The 0.25mV 103D into my phono pre+active stepup's 64dB gain yielded 0.4V, that just didn't cut it into either of my amps, especially the 1.5V monoblocks.
Looking at those measuremnts you should have had 2.65v output more than most cdp's.
You either have a fauty phono stage not giving 64db of gain, or the cartridge is faulty and not giving .25mv
Cheers George |
You are right Pubul57, the momment you start to lower a digital based volume control, you start to whats called "bit strip" the name says it all, and this lowers the resoution. Aways leave it full up when using in conjuction whith the Lightspeed.
But also be carefull some dac's and cdp's have analog based volume controls and these are usually sub grade passive motorized or non-motorized on the output of the cdp or dac. Most times you have also fixed ouput rca's as well on these cdp/dac, always better going for it and a Lightspeed than the analog variable output. If there are no fixed ouput rca's, better to open it up and bypass the passive pots by taking them out of the output totaly.
Cheers George |
That is correct, buffers are there to drive long interconnects, and or low input impedance amps. There is no perfect buffer they all sound different As there is no perfect interconnect they all sound different, unless the amp is so low in input impedance then this puts all interconnects (within reason) on a level playing field
I keep comming back to this, plug the cd player straight into poweramp ( with more than 47kohms input impedance) with good quaity 1-2mt interconnects, this is the perfect transmission of the cd players sound (save for the interconnects signature), then put in the Lightspeed Attenuator without buffer and see how little it influences that sound compared to any other active pre or buffer.
Cheers George |
I assume you are using the input that is full range not the one with the roll off at 70hz.
The diagram actually shows it's 27k input impedance this together with the 1uf coupling cap (yuck) gives a 5hz -3db roll off, the opamp is a 14watt chip amp that has an input impedance of 5megohm, you could change the 27k for 82k and you will see it will sound better. And go lower theoretically as the 82k resistor now with the 1uf coupling cap (yuck) will now go down to 2hz -3db but you will not hear this as the interstage transformer and output transformer cannot go down this far anyway.
But don't switch on the amp without the Lightspeed attached as that will let the chip amp see only the 82k resistor and chip amp may complain, when the Lightspeed is attached it sees the 7k odd of the Lightspeed and it will think everything is fine still.
Cheers George
by Georgelofi |
I much prefer direct coupling, I can hear coupling caps. But I do not think this circuit can be direct coupled as I think there is a high input offset from the chip amp, I could tell better if they gave a the internal circuit diagram to it, but they don't.
Cheers George |
You are right Paul, in some instances.
But in our case with the parameters I have outlined for the Lightspeed Attenuator, if you have a source that has low output impedance, less than 100ohm and 2v or more output voltage, and the power amp at more than 50kohms input impedance, there is absolutely no need for preamplification, (it's Ohms Law). No amount of extra current in the form of extra active current stages will have the desired effect of giving more drive and control, if fact the opposite could be had and heard, as well as a loss of transparency.
Cheers George |
Agear hi your questions to me are answered here it's a big read, I started doing these some 30 years ago it was new then, I should have patent it, it is refined sonically to the max today, without being overly complex, and pricing kept low as possible.
http://www.diyaudio.com/forums/analog-line-level/80194-lightspeed-attenuator-new-passive-preamp.html
www.lightspeedattenuator.com
Cheers George
|
Agear and guys, let's get the name of the main devices inside the Lightspeed Attenuator correct, they are not solar cells, they are not optocouplers, they are LDR's (Light Dependent Resistors). These LDR's are made from (back in the old days) Cadmium Sulphide, which back in the 70's were almost impossible to match up for me, and were unreliable, I needed about 100 units to get 4 matched. Now the ones I use are made of something else they won't say what they keep it close to their chest, but it has enabled me to match 4 out of about 30 units and the reliability is no longer a factor.
Some of the "tech heads" are saying that it's the Cadmium Sulphide or whatever it is that these LDR's "Light Dependent Resistors" are made out of , is what makes the Lightspeed sound better than metal or carbon film resistors, this could have a slight bearing.
But you know what I believe and have seen if you read my explanation on DIY Forum or read my PDF broucher about "Dynamic Contact Bounce"
Cheers George |
1mt each each side of the Lightspeed Attenuator of Eichman Express Six unshielded.
I have found shielded cable a little constricted compared to unshielded of the same construction, I believe these days with CD high level output 2v there is no RF break-through any more. Shielded cable is just a hang-over from the low output mV phono days.
Cheers George |
For those who follow Stereophile, the Lightspeed Attenuator has made it into "Class B" Recommended Pre-Amps with a $$$ value tag in the October issue.
I received an anonymous email saying if your wondering why it didn't get into "Class A" just have a look, nothing below 10K, it will never happen.
Cheers George |
I could but instead of $450 it would end up costing $1450 and would then be no longer a "Stereophile $$$ value"
Quality remote, reciever, motorized pot, different power supplies and larger chassis.
Cheers George |
And here is the Flecher-Munson 1933 study done, for percieved frequency loudness at different db levels. As you can see at low levels one needs to boost bass quite a bit to remain flat to the ear. Half way down the page is the graph everyone can understand.
http://www.jimprice.com/prosound/db.htm
Cheers George
|
10-11-10: Agear
CDPs generally suck when driving amps directly due to a lack of current (not merely a function of output voltage as some have suggested above in this thread). Current, from my limited understanding, is needed to reproduce in particular lower frequency information. Agear
CDP's (not tube output ones) have lower output impedances and therefore and higher driving current ability into power amps than 99.9% of tube preamps.
Cheers George |
10-09-10: Grannyring
Quote>Perhaps the 2D forwardness speaks more to my system synergy then to anything else. This may well be the case as I have no way of really knowing.
What you should do Grannyring to see if the Dude is artificially giving depth, is what I preach all along, put your CDP straight into your poweramp (Bolero Test) no preamp. Put on a quite cd so you can then ascertain a good level of cd to play, then swap in the Lightspeed then your Dude and see which is closer to no preamp. The one that is, is the one that is truer to the source.
BTW the English and some Aussie speaker manufacturers back in the 70's & 80's use to purposely -3db the level of midrange units to give the illusion of greater 3D depth.
Cheers George |
Well said Pubul57.
If I could just add to that, as I tell many customers, a direct connection between CDP (non tube) and Poweramp is the best/truest transfer of what is recorded on the disc. And the Lightspeed Attenuator comes the closest to that direct CDP (non tube) to poweramp connection.
If an active preamp sounds better than this, then it is the system that has a problem and needs the coulouration of the active preamp to band-aid fix that problem, it would be better to fix the problem than to cover it up.
Cheers George
|
Just looked at the Merlin VMS impedance graph (Stereophile), at it shows it only drops (over a very narrow band)to a minimum of 6 ohms at around 170hz at 0 phase angle.
This is an easy to drive speaker and should be driven off the 8ohm tap. Buy driving of the 4ohm tap all you achieve is loss of power from the amp, vitually no gains whatsoever.
Cheers George |
You may find that the OTL's are getting loaded down some by the speakers and dropping their output requiring more push from the Lightspeed/CDP. The sensitivity of the OTL was probably judged/measured with an unloaded outputs.
Cheers George |
Sorry forgot the other part of your question, as for the position of your Lightspeed control, that's fine it's a good place to be, at around 2o'clock it really starts to increase/rampup the gain rapidly, so you have oodles of gain left.
Cheers George |
Hi guys and Knghifi, glad your enjoying the Lightspeed Attenuator, you've put it up against a very expensive ($14,000) in the Vac Signature pre, I'm over the moon it held it's head up high against to it. One question Knghifi the poweramp TRL Samson you used for the comparison, it's a solid state amp and I have looked everywhere for it's input impedance and nowhere can I find it, as the Lightspeed Attenuator likes to see 47kohms or higher for the poweramp it's feeding. If the TRL is lower than this you have still not heard the Lightspeed at it's best, and heaven forbid it may beatout the VAC Signature if it sees an amp with the input at 47kohms or higher. Does anyone know the input impedance of the TRL Samson?
Cheers George |
It would be good for someone to grab an ohm-meter and measure the input terminals (Amp off naturally).
If it is bipolar transistor input you can bet you life it's going to be less than 47k.
If it's fet transistor input there's a chance it's 47k, maybe higher.
If there is no reading or wavering all over the place it has capacitor coupled input (yuck and double yuck)
Cheers George |
Banquo363 As stated in the instructions, you can leave it powered 24/7 just leave the volume at half position when not playing music. If using battery supply this is not possible, but still you can plug in the wall wart and leave it on 24/7 and you can change over to the battery even while playing music just pull out the wall wart plug and insert the battery plug quickly, no bangs or crashes.
Cheers George |
So if he doesn't like Fets or Mosfets the input stage to the Samson must be bi-polar, it is not tube as it's all solid state.
This means the Lightspeed Attenuator is maybe not giving 100% of it's self, if the Samson is bi-polar input they (poweramps with bi-polar inputs) are usually below 30kohms input impedance.
We have found in listening tests with many "golden ears" that the Lightspeed Attenuator just starts to compress the dynamics slightly at 33kohms, 47kohms was fine, indistinguishable right up to 200kohms.
Cheers George |
Clio09 mentioned that "Paul Weitzel told me he does not like FETs or MOSFET" so all I said was "being the Samson is all solid state, that if it's not fet or mosfet input it would have to be bipolar", and any tech will tell you that it's highly unlikely to be 100k input impedance with bi-polars on the input stage. Look at any bipolar input amp, most are below 47k usually 10k to 30k, as dc offset problem arise if higher than this with an open circuit input.
Cheers George |
As for having my own Lightspeed Attenuator forum Tsciame, it would be nice, but I would have to maintain it, moderate it as well, and I'm flat out building them, it's nice when I have a break in orders to go down and have a surf at the beach come back refreshed and get stuck into it again. A smart man would sell it all off to someone in China, then one could surf all day, but then who said surfers were smart?
Cheers George |
No need to do the A typical Audiophile thing here, and go overboard.
So long as you have 100pf or less per foot, that's fine as this gives with 1mt a -3db 76khz, this is over kill anyway as the cd filtering is lets face it already cutting in the mid 20khz anyway, so you won't gain anything by going even lower that 100pf per foot. No need to waste money, KISS
Cheers George |
OK now here's the truth from the horses mouth, the Lightspeed was seeing a happy load and Grannyrig check your facts before you double swear it was 100k input. And as I said for it to be over 47k it would have to be Fet input. Below is Paul Weitzel email to me. Cheers George
Hi George. Answers to your questions.
1. 68K ohms
2. FET input
3. DC coupled
They have pretty normal specs.
Take care.
Paul
TRL, Inc.
tuberesearchlabs.com |
Pubul57
The TRL amps are all SS? Not hybrid?
Pubul57
Samsons are all solid state 68k, fet input, dc coupled.
Cheers George |
Apologies to Grannyrig, you did not state the Samson was 100k, I misread the statement you made, it was the Atmasphere you were eluding to.
This has got a bit out of hand, because I had mentioned "was the Lightspeed seeing the ideal load".
It was and was beaten out by the $14,000 VAC Signature MKIIa in Knghifi's setup, I wait for future posts Knghifi after the dust has settled and all has burnt in.
Cheers George |
If the signal/source is poor, yes you can alter, soften,bloat thin out or whatever with an active pre, depending on it's own character of sound, yes a it's band-aid fix as you said.
With the Lightspeed Attenuator you hear what the source/signal is giving, if it's not to your liking, I say fix the problem (get a better source) not bad-aid fix, as when you fix the source/signal or put on better cd's on then the band-aid is still there in the signal path, with the chosen active pre.
Cheers George |
If I can take this tube microphonics a step further for those who have tube preamps.
Try this little experiment.
Take the top off the tube preamp and turn on your system and set the volume control at your normal listening position. Then with your finger nail flick down the tops of each tube and have someone listen at the speaker for any sound comming through the drivers. This will show which tubes are microphonic. And in the case of really bad ones it will be heard out in the kitchen. And really really bad ones will feed back to the said tube and cause a runaway effect that can't be stoped.
Cheers George |
Just to add to my previous post on microphonic tubes, if you hear a bonk in the speaker because of a slightly microphonic tube/s, this in turn when playing music will give and artificial effect of adding ambience/echo to the music, if that your bag so be it.
But I will always be in the camp of what goes in should be exactly what comes out, nothing added nothing subtracted, warts and all.
Cheers George |
I will try George's tube test soon.
Grannyring
It's not my tube test, it's a well known way of checking for microphonic tubes out in the field among techs without the use of expensive equiptment. You can aslo do the same with tubes in poweramps, not just pre's.
Cheers George |
Looks like Morley muso's equipment has jumped on the Lightspeed band wagon as well claiming better sound than with normal foot operated volume pedals, pity it's mono otherwise it could have been utilised.
http://www.diyaudio.com/forums/analog-line-level/187051-optical-attenuator-cheap.html#post2538367
Cheers George |
Knghifi
wish he could answer some of my questions. I'm still stuck on " ... is the closest you will get to playing music that's truest to the source (cdp, phono ect), and the truest way to hear the way the recording engineers wanted you to hear how they have recorded the music.
As it adds nothing and subtracts nothing ..." Knghifi
:
:
:
I'm still a little black and blue from the last mugging, I'll try for a few broken ribs this time around.
What comes after the source can only be corrupted by more electronics, it cannot fix a "bad" cdp dac or cd. It can only add corruptions, colourations, distortions and euphonics.
There is no magic preamp that can add bits or remove distortions from a cd or re-design a bad output stage in a cdp or dac, you can however falsely add colourations, euphonics, and tonally change things, but this is not "being true to the source".
I believe if you have a source that needs this sort of treatment, you need to get a better one, and one you like, without having to bandaid fix it.
Cheers George |
I would say that with this low 83db speaker, your amps input sensivtity would have to be between 1-1.5v to have compatibility with the Lightspeed Design.
But then seeing your amp is beastly 500watt rms, you would never use all this into (I take it their Maggies), so you could afford to loose some., I doubt you would even hit 250watt into them.
Cheers George |
Sources these days, esspecially CD can drive (have enough output) by themselves (without the need of a preamp) a poweramp into cliping, there is your strait wire with no gain (the interconnects only).
All we need is something so we can attenuate that source so it doesn't blow the amp up, we do not need a preamp that preamplifies again on top of the source that can already by itself blow an amp up with the amount of output it has.
Cheers George |
How to kill a thread "By Grannyrig"
Well this thread was going along nicely, until it turned into something personal, discussing the virtues of zero colouration of one resistor in the signal path against active preamps, tube or solid-state with 100's of different components in the signal path.
The simple truth, active preamps have colourations/distortions, because no two ever sound the same, (even though they measure flat 20hz to 20khz) that's why people keep changing them to get the colouration/distortion they can live with.
Cheers George |
Yep, no arguments from me there Paul.
If you have less than 2mt interconnects a CDP or Phono Stage with less than 200ohm output impedance (which most are) and a poweramp with 47kohm or more input impedance (which most are), there is no need for preamplification, as it is a backward step in transparency to be true to the signal from the source.
Preamps in my view are a left over dinosaur from the analogue days before the advent of stand alone high gain phono stages and cdp/dacs with high outputs.
Cheers George |
Atmasphere IOW, as you reduce the volume, **no matter how good the passive is**, you will loose bass and impact. This is why a buffer between the volume control and the interconnect cable is so important; as you reduce the volume setting the bass and impact is not affected.
Atmasphere
Not totaly correct sunshine, the ear becomes less linear at low levels and this is why the "loudness switch" was invented to boost the bass quite a bit and the highs slightly also, but these circuits like all active circuits, introduced too much of their own colouration/distortions as well for high end audio use. So now you'll only find them in car stereos and midfi home equiptment. |
To give some measured examples of interconnects
1: With IC's at 100pf per foot a 1mt = approx 300pf
this with the Lightpeed at it's highest output impedance has a HF rollof of -3db at 76khz
2: Same with 200pf per foot 1mt = approx 600pf
this with the same gives HF rolloff of -3db at 38khz
3: At 300pf per foot same gives a HF rolloff of -3db at 25khz.
Good interconnects are usually below 200pf per foot. As not to create low pass filters with passives and quite a few tube preamps.
Cheers George |
01-13-11: Almarg Ralph & George, thank you for the explanations concerning bass effects. Just to clarify, though, it should be noted that those effects are essentially unrelated to cable capacitance or cable length.
Best regards, -- Al Almarg
If it's not the Fletcher Munson low level listening curve, that your bass deficiency is stemming from, then I would seriously look at the value of your source series output cap, and make sure it is
1: That is at least a good quality polyprop cap
2: That it is at least 5uf or bigger.
Cheers George |
Some sources that have output coupling caps (series caps) and are in the order of 10uf, and are usually bi-polar (2 x 20uf electros back to back), but these sound like cr-p, better to use plastics (polypropylene) but good ones of this size are expensive, and big. All these caps are there to stop dc offset, which if gets through can take out amps and or speakers.
In my view the best source output stages are direct coupled (no cap) this is harder for the designer, as usually they need to also to design a DC servo to keep the dc offset at a minimum 1-5mV, but when done right sounds the best as they are more transparent and have less colourations of a cap coupled output, and have none of the low frequency limitations.
Then there's transformer coupled output stages, don't even get me started on those, with high and low frequency limitations, ringing and well as current limitations.
Cheers George |
01-14-11: Pubul57
George, does that mean that with some sources, Ralphs argument is correct due to the coupling caps? Pubul57
Yes, as I pointed out in my last post, but I will say that the said caps are usually big enough, not create any problem, except colourations, it's the moded one's that have had their caps and output stages modded that have to be sometimes questioned, are they large enough?
And Ralph's answer to your line stage question is answered with diplomacy, looks like he has the same view on coupling caps, (the best cap is no cap), Except still I do not believe anyone has yet built the transparent active preamp, as they all sound so different, more so I believe than the sound of different interconnects.
Cheers George |
01-14-11: Banquo363
Differently, remaining with this type of amp (with volume control), if we use it in conjunction with the LSA:
1. could we use the amp's volume control to mitigate any impedance mismatches between the two? I understand that these are two different variables, but they are physically related.
2. and, what adverse effects can we expect to remaining faithful to the source? 01-14-11: Banquo363
To your top question, yes if the volume control was 50kohm or more and there were no resistors to ground after it before the first active input stage. But saying that it will still be better if the pot was removed completely and a say 100kohm input resistor was inserted to ground.
Q1: No, not with the Lightspeed
Q2: If just using the pot on the input (no Lightspeed)it all comes down to the quality of the pot, as I say even the top ones Alps, Bournes, Penny @ Giles ect, all sound different because of light wiper pressure, materials used eg: wiper being metal on carbon or plastic conductive track, it is lightweight contact that mimics in a small ways a diode, (a proper diode in the signal path would try to rectify and AC music signal into DC), very bad. And it is this that a soldered in resistor gets rid of, the lightweight contact.
Cheers George |
The best possible way of getting the best transfer from source to poweramp is to have a cdp with a digital domain volume control, and have no preamp Lightspeed or whatever. This would be the ultimate, however a digital domain volume control in the cdp usually at less than 75% of full output starts to what's called "bit strip" and instead of getting full 16 bit resolution you start to loose resolution to 14 bit and so on the more you attenuate. I have yet to see one that doesn't "bit strip" at 75% or less level.
Even a question I emailed DCS (on behalf of a Lightspeed customer who believed his Puccini and Elgar did not "bit strip") about these very expensive units, and I asked "do these units bit strip at half volume?"
David J Stevens (of DCS) reply was.
"Hi George
Thanks for your interest in dCS and your questions.
As for the digital volume control, provided that you perform the volume adjustment at the correct point in the processing chain and you do it with sufficient precision then there is no reason to lose any signal.
Best Regards
David
Now this reply says to me that they do "bit strip"
Cheers George |
Pubul57
The one you have is already MkII. The only way I could improve to a MkIII is the versatility of the unit, that would be to have a discrete unity gain buffer (like Nelson Pass designed for the Lightspeed) on the output to lower the output impedance so it could then drive poweramps of lower than 47Kohm input impedance, and of course longer IC cables. But this would be a compromise as they would never sound as good as the MKII, I have done it, and would cost more, this is not what I'm about.
As for the switching for multiple inputs this is also not the purest way for me. If someone needs multipule inputs I suggest one of these.
http://www.decware.com/newsite/rotary.htm
That way you can still go direct to get the purest best sound when needed to show off your system.
The whole reason is to get the signal from the source to the poweramp with the least amount degradation.
And in case your wondering, all MkI's were converted to MkII years ago.
And if your wondering.
MkI = series resistor and shunt LDR "very good" still better than the best pots.
MkII = series LDR and shunt LDR "even better" more consistent i/o impedances, better bass and dynamics were noted by all previous mkI owners.
Cheers George |
To all who are not very technical, try to bear with me here.
The (diode thingy) "diode effect" is bought around by very fast music transients from the source, CDP or phono, which can be in the order of 100's of volts per micro seconds (volts per uS) these happen as the name implies in micro seconds.
When we view a static (constant) 1k-10k square waves on an ocilloscope all looks fairly fine going through a good pot (Alps Bournes ect).
But it's when you pulse the waves at micro seconds (transients) an then store it on a very good super fast digital storage ocilloscope play the micro second pulses back frozen in time, and magnify the corners of the waves it's then you see the ringing/occilations effect of the light wiper contact on the conductive track of the pot, (I have named it for want of a better word, "dynamic contact bounce")It's the wiper being ever so slightly bounced on the track.
When pressure is applied to this wiper (with in my case a wooden skewer) so the wiper presses down hard down on the conductive track, the ringing occilations stop when doing the same test. When a soldered resistor is in place of the pot it's not there either, when a Lightspeed Attenuator is in place of the pot it also is not there either.
All pots no matter how good bounce to a certain degree this is why they all sound different, they are all fundamentially flawed.
Cheers George
|
Sorry I should clarify a bit better the first paragraph above, the figure 100's of volts per micro second, is in reference to "rise time" in volts per micro second of musical transient figures, not what a cdp or dac can put out of it's output in volts, but rather the speed at which they can happen.
Like when very fast opamp specifcations of some of the better opamps, they can have rise times of 1000's of volts per micro second, but they only have 18v rails, these are very good for the i/v (current to voltage) converters after digital to analogue chip before the ouput stage in cdp's
Cheers George |
We are on the same wave length, like I've said on the Lightspeed Attenuator web site. The diode effect with contacts (potentiometers or single leaf switches) is a fact, and I would go further to say, that they are probably the worst component in the signal path.
Even rca's have this problem also, even with their much larger contact areas and tighter fitment, ever wonder why your system sounds better after a good clean up of all your interconnects rca plugs and sockets? Same deal, "diode effect" because of bad contact area contamination.
Cheers George |
Here's an interesting snippet from the latest Stereophile review of the DCS Debussy Dac which has volume control in the digital domain. Yet two stage adjustable analogue level output stage 2v or 6v, this means you can use it straight into a poweramp and only use the top 25% of the digital volume control, this way you should not be "bit striping".
Quote: The differences in sound between the Debussy into a preamplifier and the Debussy directly driving a power amplifier will depend on the preamp's transparency, dynamic capabilities, and overall quality. Sending signals directly to my Musical Fidelity Titan amp from the Debussy or through my reference preamp, a darTZeel NHB-18NS, I heard minimal differences. Given the darTZeel's price of $29,500, that's what you should expect. :Quote
Guess what the Dartzeel uses as a volume control? you guessed it. A Lightspeed setup, they tried to patent when they first released it but I quickly put a stop to it showing proof to the patent offices in Eu and USA that I first did the LDR volume control back in early 1974.
Cheers George
|
