Digital Volume Control - understanding how it works.

[Monkeyboi]

I know because I built 2 identical preamps and found that the newer one I built destroyed the 1st one in SQ that was completed 6mths earlier! The culprit: volume pot !!!

I'm shocked, shocked and yet again shocked to read the older pre-amp didn't present a better SQ :eek: . After all it's had an extra 6 months of operational time to "burn in".

BTW, if an analogue pot doesn't last 6 months of use it was either crappy to begin with or subjected to an IKEA mobilfactor test.

Cheers,

Alan R.

[Addicted to music]

Using the reasoning in the opening post of this thread will I get a better result in using less digital attenuation of the NAD M51 and using passive analog attenuators at input to the power amp?

I doubt it, I have been listening to the M51 at -54DB today and yesterday morning between 5am - 7 am, this is because everyone apart from my son and I are awake. If you can distinguish and be able to tell distortion levels at -70db to -45db than that saids alot about my set of ears

[Addicted to music]

I'm shocked, shocked and yet again shocked to read the older pre-amp didn't present a better SQ :eek: . After all it's had an extra 6 months of operational time to "burn in".

BTW, if an analogue pot doesn't last 6 months of use it was either crappy to begin with or subjected to an IKEA mobilfactor test.

Cheers,

Alan R.

Welcome to the real world! I deal with electronics failing on a daily basis and even today! And yes when you use a pot it is as you say subjected to the IKEA test!!! :lol:

[Jeff65]

....Potentially yes BUT only when listening at less that say -50Db - and that's the sticking point - you're unlikely to every listen at such a low level anyway.

This depends on the voltage output level of the DA converter and the analog gain in subsequent stages of the system. A lot of systems have much more gain than is necessary.

Using a digital volume control in a system with too much analog gain is not optimal.

[Guest Misterioso]

Using a digital volume control in a system with too much analog gain is not optimal.

Unless you can adjust the analog gain...

[Jeff65]

Unless you can adjust the analog gain...

True enough, but many audiophiles do not consider the amount of gain in their systems. All should do so, but this is particularly important if one wishes to use a digital volume control.

[Addicted to music]

Hi pchan.

1. The settings are quoted in the table.

2. Won't make much different 6 months work. Most users, myself included will only adjust the volume from time to time. Though I do have a 3~4yo Troubadour with an Alps volume control in it. If I get some spare time I'll remove it and test.

Usage and consistency will be fine for a while. It will not last like a digital volume control, but that is the price you pay for sound quality.

3. Heat isn't an issue. I'm a very speedy solderer

I'm not sure how you came to your conclusions in you "FWIW" para.

As the volume reduces, it gets closer to zero volts, not dc. It will still be the music (ac) signal, just at a reduced level p-p.

Why would the highs be affected at just say 16khz to 20 khz. A pot can transfer frequencies up to and beyond 1mhgz. Anyway the contact resistance doesn't come into it. At low volume, say a 0.05vac signal, driving a typical small signal valve. The input grid resistance will be well over 1meg ohm, and as high as 2 in some cases. The voltage drop across any contact resistance will be so insignificant I doubt it could be measured, maybe a few pico-volts. The straight "curve", will look, like a straight line 20hz to 20khz.

Cheers, Earle.

Hi Earle,

@1

Sorry I didnt make myself clear, what resistance was the pots set at when you took these measurements, ie: 10, 200, 3K, 10K, 100K ohms

@2

Deterioration is noticable only if its constantly rotated and electricity applied.

@3

No matter how quick you solder to are going to hit that point at 200-300 degree C, because it is connected to a wafer with a metal point it is going to transfer most of that heat onto the wafer close to that point.

In reference to my FWIW,

if 2 preamps are built with identical components at 6-9mths apart and you find that the 1st preamp doesnt sound as good as the 2nd. You investigate and find that the volume pot fixed the SQ issue of the 1st built preamp, then I would conclude that the volume pot was the issue due to wear and tear. This is not a theory but actual finding,

The reduction on the highs, well I dont have an explanation for that, that what was observed on a worn pot compared to a new one. However the explanation for the reduction of bass I believe is valid as bass energy is close to DC so any increase in contact resistance is going to effect the overall transfer of audio regardless of 1-2M ohm input resistance. In your table there is a difference of 0.4db between the Goldpoint and the no name pot. But as you decrease the input the goldpoint doesnt seem to deliver as the no name carbon base pot!!!! Very interesting considering the difference in cost. Note also that the figures vary in db between them all inside the table.

[ehtcom]

Hi pchan.

Didn't measure the resistance. All the pots were 100k audio taper. Using google, find a db to volt conversion calculator, then math out a resistor divider network and you will be able to calculate the values.

Agreed, some pots wear quicker than others. Stopped using a couple of the above for that very reason.

No, most of the heat stays with the solder lug.

Were they really identical components? Did you measure them all? Transistors?, opamps? tubes? some caps have tolerances -10 to +30%. Maybe you scored a partially faulty pot with the first unit?

The measurement isn't an indication of frequency response (all tests done at 1khz BTW) it indicates signal to noise ratio. I think you will find the frequency response remains very flat.

20hz it nowhere near DC, it's 20hz AC, a constantly varying voltage.

I did notice that the signal to noise ratio deteriorates with the introduction of a pot. No surprise there, every component comes with added noise for free.

Cheers, Earle.

Edited July 31, 2012 by ehtcom

[gainphile]

Surely any perceived "lack of highs" can be measured at the tweeters, and traced back accordingly?

Even 0.1db is readily visible with 5ms gating.

[jkeny]

The original article is wrong & misleading, I believe & if you read the following posts you will see that it is corrected almost immediately by Patrick Dixon's posts

If you play back a 16bit audio file through a 16bit digital pipeline (this includes all the digital channels that it has to pass through before it reaches the D/A stage) then it has a SNR somewhere <96dB

If you play back the same 16bit file through a 24 or 32bit digital channel which has a SNR somewhere < 120dB. it's not "magically" adding anything other than playing it in a higher resolution channel.

So his statement

"

By "expanding" a 16 bit signal to 24 bit, all we are doing is saying "these 16 bits go in the most significant slots of the 24 bit word". We haven't improved the SNR of the signal, any more than you can "enhance" a digital photo the way they do on CSI."

is misleading. You have improved the SNR of the channel you are using to play back. Nobody is saying you are getting an "enhanced" audio signal but there is no doubt that using digital vol control on 16bits channel will lose data bits immediately - 1bit for each 6dB of vol reduction. Using 20 bit channel you have 4 bits that can be discarded before bits with data are lost or 24dB of volume reduction available.

Edited July 31, 2012 by jkeny

[Addicted to music]

Earle,

All the components where the same, no I didnt measure them before placing them in the circuit, but replacing the pot in the first did bring the SQ up to where the second unit's performance and matched it and that was audiably clear. What I did do was replace a pot within a 6mth period and also noted the difference. Its simple wear and tear. If you open one up the wear marks on the taper are evident, cleaning it worstens the pot to the degree where it develops cracking noises and that not with DC in the circuit. You are able to hear the pot turn audible through the speakers.

I began to use plastic film types and they are worst than carbon types. Any type with the letter ALPS marked on it where the ones that sounded the best and lasted the longest. I was buying an ALPS 100k thru Tandy Electronics when they where around @$4.00. Infact I went to every Tandy store in Melb and purchased the stock they had!

I have also a stock standard Alps that came with the Stax Lambra II, this pot is starting to get audiable as you rotate it!

My next step was to go digital and use a intergrated control such as the WM 8816 or one of these: http://www.diyhifisupply.com/node/734. & http://diyclub.biz/c...products_id=112

Edited August 1, 2012 by pchan

[BradC]

What if the noise is inaudible?

Exactly, even though the SNR is degraded with digital attenuation, the SNR is so much in excess of the SNR of the room (and your ears) that you can afford to lose SNR without perception of any ill effect

[andreasmaaan]

I've nothing specially against digital attenuation, and in fact I use it in my system. However, I think the original poster's argument is that when a signal is digitally attenuated, the noise from the ANALOGUE output stage of the DAC is not so attenuated (ie remains constant), resulting in a poorer overall SNR of the signal from the DAC's outputs.

With an analogue attenuator, the noise from the DAC's analogue output stage is attenuated along with the signal. However, added to this is the analogue attenuator's own noise and any noise/distortions introduced by additional cables/connectors.

Presumably, the best solution is system-dependent.

[Arg]

The original article is wrong & misleading,...If you play back the same 16bit file through a 24 or 32bit digital channel which has a SNR somewhere < 120dB.....

Thank goodness. I thought I was going nuts.

[jkeny]

Actually, I came across this video presentation form CTO of ESS & he makes the case that analogue vol control is better than digital.

However, (isn't there always one), it is not easy to implement an analogue vol control that beats the -135dB of the ESS DAC vol control.

So like a lot in this audio field, it's easier to achieve reasonably good vol. control in digital as it is to achieve reasonably good audio - that's probably one of the reasons why digital is so popular?

Edited August 1, 2012 by jkeny

[Addicted to music]

Actually, I came across this video presentation form CTO of ESS & he makes the case that analogue vol control is better than digital. [media=]

[/media]

However, (isn't there always one), it is not easy to implement an analogue vol control that beats the -135dB of the ESS DAC vol control.

So like a lot in this audio field, it's easier to achieve reasonably good vol. control in digital as it is to achieve reasonably good audio - that's probably one of the reasons why digital is so popular?

Interesting that you have posted this article JK,

I was beginning to disbelieve every claim made by ESS as there products specification are not published in the public domain, there latest venture in the Renaissance Dac, there head MM: Mark M made a false claim on a AD797 opamp and when finally challanged on the HeadFi forum quickly corrected there claim in there marketing material.

In my previous post I have made it obvious that using traditional volume control with traditional mechanical point contact is one of the most destructive component that could ever be used to control volume in all amplifiers. In fact its the first thing that destroys the integrity of original signal coming into the amplifier. In the 80s to 2005 alot of amp manufacturers such as Marrantz and Mark Levenson are using very expensive ways of implementing this via integrated circuits such as the WM 8816 or the Cyrus chip, this progress to stepped resistors using relays where WFS and many other manufacturers uses. Silicon Chip also had a project to control volume in a digital manner but the chip they used was no longer produced because the cost was $50-60 a chip and to build it to there specs was going to total up to $500AUS in parts alone. It isnt till dac manufacturers place digital volume control in there packages that have made this cost effective but at a reduction of bit resolution and hence the debated topic of Analog Vs Digital Control. So what ESS is say is no secret, digital volume is popular because of the destructive nature of normal volume potentiometers, and that includes all of them regardless of make and material used.

[jkeny]

I agree that analogue volume control to be better than good, well implemented digital volume control, needs to go to extraordinary lengths to beat the specifications of good digital vol control. The mechanical point contacts can be improved with mercury switches, which switch in low noise resistors in a stepped attenuator or relay based volume control. I've also seen the use of Light Dependent Resistors (LDRs) as volume controls (Lightspeed) to avoid this mechanical point contacts that you mention & it seems to have a big following but I don't believe that it's specs are anything special. However, if there is anything to learn from the video it's that specs & measurements are not the final arbiter of sound quality.

Edited August 1, 2012 by jkeny

[jkeny]

BTW, I disagree with Martin Mallinson about both their Jitter reduction ASRC technology - their DACs sound better when their ASRC is bypassed by being driven synchronously. A number of people have verified this & I therefore I wonder have they listened & heard this? I guess it's asking a bit much for them to admit this aspect but even Dustin Forman (one of their chief designers) has said it sounds better, AFAIK?

[aechmea]

And of course in the real world, "it all depends".

In my system's particular case, I have to run the (digital) volume at -5dBFullScale to get about 80 - 85 dB SPL at the listening position from a digital source most of which seem to be recorded at about average -3dBFS (with peaks at 0dB presumably). The analog part of the volume system is unity gain. The speakers are inefficient (85dB), the power amps are insensitive (2V input sensitivity) and the room is large (115 cu metres). So what this means is that I am probably not dropping at bit at normal listening levels (even if it were a 16 bit system) and if I turn the vol down it is unlikely that I will drop any more than 1 bit, maybe 2.

So if I do drop a bit, that means it is the lowest order bit, and the SNR or dynamic range or whatever is reduced from 96 to 90.

OK, so if I listen at 80dB and the noise floor is -90dB (given that there is an approximation between SPL, sound intensity and power dBs) does that not make the noise lower than 0 (threshold of hearing).

Out of curiosity I played something at -50dBFS on the digital volume as suggested a few posts ago. The result was that I could barely hear any useful signal above normal household ambient noise. I also added an old but respected pre-amp (ME25 set to roughly unity gain) to the system prior to the digital section, and the noise as observed in the digital domain could be seen bumping along at about -80dB. Sure the pre-amp is old and there was an AD conversion step in there but it still gives a ballpark idea of preamp noise.

Then add to that the gross distortions added by the room (likely to be +/- 15 dB at some frequencies) and add some eardrums whizzen by age and its all a bit of a moot point. Therefore I echo other comments about 'is it audible and therefore does it matter at all' and 'it is system dependent'.

[BradC]

That's the point, a quiet room will have a noise floor around 30-40 dBA. Playing at 90dBA only requires an SNR of 50-60dB.

You can afford to lose 10's of dB in SNR and not notice a difference.

If you play silence on your system, that is your noise floor. Turn the digital volume up and down, can you hear the noise? If not, digital volume control is not an issue.

By the same argument, you can never actually use a dynamic range of over 90dB. And since most music, except the best recorded classical music, will barely have a dynamic range over 10dB.

And of course in the real world, "it all depends".

In my system's particular case, I have to run the (digital) volume at -5dBFullScale to get about 80 - 85 dB SPL at the listening position from a digital source most of which seem to be recorded at about average -3dBFS (with peaks at 0dB presumably). The analog part of the volume system is unity gain. The speakers are inefficient (85dB), the power amps are insensitive (2V input sensitivity) and the room is large (115 cu metres). So what this means is that I am probably not dropping at bit at normal listening levels (even if it were a 16 bit system) and if I turn the vol down it is unlikely that I will drop any more than 1 bit, maybe 2.

So if I do drop a bit, that means it is the lowest order bit, and the SNR or dynamic range or whatever is reduced from 96 to 90.

OK, so if I listen at 80dB and the noise floor is -90dB (given that there is an approximation between SPL, sound intensity and power dBs) does that not make the noise lower than 0 (threshold of hearing).

Out of curiosity I played something at -50dBFS on the digital volume as suggested a few posts ago. The result was that I could barely hear any useful signal above normal household ambient noise. I also added an old but respected pre-amp (ME25 set to roughly unity gain) to the system prior to the digital section, and the noise as observed in the digital domain could be seen bumping along at about -80dB. Sure the pre-amp is old and there was an AD conversion step in there but it still gives a ballpark idea of preamp noise.

Then add to that the gross distortions added by the room (likely to be +/- 15 dB at some frequencies) and add some eardrums whizzen by age and its all a bit of a moot point. Therefore I echo other comments about 'is it audible and therefore does it matter at all' and 'it is system dependent'.

[Arg]

Good comments above, I just want to add that the best solution is good gain management through your system.

With good gain management you will have:

• the minimum number of attenuators (ideally only one) in your system, digital or analogue;
  
• attenuators near their 'max volume' setting; and
  
• a small range of attenuator (volume) settings no matter what source unit is switched in.
  

With good gain management it will hardly matter what volume control you use, as long as it is good quality and well implemented.

[Addicted to music]

From my experience I have always wanted to to use a intergrated volume chip like the WM 8816 or a step resistor relay control. Its the 1st I have heard of mercury switching jkeny and Im glad that you have mentioned it, as you can see the innovation is great to see but, mercury is not a very good conductor as: http://environmental...electrical.html as you can see from this table. So the benefits of using mercury defeats this purpose.

I also though I give this a mentioned that on the EE mini max dac forum on Headfi, there are cases on this site that they have removed the volume pot and the SQ improves, well guess what the latest EE mini max dac plus doesnt come with a volume pot!

There is a lot about the debate of bit length resolution in using the inherent feature of digital control and the reduction of bits to achieve resolution. FWIW here is what I have set up and loving it:

Sony s470> NADM51> 240W power amp.

I am listening to this with the NADM51 set to -54db to -52db and wow, I cant tell the difference at this level or -20db in resolution as this set at 16bit.

The other advantage that a mechanical contact point cant provide is channel balance at all volumes!!!!

Removing the pre with contact point volume control regardless of using a typical Eltronics motorised no name brand or Alps, I get better resolution, imaging, clarity and the list goes on.... I refused to go back to a pre, the benefits of the M51 and utilising the digital volume control is the best investment I have ever made period!!!

This is what I have experienced and recommend, but its whatever tickle your toes So if you what to control volume with a mechanical contact point be my guest

Edited August 2, 2012 by pchan

[andreasmaaan]

I had a NAD M51 in my system for a couple of months and I can second that it has an excellent digital volume control. Compared to the two software attenuators I A/B'd it against (jriver and jplay) it always sounded better. It can't simply be down to the number of bits, as jriver (which unsurprisingly sounded the worst) uses 64 bit processing in 'internal volume' mode, vs. NAD's 35 bit processor. So I suppose NAD must be doing something else right...

(It wasn't my favourite DAC overall though.)

[Dr X]

I had a NAD M51 in my system for a couple of months and I can second that it has an excellent digital volume control. Compared to the two software attenuators I A/B'd it against (jriver and jplay) it always sounded better. It can't simply be down to the number of bits, as jriver (which unsurprisingly sounded the worst) uses 64 bit processing in 'internal volume' mode, vs. NAD's 35 bit processor. So I suppose NAD must be doing something else right...

Perhaps the JRiver volume is superior but you simply prefer the sound of the inferior (NAD) one?

[andreasmaaan]

Perhaps the JRiver volume is superior but you simply prefer the sound of the inferior (NAD) one?

Will let that one through to the keeper.