at for ex Percy Audio and PartsConnexion.
Nothing out there has yet beaten the TX2352 resistors. But I have not tested the Z201 resistors. The cost of those cross my cost limit.
I have tried the tantalum film SMD resistors in a stepped relay attenuator, but those are not for me. Changing all the resistors in this attenator to TX2352 really lifts sonics. At least if you like highly transparent sound with nothing added nor subtracted.
I found the tantalum film SMD resistors to be a bit lame and mellow.
Sigurd
Yes, it is the PRP 9372 that I use. They are actually co-designed by Charles Hansen of Ayre, who uses them extensively as far as I have seen. I think that CH mentioned that he uses onl one 9372 type that has a given TCR, and that only the 9372 with this special TCR is the best one. Not sure which TCR that is. I use the ones with TCR = 50 when possible.
The PRP:s can be bought at for ex Percy Audio and PartsConnexion.
Nothing out there has yet beaten the TX2352 resistors. But I have not tested the Z201 resistors. The cost of those cross my cost limit.
I have tried the tantalum film SMD resistors in a stepped relay attenuator, but those are not for me. Changing all the resistors in this attenator to TX2352 really lifts sonics. At least if you like highly transparent sound with nothing added nor subtracted.
I found the tantalum film SMD resistors to be a bit lame and mellow.
Sigurd
audiowolf said:
Ears.
Hi,
audiowolf's comments particularly in post #13758 are interesting, and I also always 'pop' my ears before any serious listening-trials.
Having 'listened' attentively to sonic differences in components for nearly 40 yrs now, I have picked up a few tips along the way, mostly entirely by accident.
The 'ear popping' realisation resulted from the aftermath of a head cold when one ear was affected more than the other, and the soundstage shifted from centre towards one side, rather puzzlingly. Popping ears (mine, anyway) even when there are no obvious problems is worth aprox +2dB increase in amplitude subjectively across the board, with slighty greater effects noticed at HF. More importantly, there is an increased resolution throughout which is not realised in the same manner by increasing the overall volume settings in my set-up.
Alcohol is another problem where serious listening is concerned, and I discovered this when 'voicing' passive speaker crossovers many years ago. After a long day experimenting with alternative crossover components, I would relax very late in the evening with a few beers for an extended listening-evaluation of the day's work, when the mains pollution was lower after most folk had gone to bed.
Several times during these late-night sessions I made some changes to components because the overall sound was apparently lacking in extension (and amplitude) subjectively at HF, and the overall sonic balance was not as it should have been.
Next morning, when listening again the result almost 'screamed' at me in comparison with an overly bright balance of sound, and I immediately reduced the HF 'enhancements' I had added late the previous night.
Not believing in so many apparent coincidences, I specifically tested the effects of alcohol during a day-time listening-session, and within a minute or so of taking the first few sips I could hear a reduction (relatively) in the extreme HF taking place, and within 30 mins. the sound was very much 'softer' overall, and somewhat less clear throughout. Minute details which to me make the performances far more realistic, and which I knew should be heard on familiar recordings, just disappeared into the more mushy background.
Maybe this is why many people joke about their systems sounding better after a few drinks, because for certain any harsher HF effects are reduced (at least in my experience), as I have been aware of for many years now.
It is known that alcohol deadens other senses, so this realisation shouldn't have been any great surprise to me I suppose, but although it may help with enjoyment of ordinary listening-sessions, I stay well clear of any alcohol if serious judgements are needed regarding any subjective sounds.
Regards,
Hi,
audiowolf's comments particularly in post #13758 are interesting, and I also always 'pop' my ears before any serious listening-trials.
Having 'listened' attentively to sonic differences in components for nearly 40 yrs now, I have picked up a few tips along the way, mostly entirely by accident.
The 'ear popping' realisation resulted from the aftermath of a head cold when one ear was affected more than the other, and the soundstage shifted from centre towards one side, rather puzzlingly. Popping ears (mine, anyway) even when there are no obvious problems is worth aprox +2dB increase in amplitude subjectively across the board, with slighty greater effects noticed at HF. More importantly, there is an increased resolution throughout which is not realised in the same manner by increasing the overall volume settings in my set-up.
Alcohol is another problem where serious listening is concerned, and I discovered this when 'voicing' passive speaker crossovers many years ago. After a long day experimenting with alternative crossover components, I would relax very late in the evening with a few beers for an extended listening-evaluation of the day's work, when the mains pollution was lower after most folk had gone to bed.
Several times during these late-night sessions I made some changes to components because the overall sound was apparently lacking in extension (and amplitude) subjectively at HF, and the overall sonic balance was not as it should have been.
Next morning, when listening again the result almost 'screamed' at me in comparison with an overly bright balance of sound, and I immediately reduced the HF 'enhancements' I had added late the previous night.
Not believing in so many apparent coincidences, I specifically tested the effects of alcohol during a day-time listening-session, and within a minute or so of taking the first few sips I could hear a reduction (relatively) in the extreme HF taking place, and within 30 mins. the sound was very much 'softer' overall, and somewhat less clear throughout. Minute details which to me make the performances far more realistic, and which I knew should be heard on familiar recordings, just disappeared into the more mushy background.
Maybe this is why many people joke about their systems sounding better after a few drinks, because for certain any harsher HF effects are reduced (at least in my experience), as I have been aware of for many years now.
It is known that alcohol deadens other senses, so this realisation shouldn't have been any great surprise to me I suppose, but although it may help with enjoyment of ordinary listening-sessions, I stay well clear of any alcohol if serious judgements are needed regarding any subjective sounds.
Regards,
Re: Re: Re: Ears.
Interesting. Now, if I would like to prove this particular phenomenon over a large number of subjects, how can I make sure they are uniformly drunk, disregarding the weigth, gender, age, race, religion and shoe size?
Also, an ear popping device seems to be mandatory.
Bobken said:
Instead of deliberately misquoting me, you might have given me an accolade for forcing myself into imbibing alcohol to prove this particular phenomenon.
Interesting. Now, if I would like to prove this particular phenomenon over a large number of subjects, how can I make sure they are uniformly drunk, disregarding the weigth, gender, age, race, religion and shoe size?
Also, an ear popping device seems to be mandatory.
I am lucky to live in a place with relatively low acoustic background noise and low electrical noise. More than in the past, but far far less than you if you live in a suburban or urban setting.
I give serious thoughts to building a faraday cage into the LR (that's Listening Room) from time to time. How about you?
Listening at night is better due to the lower levels of cosmic radiation - we are shielded from the sun by the earth.
And, you also might be more relaxed. An important thing for listening, or so it seems.
Stepped attenuator? After much thought, and consulting the oracle (it's out back, deep in the woods) I chose 25kohms. We discussed 27.5328kohms also... fyi. I use a discrete stepped "L". The impedance that the driver sees is the same no matter what the attenuation. The case could be made for a "T". Everything here can drive a real 600 ohms, but not everyone brings over something that can, so 25kohm seems plausible for them.
The "switch" is a stereo "pulled from service" Daven. I use 2dB per step, only on occasion I wish I had smaller steps. The contacts are large, about 0.250" diameter, wiping, make before break type.
I've got one pair of really old "mono" ones from a similar company that have what look like solid thick silver contacts. As the manufacturing of these things proceeded forward in time the thickness (and maybe alloy quality) of the contacts reduced until the stuff in the late 70's is wafer thin sitting on top of plated brass.
But you can't hear the difference between a thick, solid silver (alloy I guess) contact and a thin nickel silver contact brazed on plated brass, right? I mean no one can...
As far as those QTP devices? Can't see how they work at this point... anyone else? They appear to be merely "spikey" looking nickel particles bound in silicone rubber... squish and the resistance drops... huh? One time, ok, but multiple repetitions? Where's the conductivity path? Wazzup wit dat?
And they're only good to 40 volts only @ 10amps.
I give serious thoughts to building a faraday cage into the LR (that's Listening Room) from time to time. How about you?
Listening at night is better due to the lower levels of cosmic radiation - we are shielded from the sun by the earth.
And, you also might be more relaxed. An important thing for listening, or so it seems.
Stepped attenuator? After much thought, and consulting the oracle (it's out back, deep in the woods) I chose 25kohms. We discussed 27.5328kohms also... fyi. I use a discrete stepped "L". The impedance that the driver sees is the same no matter what the attenuation. The case could be made for a "T". Everything here can drive a real 600 ohms, but not everyone brings over something that can, so 25kohm seems plausible for them.
The "switch" is a stereo "pulled from service" Daven. I use 2dB per step, only on occasion I wish I had smaller steps. The contacts are large, about 0.250" diameter, wiping, make before break type.
I've got one pair of really old "mono" ones from a similar company that have what look like solid thick silver contacts. As the manufacturing of these things proceeded forward in time the thickness (and maybe alloy quality) of the contacts reduced until the stuff in the late 70's is wafer thin sitting on top of plated brass.
But you can't hear the difference between a thick, solid silver (alloy I guess) contact and a thin nickel silver contact brazed on plated brass, right? I mean no one can...
As far as those QTP devices? Can't see how they work at this point... anyone else? They appear to be merely "spikey" looking nickel particles bound in silicone rubber... squish and the resistance drops... huh? One time, ok, but multiple repetitions? Where's the conductivity path? Wazzup wit dat?
And they're only good to 40 volts only @ 10amps.
Sigurd Ruschkow said:
Do you have a type number/source for those? I'll be eager to try them out. The old Shinkoh are still among the best resistors i've tried and even the magnetic Audio Note are not half bad. Quite possibly coloured, but in a way that pleases my ears. Smds will fit nicely on a Shallco.
Has anyone tried Pass-style bipolar switches for an attenuator? I never had much luck with relays and wonder if semiconductor switches may sound better.
- Status
- Not open for further replies.