That's OK, if there is measurable distortion which maybe is audible it's on the table as a cause for user preference, i.e. effects box.
BTW "collective of manufacturers" sounds very ominous.
You are conflating two different uses of the term "linear." A LDR's non-linearity will change the harmonic structure of the waveform i.e., add distortion. The non-linearity in potentiometers you refer to is only regarding the taper of the resistive element, so high quality a log taper pot will have excellent linearity, that is, it will not distort the signal, just attenuate. OK?
Mike
Last edited:
It doesn't take much to get a preamp/buffered volume control well below any sort of audible level, by anyone's standard.
CMRR being the hardest spec to hit, probably, but a nice line-level transformer or THAT receiver will go far towards that goal.
I know what he means. My point is that he's used the term "nonlinear" 93 1/2 times in this thread without explaining what is so nonlinear about a LDR or why a pot is so much better.
Interestingly, I've spent half an hour looking around on google and haven't seen anyone provide distortion charts for a regular old pot. I did find this however, which suggests that a good old alps blue isn't all that "linear" either. But, as I said, I can't find a whole lot more evidence one way or another.
Anyway, my point is that 'flat earth' types are supposed to be all about measurements and whatnot but I haven't seen DF96 or anyone else drop science about why LDR are bad and pots good.
The pot was measured in-circuit, so it appears that affected the result. If pots are as bad as claimed, it is hard to explain how any quality audio device incorporating one could possibly have the vanishingly low levels of distortion that most do.
Mike
Mike
You appear to be confusing two completely unrelated issues: what happens to resistance when you turn the volume knob, and what happens to resistance when you apply a signal. This confusion seems to be surprisingly common among LDR fans.jrubins said:
If you don't understand why nonlinearity in audio is a problem then you don't need to see an image; you need to read a good book.
Now someone else is using the "we" word! I suspect that other LDR people might prefer not to be associated with someone who, I seem to recall, tried to get output from an opamp input pin and sprinkled diodes throughout his circuits.Chris Daly said:
I never said that a pot is "so much better". It is better, but not hugely so. A well-designed LDR volume control could even approach (but never reach) the linearity of an ordinary resistor, provided that signal levels are sufficiently low - but this might bring problems of noise.jrubins said:
I don't know why people find it so hard to believe that a device which is somewhat nonlinear cannot improve an audio signal, but could damage it in a way which some people find pleasing to the ear. I guess most of them have economic interests, as they are either selling them or have spent money buying them.
Let us suppose that I gave in to the baying mob and actually listened to one of these devices. Let us further suppose that I actually liked the sound. Then I would not conclude that the LDR was better than a pot (as I know it cannot be); I would conclude that I too am one of those people who prefer slightly distorted sound.
Last edited:
Never said I was an LDR fan. I'm just making the point that you are making many claims without providing any evidence while condemning others who are reporting what they hear as not providing evidence.
Edit: to further clarify my point: I've never heard an LDR attenuator, but I suspect that you haven't either. Your response to the OP and others however seems to be "it can't possibly sound better, it's non-linear". However, you are not backing up your claims that any other form of attenuation is *more* linear, or that the type of nonlinearity displayed by LDRs is worse than any other type, or that nonlinearity is detectable by human ears, or that LDR nonlinearity is worse for music, or that the distortion of LDRs at listening levels is higher than pots, or that there isn't some other explanation such as input-output impedance matching and on and on. All you have done is said "LDRs are less linear" over an over and we should just take it on faith that makes them worse in actual circuits.
I, for one would rather have some evidence. Since there seems to be a dearth of that on this thread, I will seek some out myself and build my own damn attenuator and see if I can hear a difference. Maybe the 'nonlinearity' will be appealing. Remember we don't listen to spec sheets and distortion figures, we listen to music!
Sent from my iPhone using Tapatalk
Last edited:
Could you give a more generalized answer than that?
Any active audio preamplifier circuit is going to have a harmonic distortion profile which will include the 3rd harmonic that LDR's have.
And the usual attraction and benefit most people have with going to a passive preamp from an active preamp is that the passive is more transparent than an active, and I include myself as one of those people.
So how does a line stage preamp with gain stage/buffer circuitry become the standard bearer for transparency?
e.g. in a simple kit form: http://www.diyaudio.com/forums/vend...-differential-driver-preamp-0-000021-thd.html
LDR distortion is linked in Chris Daly's post above. Have a look at the differences. Without any sort of desire to be confrontational, I think we're operating on different definitions of transparency (I think input = output x scaling). I'm glad you enjoy the system you have.
Thanks for the link.
Yes, we likely have differing views of transparency.
Have you read post 134 about the measured distortion and it's effect?
Post 2 of that link, says it all, "how to configure the THAT Driver as a preamp by adding a volume pot" See how the volume pot is NOT measured, and just accepted as being what you have to use.
Imagine instead if it read as, "THAT Driver as a preamp by adding a LDR"
Last edited:
It is a known and accepted fact that LDRs are somewhat nonlinear as resistors. How nonlinear may vary from LDR to LDR, but none of them will be as linear as a perfectly ordinary resistor or pot. Anyone who disputes this is simply showing ignorance. When I first heard that LDRs were being proposed as volume control (a few years ago) I thought "Surely a light sensitive device like that is likely to be nonlinear? I will do some Googling." I discovered that my hunch was correct. You can do the same.jrubins said:
I never claimed that LDR nonlinearity is worse than other sources of nonlinearity. It is worse than linearity, as exhibited to a high degree of accuracy by ordinary resistors and pots.
I should not need to demonstrate that nonlinearity (if sufficiently great) is audible. This again is a known and accepted fact.
I have not claimed that LDR nonlinearity is "worse for music". Worse than what - spoken voice recordings?
As precision test equipment uses pots, and studios use pots then we know that the distortion of pots is sufficiently low for audio signals conveying music and voice. That means that the only question of interest is whether LDR distortion, although higher than a pot, is still low enough to be inaudible. If so, an LDR volume control will be indistinguishable from a pot. If not, a preference for LDR is a preference for audible distortion.
Any problems of impedance matching could be solved for any volume control method by adding genuine buffers. With pots these are usually not necessary, but can be added if they are. I am not aware of any reduced need for buffering with an LDR volume control.
Why are so many people wriggling and squirming to avoid or deny simple facts? Why are people so surprised to discover yet another example of how a small amount of low order distortion can be pleasing to many people (single ended triode amps are another example of this phenomenon, as are most 'tube buffers')?
Sigh. Who kicked your respective dogs? I was asked and answered a question--linking the one that came to mind first.
Yes, I read post 134, and there was no follow up/closure. So, what are we supposed to take from it?
Chris--ask Tom nicely to test with an Alps volume pot at half-volume in the recommended location. Expect vanishgly low distortion to remain, even if it goes up a decade.
Yes, I read post 134, and there was no follow up/closure. So, what are we supposed to take from it?
Chris--ask Tom nicely to test with an Alps volume pot at half-volume in the recommended location. Expect vanishgly low distortion to remain, even if it goes up a decade.
I use a stepped volume control with 1% film resistors as do others, from that article it looks like the LDR can do <.05% (which actually is probably inaudible to most folks). The film resistors are several orders of magnitude better, looking at THD alone (easily verified).
IMO until other effects like the excess noise behavior of the CdS are taken into account the story is not over.
Would it upset you if audible artifacts of whatever nature cause a user preference i.e. LP vs CD?
You guys have got this all wrong, there is a slight 2HD when the LDR's are over driven, and seen as a very slight rounding of the tops of the sine waves, just like triodes do.
This only happens when the input signal are way over the 2v standard at 0dbf for Redbook, the only ones I've seen to do this are bench test cd's that have 0dbf sine waves playing on rare high output >3v cdp's.
I have posted many pics and screen shots many findings here, you may have to sign in to view them.
Lightspeed Attenuator - Page 4
And Chris please do not include me as "we the collective manufacturers" as I certainly do not agree to any form of "forced quad matching" of LDR's as there are many problems with this, including what's being discussed here buy the doubters.
Cheers George
This only happens when the input signal are way over the 2v standard at 0dbf for Redbook, the only ones I've seen to do this are bench test cd's that have 0dbf sine waves playing on rare high output >3v cdp's.
I have posted many pics and screen shots many findings here, you may have to sign in to view them.
Lightspeed Attenuator - Page 4
And Chris please do not include me as "we the collective manufacturers" as I certainly do not agree to any form of "forced quad matching" of LDR's as there are many problems with this, including what's being discussed here buy the doubters.
Cheers George
Last edited:
while everyone is hooked on proving distortion figures. what the life span of carbon track and resistor stepped attenuator mechanisms before they start to show signs of wear which in turn leads to distortion. then at the same time whats the life span of 'LDR' volume devices before the resistors start to degrade.
lets say we took each device and used it from full off to full on 10,000 times, which would still be nearer the original figure of manufacture.
carbon track no matter how good all wipe the carbon from the surface of the wafer rendering it useless. resistor based stepped attenuator is like carbon track but the contacts will wear out instead of the carbon track rendering it useless or in need of repair. 'LDR' have no moving parts so its down to how long the resistor/led life span is, ok so the LDR still rely on a conventional pot or rotary encoder to control the currant to the led so will need replacing as the pot/encoder wear but you can use the cheapest parts you can get your hands on as a replacement.
lets say we took each device and used it from full off to full on 10,000 times, which would still be nearer the original figure of manufacture.
carbon track no matter how good all wipe the carbon from the surface of the wafer rendering it useless. resistor based stepped attenuator is like carbon track but the contacts will wear out instead of the carbon track rendering it useless or in need of repair. 'LDR' have no moving parts so its down to how long the resistor/led life span is, ok so the LDR still rely on a conventional pot or rotary encoder to control the currant to the led so will need replacing as the pot/encoder wear but you can use the cheapest parts you can get your hands on as a replacement.
10,000 rotations? That's easy to do. When I was making potentiometers, our test was 10 million. FWIW, the Alps pot in my linestage has close to ten years of daily use on it, no problems at all.
edit: The cheapest pots with the softest tracks that I could find were rated at >25,000.
edit: The cheapest pots with the softest tracks that I could find were rated at >25,000.