PMA said:
Yeah, me too, but it wasn't practical: http://www.diyaudio.com/forums/showthread.php?postid=839082#post839082
That was the most "advanced" version: http://www.diyaudio.com/forums/attachment.php?s=&postid=863703&stamp=1141741699 
I should probably step in here on the volume control topic....Back in the 80s we were using C&K 23 position rotary switches in multi-deck format to switch between individual resistor pairs, for single ended signals. This allowed a pF size trim-cap to be put in parallel with the sppropriate resistors to shunt a but of HF back into the output for good scuare wave response, unto the low MHx region. I always found that 10K impedance was a good palance between being able to drive it, having moderate self noise, and a low enough output Z that you didn't loose everything to stray C. Later on we graduated to Shalco 32 position switches..better materials and more positions....
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.
The DACT controls are a series chain of resistors, with a tap moving between them, so the output may have a lot of solder joints and resistors in the path, depending on position. They are OK for budget applications about 200 bucks a pop, and do sound better than all the pots we have ever listened to, including the hi dollar Alps and assorted conductive plastic, carbon and cermet pots from a host of makers. I have used the DACT as a common mode variable resistor between the balanced lines with good success, however, a large rotary switch to select an individual resistor would be better. Goldpoint makes DIY switched attenuator kits, but I have never used them.
I am curious as to the sound of the best opto-resistors, [ there have been some threads on the subject on the DIY audio forum] and would like to experiment with a digi-pot running an opto.....The digi pots themselves have become popular in consumer gear, even some hi-end and ecpensive stuff....but I don't like the way they sound either. Perhaps using one to operate a better sounding [ unknown if this is true] optical resistor may be a way to get better sound and 100 or so resettable positions. Hmmmmm....
So far the best sound I have acheived in an attenuator is a large frame 32 position Shalco with individual resistor pairs at 10K impedance. Driving a tube grid only a few inches away I have not needed compensation bypass trim-caps, but with some C in the output lead from shielded wire or stray C, and lower impedance loads, you would want to experiement with some trimmers...After all, that is how oscilloscope attenuators are made to have hi bandwidth....
Relay banks could also be used to switch between resistor banks, Matt Kamna has recently developed a 15 step fully developed product with remote control to do this, soo to be commercially available....You have to watch for dielectric absorption from all that circuit board and long traces rather than hand applied wire....the usual trade-off, labor and potential for errors VS fast assemble and low potential for mistakes, with best sound in the high labor version...
Maybe more thoughts later.....just some $0.02 worth for now
Woof!
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.
The DACT controls are a series chain of resistors, with a tap moving between them, so the output may have a lot of solder joints and resistors in the path, depending on position. They are OK for budget applications about 200 bucks a pop, and do sound better than all the pots we have ever listened to, including the hi dollar Alps and assorted conductive plastic, carbon and cermet pots from a host of makers. I have used the DACT as a common mode variable resistor between the balanced lines with good success, however, a large rotary switch to select an individual resistor would be better. Goldpoint makes DIY switched attenuator kits, but I have never used them.
I am curious as to the sound of the best opto-resistors, [ there have been some threads on the subject on the DIY audio forum] and would like to experiment with a digi-pot running an opto.....The digi pots themselves have become popular in consumer gear, even some hi-end and ecpensive stuff....but I don't like the way they sound either. Perhaps using one to operate a better sounding [ unknown if this is true] optical resistor may be a way to get better sound and 100 or so resettable positions. Hmmmmm....
So far the best sound I have acheived in an attenuator is a large frame 32 position Shalco with individual resistor pairs at 10K impedance. Driving a tube grid only a few inches away I have not needed compensation bypass trim-caps, but with some C in the output lead from shielded wire or stray C, and lower impedance loads, you would want to experiement with some trimmers...After all, that is how oscilloscope attenuators are made to have hi bandwidth....
Relay banks could also be used to switch between resistor banks, Matt Kamna has recently developed a 15 step fully developed product with remote control to do this, soo to be commercially available....You have to watch for dielectric absorption from all that circuit board and long traces rather than hand applied wire....the usual trade-off, labor and potential for errors VS fast assemble and low potential for mistakes, with best sound in the high labor version...
Maybe more thoughts later.....just some $0.02 worth for now
Woof!
john curl said:
Hi John,
Yes, listening tests. I'm for whatever works. I'm not a stickler for double-blind AB tests, for sure. I just like some convenience and reasonable apples-to-apples discipline. Look on my website for what we did at RMAF 2006 and HE 2007. Certainly not perfect, but better than totally un-disciplined tests. Whatever works. Even whatever kind of listening tests (I prefer the term listening comparisons) you and your friends do.
With or without Remi Martin pre-conditioning
.Cheers,
Bob
dimitri said:
I listen in the evening. Sometimes I’m overwhelming with emotions of the day and they don’t allow concentrating. Soft reset is required.
A well and proper TRUTH!!!! dimitri, you got THIS so right.
No one that I know has the time and circumstances to listen 'as well' as we might like, just when it might be fun to do so. The real world intrudes all to often on the desires of the aesthetic sensibility.
This is a REAL problem that all of us 'serious listeners' face....At some times our music systems are tools of escape from the stresses of life, a place to bask in the 'sun' of musical glory as an antidote to the pressures that real life imposes. At other times we must wear the hat of the critical and unbiased reviewer of our own and others work, to evaluate the progress or recession we have wrought. When entering into a serious listening session, I have found, for me at least, that it is best to have a clear and peaceful mind, to have cleaned out the earwax, AND here is one that very few listeners do...to clear out the Eustacia tubes to equalize the pressure on both sides of our eardrums. [Like popping the ears when changing elevation..Scuba divers get REAL good at this, cause you gotta do it every few feet of depth change ] Listening is both a physical activity, involving the ears, and neurological one, involving huge amounts of our brain processing power. The way we are "wired" our visual system has precedence in the brain for processing power, displacing other functions such as hearing and smelling to lower priority. When there is an absence of visual input, such as when listening in the dark, the brain will 'switch over' to offering much more of the resources to processing audio. I suppose someone would argue that if we can't see, we will optimize the rest of our best senses for survival...and for whatever reason, it is true, when we listen in the dark, we DO hear better.....There really is MORE brain power available to process sound when your in the dark.
The process becomes much more complicated when involved in live recording...especially when only headphones are available for monitoring. There is usually a lot of stress as you have little time to set up and evaluate the acoustics as they impact the recording decisions, ....the musical events are usually non repeatable and you HAVE to get it right because there is ONLY the one chance, ....Recordists have it tough, as your are never in the dark so the brain is focused on processing vision. You have learn how to keep the eyes closed to hear better in the phones, and to also take time away from them to listen to the REAL acoustic performance, .... During the intermission when I like to try and be quiet and relax, everyone wants to come and bother me for a listen...At least these days we don't have to keep 2 large tape decks streaming tape and changing reels. It used to be a real battle to keep things going, as even with the large reels, at 30 IPS you only get about 15 minutes, so you have to start one before the other ends with most performances, and splice later. Today ya' just push the start button and watch the levels, which you HOPEFULLY got right in rehearsal so you don't have to touch them.... By the way, EVERY time I record an orchestra, they have gone about 3 dB higher in actual performance than in any rehearsal, so just keep that in mind if you ever get the chance to record one, it will save your headroom butt if you adjust for that before the music starts!
When I record an orchestra, or anything else in the way of location work, for that matter, I always listen to the work just completed after we get home. It usually quite late, and I'll have left the playback system running to be well warmed up....There really is nothing else like a well done DSD master file heard for the first time over a decent system. It is a time of de-stressing from the just completed job of recording, and to just enjoy the fruits of the labor. No one else in the world has heard the 'just made' recording on a stereo before this moment, and with all the production steps to make a released disc removed, the purity of sound in the recoding from the raw file, well it just does not get any better unless it IS the live event. I feel quite blessed to have the opportunity to experience this over and over. Its too bad that more people don't have the experience of being at the live event and then hearing it only slightly later from the live-master machine. It really does reveal the limits of recording technology with the memory of the live event still strong in the mind, but it IS great fun too to hear just how good it can be done.
Listen for fun, Listen to escape, listen to evaluate! All are part of the larger audio/music lexicaon and each IS totally valid, but they all quite different experiences.
cheers!
More on 10K, my personal attenuator impedance preference ....I was just reviewing a few pages upstream from here and saw the posts about problems with interfacing with tube gear.....I have a vintage Marantz 10B tuner that was originally designed to drive about meg ohm or maybe half a meg...Well, what to do....
Back in the late '70s when I bought the darn thing , I just ripped out the 12AX7 cathode follower and the tiny 0.1uF coupling cap and replaced the wimpy tube with a 5814A running at MANY time the current, and shoved in a much larger output coupling cap set. The caps were the hard part...not much room in there. I ended up using several 3 uF caps in parallel, and putting them where I could fit them and not cause cross talk and other pickup issues. Yeah, a kludge, but WOW, did that tuner ever sound better, even with really hiZ loads. One of these days I'll build an additional outboard supply and add-on a small chassis above the main chassis, to run a really higher power diff out set of modern tubes, or maybe go solid with a really good FET stage. I'm not hung up on the idea of keeping it all tubes so who knows.........Anyway, once you get past the 'collectible' issues
with vintage gear, making some of this older stuff work MUCH better is fun, AND prudent, at least if you want to actually listen to it like I do. 
But then again, every piece of audio gear I use that I did not actually build, is modified, even my microphones! To me, no gear is sacred, and tomorrow's 'mint' collectors will likely curse and bitch at me in my grave when they find my stuff.
Back in the late '70s when I bought the darn thing , I just ripped out the 12AX7 cathode follower and the tiny 0.1uF coupling cap and replaced the wimpy tube with a 5814A running at MANY time the current, and shoved in a much larger output coupling cap set. The caps were the hard part...not much room in there. I ended up using several 3 uF caps in parallel, and putting them where I could fit them and not cause cross talk and other pickup issues. Yeah, a kludge, but WOW, did that tuner ever sound better, even with really hiZ loads. One of these days I'll build an additional outboard supply and add-on a small chassis above the main chassis, to run a really higher power diff out set of modern tubes, or maybe go solid with a really good FET stage. I'm not hung up on the idea of keeping it all tubes so who knows.........Anyway, once you get past the 'collectible' issues
with vintage gear, making some of this older stuff work MUCH better is fun, AND prudent, at least if you want to actually listen to it like I do. But then again, every piece of audio gear I use that I did not actually build, is modified, even my microphones! To me, no gear is sacred, and tomorrow's 'mint' collectors will likely curse and bitch at me in my grave when they find my stuff.Optical attenuators or really linear optocouplers are not very linear. A British company makes probably the best but they all have considerable distortion. The distortion vs voltage curves are published for some of them. To get low distortion you must keep the voltage low (10 mV or less). if you look at the oscillators that have used them for AGC there are tricks for keeping them pretty linear. They also have considerable lag in response that is much slower going up than down.
I spent some time looking at pots and other solutions for volume controls. Measured the distortion of many samples and was disappointed in quite a few. Alps pots are much better than the generic Taiwan clones. A Vishay metal foil trimmer is much better still.
Sonically, I found the tech-labs switch with Vishay (Texas Components) foil resistors to be a real benchmark that is hard to exceed. But a relay network of the same resistors and the Teledyne TO5 can relays is close, and needs far fewer parts. Managing capacitive parasitics isn't easy but 10 K works up to 500 KHz or a little past. The stray capacitance in the wiring on the output and the input C of the next stage is what you are compensating for. Above 500KHz you must go to lower values. There are a of of tricks that can be learned from Tek scope input circuits for dealing with these problems.
I spent some time looking at pots and other solutions for volume controls. Measured the distortion of many samples and was disappointed in quite a few. Alps pots are much better than the generic Taiwan clones. A Vishay metal foil trimmer is much better still.
Sonically, I found the tech-labs switch with Vishay (Texas Components) foil resistors to be a real benchmark that is hard to exceed. But a relay network of the same resistors and the Teledyne TO5 can relays is close, and needs far fewer parts. Managing capacitive parasitics isn't easy but 10 K works up to 500 KHz or a little past. The stray capacitance in the wiring on the output and the input C of the next stage is what you are compensating for. Above 500KHz you must go to lower values. There are a of of tricks that can be learned from Tek scope input circuits for dealing with these problems.
- Status
- Not open for further replies.