Ok some normal hifi speak for a change, I never been a big fan of balanced as 90% of the time to make them balanced a lot of the equipment manufacturers just bundle in more opamps into the single path to make it balanced input or output, very few are true balanced and discrete all the way through, you need to see the circuit diagram to find this out. Even the advantage of it does nothing for me, it's just so I'd be able to run 20mt interconnects with good noise figures. In the end most hi end poweramps speaker outputs are + and ground, single ended again.
I have tried doing balanced Lightspeeds quite a few times and even when double quad ldr's are matched perfectly they go out of calibration after just a few hours sometimes days because they are exponentially more critical to temp drift. And it is exponentially harder to match double mathed quads for balanced as it is single matched quads for single ended.
It can be done with sensors that would measure a test tone before each listening session and then do an auto calibrate circuit, but that then takes them away from being a matched set with the same i/o impedances for each channel, they would get wildly different readings and therefore each channel will sound slightly different to the other at different levels. Or you could do a feedback type arrangement then your asking for more crap in the signal path and making them active and creating distortions /colourations.
After one and a half year of developing Wapo seams to have hit a brick wall with his on diy, http://www.diyaudio.com/forums/analog-line-level/170381-precision-led-ldr-based-attenuator-15.html#post2687081
what I presented to them is the way it is, with development going back to 1974 in all sorts of formats, but they love to tinker, and good on them, this is how man reached the moon, but it's sad to see them as he said "This project has hit what appears to be an insurmountable brick wall."
Cheers George |
Yes, Nelson Pass very kindly did get involved, and designed a nice direct coupled discrete (no yuk opamps) DIY buffer for the Lightspeed Attenuator, because some of his amps were 12k and 20k input impedance, and this made Lightspeed Attenuator match those low impedance amps correctly.
Cheers George |
Let's get this right!!!!
The Ravel's Bolero test is just a level test, that all, as it starts off very quite and builds slowly so you can gauge level when directly connecting CDP to poweramp, so you don't blow up your amp/speakers.
You can use any cd that starts off quietly, not just the Ravel's Bolero cd!!
Cheers George |
I am so sad/shocked about this news, Paul was a great audiophile, who spoke from the heart, and I will be eternally grateful to him for starting this thread, and he will be missed dearly.
Love you Paul, cheers George |
miller65Thanks George the temperature variation in the power supply makes sense as it is more stable in line with a stable room temperature I also intend trying with a battery in the next few weeks.
As a next stage I am considering using my CD player as a transport and introducing a DAC into my system this will also enable me to include a network player at a later date. However having checked the output spec of the DAC Beresford Caiman mk11 I have a couple of questions in terms of how output impedance can effect the sound quality . The output voltage of the DAC is 2v which is in line with LSA recommendation but the output impedance is 1k ? appreciate it should be 200 ohms or less ideally . Iwill use with a very low capitance Stager silver solid i/c 0.5 metre length .
Does anyone have experience of using the LSA with output impedance this high and if so what if any effect can it have on the sound / volume control . I love the transparent dynamic sound the LSA currently gives me and wonder if this potential impedance mismatch would have any adverse effect on the sound quality or use of the volume control which
currently operates at 12 o clock , which is generally plenty loud enough
Hi, yes please post up what you think of the battery once you've listened to it. Your Dac at 1kohm is fine still into the Lightspeed's 10kohm, but then your poweramp comes into play and series up with the 10kohm of the Lightspeed, I would say amps with over 47kohm all will still be fine. As for interconnect cable capacitance, stay below 100pf (capacitance) per foot and 2mts max, good cables are usually below 100pf per foot. Cheers George dgarretsonHi George, I'm thinking of an DIY LDR attenuator to replace the stock 10K Alps dual pot in a Sennheiser HEV70 energizer for HE60 electrostatic headphones. I need to step down from 25Vdc to 5V for this application. The chassis is cramped and there is little room inside for a heat sink on the down regulator. An LM2596 looks like an easy fit in terms of low heat dissipation and compactness.  Do you think this would compromise sonic performance relative to a LM7805 linear regulator? My concern about the LM7805 is overheating from the 20V step down. Thanks, Dave   Hi Dave, as for sinking 20v on a 1amp 7805 this is too much and will get very hot, best of getting a 12vdc supply for it. Cheers George  |
Thanks for the links banquo, I looked at the "musicdirect" link and it seems they do do international shipping also, and $49 is a great price. This supply could be a no brainer, as smp's are very good everywhere save for the noise most have, and this one seems to have addressed that and made it even quieter than linear supplies that are available. http://www.musicdirect.com/t-shipping.aspx
Cheers George |
Thank you Sean for the very favorable mention of the Lightspeed Attenuator. I just had a good look at these highly regarded Linkwitz Labs speakers of yours, very elaborate for a diy'er to do, one needs to be on the ball to make these. (and have some time on their hands).Â
https://www.linkwitzlab.com/LX521/LX521_4.htm
Cheers George |
|
The ratio impedance between the series ldr to the shunt ldr is what causes the level difference, this is why it's called a voltage divider, but one that can be varied in both planes. EG: If they are at the same impedance you have half volume coming through from the source.
As for the 2v or 4v choice, it depends on the poweramps gain and speakers efficiency, go with the one that gives you good reasonably loud listening level at mid position, but it doesn't really matter too much. With one of my very low gain amps I listen around at 3 o'clock with the others around 10 o'clock and you'll never guess which one has more dynamic headroom, yes the one that's at 3 o'clock the low gain amp, not because of the Lightspeed but because the amp has itself has less restrictive (compressive) circuitry.
It does not matter where the volume is for listening, so long as all the i/o impedances are right, and you can reach your loudest level you want to hear.
There is no such thing as dynamic headroom compression with the Lightspeed Attenuator volume control as it can pass anything the source gives out, unhindered. The only thing that will sound artifically dynamically better is a DBX unit (dynamic range enhancer) but we all (old school audiophiles ) know how disgusting they make the music sound.
Cheers George |
Hi James, I'm sure Alex mods the output stage and even the i/v stage of the Esoteric X01 when he does the mods, it is possible he has lowered the gain with his mod, also on inspection they are an opamp based i/v stage and opamp based ouput stage and it would be quite easy for Alex to give you more gain with these, just by jiggling around the feeback resistor value that are around each of the opamps. |
Hi guys, I have a potential future CA distributor, who asked me about the comparison between the Lightspeed Attenuator and the Music First TVC, I remember someone did the comparison and it was very favourable for Lightspeed can anyone give the link to it for him?
Cheers George |
Teak does not shield against rf, it maybe a step forward in glitz, but a step backward in sound quality.
One day I may do a remote and have a retrofit scheme for present Lightspeed Attenuators, but it would have to be a quality one, not those chintzy Chinese made ones that people are using that would come back and bite me in warranty claims.
Cheers George |
Hi guys, interconnect cables can be a mine field, not just with the Lightspeed Attenuator but with all forms of audio gear.
Cables can form a hi frequency filters with their capacitance (pf picofarad)) when combined with the input impedance or output resistances of sources and loads, as can output/input coupling capacitors of some sources and amps, they can create a low frequency filters with the load of the next stage/s. Then there's the inductances and resistances of cables which can also form filters (hf, lf, even band pass) of some description.
A good general rule I find is to have direct coupled outputs (no Caps) on sources you own, and direct coupled inputs on the poweramps you own, and to keep interconnects as short as possible (even if it means reconfiguring the system) and to get ones that have low capacitance less than 100pf per foot.
Cheers George |
Sorry, I should have made it more clear, it's not for more $'s. As it is I'm finding it increasingly harder to keep up supply. At it's current shipped pricing of $490usd it is only viable if the overheads are kept low as I have done, family based, with no outside employees or premises.
This would be more of a licensee/distributor to manufacture the Lightspeed Attenuator in the US for US customers only.
Cheers George |
Paul just looked at your system, that is a Esotar D330 tweeter in your speakers? I fell in love with that tweeter when I listened to a pair of Sonus Faber Extremas, it was the first time I liked the sound as much from a dynamic tweeter as my preference for ESL's for high frequency repoduction.
Cheers George |
At line level, at 2vrms which a cdp hardly ever hits, even with peaks, you get .2% of 2nd harmonic which is pleasing to the ear, and as we know with tube amps this can be 10 x this figure.
All tech info/graphs and circuits can be found at this site http://www.diyaudio.com/forums/analog-line-level/80194-lightspeed-attenuator-new-passive-preamp.html which is more for the diy'er in which I give all info on how to make one yourself, sounds as though you maybe capable, you should try it, and maybe be converted.
Even Nelson Pass has designed a buffer in this diy thread for the Lightspeed so it can drive his low input impedance amps which some are 20k and 10k.
It is a massive thread 100x more than this one with over 4,000 posts with some pretty knowledgeable tech guys giving their input for the diy'ers.
Cheers George |
If you wish to see the measurments at CD playing levels, even better than what I said in the last post, I have posted the Silonex graph and an Audio Precision distortion measurement graph, by the measurers calculations is said to be .0025% only at cd levels.
http://www.diyaudio.com/forums/attachments/analog-line-level/232027d1311212462-lightspeed-attenuator-new-passive-preamp-silonex-ns32sr2s-distortion-measurments.jpg
Cheers George
|
An interesting little snippet for those of you who own or have built a Lightspeed Attenuator, this will give you maybe some bragging rights.
Not only does the $28k DarTZeel NHB 18-NS preamp have a form of Lightspeed Attenuator controlling the volume (which they tried to patent).
But now there is a new player on the block Constellation Audio with their $60k fet line stage preamp the Altair which also uses a Lightspeed Attenuator form of volume control. (at least they didn't try to patent it)
http://www.constellationaudio.com/ca/reference-series/altair/altair-features
Cheers George |
Clio09: What exactly is an array of 48 optically controlled resistors and why can't my LSA have this feature? ;)Clio09
Not too sure Tony, could be 48 1% trimming resistors to go with the quad unmatched ldr's, which in my opinion would detract from just a quad matched set, as the i/o impedances for both channels would not remain constant, and there would be another component in the signal path.
Or it could be 48 x ldr's in some paralleled setup for series and shunt ldr's, again this would detract from a simple matched quad set. Or it could be a 48 position switched volume control instead of continuous rotary pot with a quad matched set, this would then sound the same as my production Lightspeed Attenuator.
Cheers George |
Dave is using it the way the first Mk1 Lightspeed Attenuator was designed. This is a series resistor with an ldr for the variable shunt to ground. This is still better than any potentiometer, but when I compared it to the MkII Lightspeed Attenuator which is series LDR and shunt LDR exponentially harder and costlier to make, it was clearly another step up in sound quality and usability, so much so that I recalled all the Mk1's and they were all converted over to MkII status, and all the owners were amazed at the difference. It was more dynamic, punchier in the bottom end and a more transparent sound stage, also the min volume went down closer to zero, and it had more stable i/o impedances at different settings. Dave you should try to make a MKII you will be very pleased with the difference again.
Cheers George |
Paul, think of it this way, you have 4 lanes of peak hour traffic coming into two different cities, one of the cities has a bridge that can take all 4 lanes into that city without any speed reduction to the cars,. the other one has less lanes on the bridge therefore the cars have to file over the bridge at slower speed than they approached it at.
Same thing happens to electrons,the bridge/input impedance compresses the music to get it through. This happens every where in the audio chain, from cartridge to phono stage, from phono stage to preamp, from preamp to poweramp, and the big one from poweramp to speakers.
Cheers George |
It actually only becomes a gain reducer when the input impedance of the poweramp is lower than the output impedance of the preamp.
Cheers George |
Hope all the testosterone has now been depleted.
Your question Unsound:George, perhaps you would discuss the balanced issue?:Unsound
Balanced can be done, for the diy'ers, but to manufacture it, it would be a nightmare, as to do match a quad set which the production Lightspeed Attenuator has, is already very time consuming. To do a double of this for a balanced setup is exponentially harder and costlier and may drift.
Not impossible for the diy'er though as it could be self calibrated. every few months by the builder. I have posted the circuit for a balanced here on diyaudio, you may have to sign up to open the attachment though.
http://www.diyaudio.com/forums/analog-line-level/80194-lightspeed-attenuator-new-passive-preamp-383.html#post2396316
Cheers George |
If I remember Batalok I think you maybe one of many I may have said the input impedance was a bit low around 30K and suggested that any competent tech could raise it up to 68k or so with 1/2hr work and a 5c resistor.
Cheers George |
Pubul57: Raises the question as to the the "effects" are, and when is gain "necessary". Pubul57:
With today's sources, cdp, dacs, even phono stages, there is clearly no need for added gain in most systems. As what most sources can give out today in volts and current, they can easily overdrive (clip) most of today's amps. So all that's needed is a way to attenuate that source no need to give them more gain with the added colouration of what an extra gain circuit gives, only to reduce it back down again to below what the source originally gave out.
Cheers George |
09-17-11: Banquo363I should add: what else should we look at other than impedance matching concerns.
If the impedances are right, nothing, so long as the dac or cdp has volts high enough to clip/overdrive the poweramp which nearly all have, and enough current which nearly all have, except some tube output dac's and cdp's.
Look at any of the user manuals of the higher end dac's and cdp from Wadia going right back now to the late 80's that had digital domain volume controls in them. The manual states going direct into a poweramp with any of these units will yield a superior sound than going through any preamp, so long as the digital volume is in the top third so there will be no bit stripping.
There are no preamps inside these top Wadia's I have personally seen inside a few of them they are just like all other dac's and cdp's on the market, with a current to voltage conversion stages after the dac chips then output buffers to the rca's. The Wadia's magic happens before the dac chips, in their propriety filtering techniques and receiver stage .
Cheers George |
You do have a bad match Teayjay with or without the use of the Bryston Xover, and what it shows is that when you said you tried the Lightspeed Attenuator you also put it into that same bad impedance matching situation.
With or without the Bryston xover, the Lightspeed saw the 20k input of the Passlabs XA60.5 or it saw the 15k input of the Bryston xover, and your $60k preamp is in the same unfortunate mismatch.
It is well documented that the Lightspeed Attenuator needs to see 47k or more load, and your preamp is even higher at 100k.
You your self said you do not care or know about impedance matching components, yet you think you have cred by just throwing money at components and bagging them if they don't sound right. |
09-17-11: Grannyring Actives also need to be well matched with the amp. My cuurent tube active was a poor match to the Sunfire Signature amp which has an input impedance of 24k ohms.Grannyring
Yes Grannyring, tube output active preamps are just as susceptible to poweramp impedances as passives are.
I looked at what Teajay is using. The output stage of his Concert Fidelity CF-080 is a 12AU7 in cascode configuration direct to the SE output rca's, I thought this is a high output impedance. So I checked with our tube gurus, they confirmed to me for this tube to be use in this configuration in it's best linear way, it's output impedance will be around 33K! And at best it will only give 10mA maximum output current.
This is fine if it sees a poweramp with 100k or higher input impedance, but he is using it into a Pass Labs XA100.5 which has 20k SE input impedance??? Bad Match this one!!
Cheers George |
Sorry correction, XA60.5, it is the same 20k for the single ended input as the XA100.5 |
Also if it is feeding any of the Bryston Xovers this is the same bad match 15k or 20k input impedance depending on model, one needs to do their homework before spending the big bucks. |
Try to bear with me here and let it sink in.
You see this impedance matching goes right through the whole system between components. EG: source to preamp, preamp to poweramp, and poweramp to speakers.
Even the internal stages of the individual components should be impedance matched if the designer is worth his salt.
eg: of a poweramp it's: input stage to voltage stage, voltage stage to driver stage, driver stage to output stage. Same for goes the pre and source.
Hopefully they are done/designed right as the end user cannot do anything about these stages inside the individual components or check them. But at least the end user can make sure the components are themselves matched from each other.
I specify with the Lightspeed Attenuator that it needs to see a poweramp with 47k or higher input impedance, I believe that tube preamp and tube dac/cdp manufacturers should be saying the same, this is why so many high output impedance tube pre's sound bad when they are mated to low impedance poweramps.
Cheers George |
Paul one of the tube guru's on diyaudio said the solid state part is at the input, and it's even stated that the 12AU7 in cascode mode is directly attached to the rca outputs in the manufacturers blurbs on a website.
I asked what other preamps used this cascode output stage, and he said Rouge Audio used to on their preamps but their poweramps are very high, the M120 monoblocks I had here were 1meg! input impedance. And also he said some early Lamm preamps were also cascode output.
It is ok to have this high output impedance if you have over 100k or more on the poweramps, but you do get the problem of having to use low capacitance and short interconnects 1.5mt or less, just like with the Lightspeed.
Cheers George |
Hi Tony,this is what a cascode 12AU7 will be judging by the tube gurus. 33k is not bad so long as the poweramp is 100k or more. What was the input of the power amp used at the CES/RMAF.
Cheers George |
Like I have always said Tony, it will work and sound ok with no damaging side effects.
But the source should be at least 3 x or more lower output impedance than the the load. I imagine then it could have sounded even better.
Now can you imagine what Teajay heard when he listened to the Lightspeed into a 15k load and drew his conclusions?
Cheers George |
Yes I too received it today, and I need to tell all, before any assumptions are made, that Sam Tellig (Tom) purchased one at full price over 2 years ago, before I had any idea that he was going to review it.
And just a couple of weeks ago one was sent off to a classical guitar buff, in New York, his son David, again at full price.
Cheers George |
This is the output description "The 12AU7 tubes are on the back panel virtually bonded to the output connectors."
This is where the solid state part is used "It uses a solid-state analogue switch for input switching and an analogue solid-state volume control selected for its sonic quality and tracking accuracy."
All this can be read here, half way down.
http://www.positive-feedback.com/Issue41/concert_fidelity.htm
Cheers George |
You have no constructive critisism Teajay, all you try to do is to get this thread shut down with your personal attack.
Not once have I attacked you personally just questioned the Ohms Law compatibility of components in your system that have not been properly addressed in the specs they have presented, and the voodoo that surrounds them. |
The HF filter that capacitance of an interconnect that is 300pf for it's total length when with the Lightspeed is 76khz -3db
200pf is 114khz -3db
and for 100pf it is 227.5khz -3db.
As you can see even the 300pf is fine as it is still only -3db at 76khz.
This is why I recommend 100pf max per foot and the use of 1mt length with that maximum capacitance.
But most good quality interconnects are well below 100pf per foot which is even better.
This also applies to high output impedance tube preamps with their interconnects, especially when say a 30kohm output impedance preamp is used, if it has an interconnect with 300pf for it's length, it works out to be -3db at 17.5khz which cuts into the audio band quite severely.
Cheers George |
Banquo363Am I right in believing that non-ideal impedance match between phono stage and LSA results in a attenuation of the high end? Perhaps it hasn't bothered me because of my old ears?Banquo363
No, the slight impedance (resistance)mismatch you have will compress slightly the dynamics, and if ever the output impedance becomes higher than the input impedance then you will attenuate the whole signal, as it then becomes a voltage divider.
What attenuates the highs is lots of cable capacitance.
Cheers George |
2: Never had any problems with the ldr side, as my prototype is still as it was when new and that's 7 years now powered 24/7. When I say in the instructions to leave the Lightspeed Attenuator at 1/2 volume when leaving it powered up and not listening, it's the led's I'm thinking of, even though when at full or min volume they are never at their max mA rating, it is in my opinion to share the use evenly so I say leave it at half way if not in use (listening).
5: The best is as it is now, it's simplicity in it's purest form, as the music signal only goes through one soldered resistor (ldr) from input to output, anything else would need to go through more devices. As for the power to drive the led's, it is known that battery is the purest form of dc power, and so many times we have A/B'ed between the wall wart and a Lithium Ion battery and nobody could detect the difference reliably some say they hear a difference but can't tell what it is. Even golden eared Sam Tellig, need to see with his own eyes if he has the wall wart or battery attached to his Lightspeed when he's listening.
As for the Warpspeed not interested in it, too much other crap being put into the signal path.
As for doing any other things to the Lightspeed, I would like to have more inputs, but the only way I would do them is to use Mercury Wetted relays for the input switching (which I believe are being made illegal) and that would be still not as good as only one input with no switching. So use one of these, and then you can still go direct when your really want to hear your system at it's best. http://www.decware.com/newsite/rotary.htm
The only thing I may do in the future if I find a reliable set is a remote control volume and I would make sure it's retrofittable to all older Lightspeeds. But I have not yet seen a system I'm happy with that has long term reliable supply and quality. For a remote volume you need, a quality remote, receiver circuit, motorized pot and power supply to run it and the receiver
Cheers George |
Hi guys, as I have found out input switching is almost as detrimental to the sound quality as a volume potentiometer, this is why I always recommend a separate good quality input switching unit so you can still go direct when you want to bypass all contacts and get the best out of your system or when you want to show it's transparency off. http://www.decware.com/newsite/rotary.htm
Now the obsessed fiddlers on diy audio have found this out also, and to their credit they have now devised their own way of input switching unit using ldr's
http://www.diyaudio.com/forums/analog-line-level/197894-ldr-based-input-selection.html
I still think the one I posted was simpler and it auto shorted the unused inputs.
http://www.diyaudio.com/forums/analog-line-level/80194-lightspeed-attenuator-new-passive-preamp-153.html#post1640716
Cheers George |
There should be only one mains earth point in a system, (if you have more than one you will have a loop), all the other components will get their earth via the interconnects from the one that is earthed. Once you have two components earthed then it will become a loop via the house earth line.
Even when I had two monoblock 805 tube SE that I built, I got a virtually no hum when only one was earthed and the other got it's earth via the interconnect to the Lightspeed and back down the other interconnect to the unearth monoblock.
PS wonder if I'll get the magic 1,000 post with this as they moderate my posts for up to 6 hours before approval.
Cheers George |
I ask one thing, for a single ended, low capacitance interconnect of 1-1.5mt, what are we trying to control in electronic and mathematical terms, when phrase "controlling the ICs" is used?
Cheers George |
Different amps will have different gains, so will sources, and speakers also have different efficiencies. This will all govern where the volume control (on any passive preamp or any active preamp tube or transistor), position will be for normal listing.
Now with the sources giving higher outputs than they ever have, and the recordings being higher in level also, too much gain is really becoming an issue with most systems.
This is why they are starting to make rca attenuators like the Rothwell Attenuators in -10dB and -20dB to take care of this problem.
http://www.decibelhifi.com.au/category42_1.htm
Another reason why there is no need for active preamps with gain, we have too much gain coming out of nearly every component in a system.
Cheers George |
The Rothwell attenuators work just like the Lightspeed Attenuator. That is, they have a soldered fixed series resistor with another shunt fixed resistor to ground, this forms a voltage divider just like the Lightspeed Attenuator.
Except the in the Lightspeed Attenuator these series and shunt resistors have variable resistance with the amount of light that shines on them. So what you have is a variable voltage divider in the Lightspeed Attenuator.
The only problem with the Rothwell's is that they change the input impedance of the device they are attached to, so this needs to be taken into consideration when using them, with passives and tube preamps.
Cheers George |
If the Rothwell is 9.5k shunt to ground and the amp is 470k input, any preamp will then see 9.3k. Not good for any passive or tube active.
Cheers George |
Sorry plus the 21k added to the 9.3k, yes 30.3k.
Cheers George |
Yes you are correct Tony. The reason I posted them up 16 posts ago was to say that there is a problem these days with too much gain in systems, and that there is almost no reason for active preamps with even more gain again, they nearly all need only to be unity gain buffers only, if one wants an active pre.
Cheers George |
Sounds like you are one of the many who have no need of a active preamp with gain because you already have an abundance of gain from either your source, amp, and or very efficient speakers.
Yes the Lightspeed Attenuator could help here, as well as improve the transparency/dynamics of the sound, being truer to the source than active preamps.
Just a couple of parameters that should be met.
1: Your source cdp or dac's output impedance should be 100ohms or lower (which most are).
2: Your amp's input impedance should be 47kohm or higher (which most are)
3: The interconnect from the Lightspeed to the amp/s should be 1.5mts or less and of good quality
Cheers George |
Dpac996 hi, yes a Lightspeed can be used to drive a poweramp that have as little as 22kohm input impedance or even lower and there is no harm to the components doing this. I have many customers that are doing just that, but I like to present the parameter specs that are well over with headroom, so 47kohm which is the industry standard was chosen, even though we use it at 33kohm at our audio society meetings.
As for LDR matching, it is a must to have all 4 series and shunt matched as this will give a better logarithmic feel to the volume control, if you only match the series L/R and then match just the shunts L/R this will still work but the volume control will either be very touchy down low or very vague down low.
Cheers George |
