DAC straight into power amp - not impressed!

[AlexM]

Hi,

My C-J pre-amp left for a service today, and while it is away I have been forced to plug my CD/DAC (with digital volume) straight into the power amp.

I am suprised to find that the sound is detailed with good stereo imaging and timbre, but a bit rolled off in the treble, and the bass is quite a bit less weighty and well defined and sounds somewhat muddy and opaque. Given that a pre-amp essentially buffers the signal and isn't supposed to add-much of itself, why should this be?.

One of the features of my pre-amp is a very low output impedance and lots of voltage drive - is the output stage of a DAC going to struggle driving the high-impedance input of a power amp? I wouldn't have thought it would be working any harder then when driving the pre-amp itself.

I suppose it could be that the digital volume control has to be significant attentuated to keep the volume sensible (-40Db).

Any ideas what is happening here?

Alex

[MartinT]

Some of us have advocated active preamps for the very reasons you state over the years. I don't like 'passive' setups as I find the sound lightweight and lacking in drive. It's that drive that a preamp is capable of that controls the power amp to within an inch of its life, giving drive, dynamics and bass weight. Attenuating via a digital control certainly wouldn't help as they work by throwing bits of resolution away.

[AlexM]

I really am quite suprised how much impact and bass drive is lost, as well as treble sparkle and detail. The resulting sound is a little mid-forward and insubstancial.

I am interested in why this is - I'm not driving long I/Cs, the input impedance of my power amp is 100Kohm, and the output stage only need to provide 1.2v to drive the power amp to it's full rated output.

As an experiment I will try adding some attenuators and running the digital volume control higher to see if the loss of resolution is caused in part by that. Cambridge Audio say that the volume reduction algorithm they use in the ATF2 doesn't result in bit truncation or loss of quality and in fact, they bill the component as a digital preamplifier. Whatever they say, I defintely think it sounds better with valve pre-amp in the loop!

[Yomanze]

I really am quite suprised how much impact and bass drive is lost, as well as treble sparkle and detail. The resulting sound is a little mid-forward and insubstancial.

I am interested in why this is - I'm not driving long I/Cs, the input impedance of my power amp is 100Kohm, and the output stage only need to provide 1.2v to drive the power amp to it's rated input.

As an experiment I will try adding some attenuators and running the digital volume control higher to see if the loss of resolution is caused in part by that. Cambridge Audio say that the volume reduction algorithm they use in the ATF2 doesn't result in bit truncation or loss of quality and in fact, they bill the component as a digital preamplifier. Whatever they say, I defintely think it sounds better with valve pre-amp in the loop!

I'm not surprised. On paper a passive or DAC volume should be absolutely fine, but I find a loss of dynamics, drive and a feeling that I need to turn the wick up a bit with passives / no preamp whereas a good active pre is the same at all volumes.

[John]

Passive pre can work fine in the right system No one would describe my system of lacking resolution but get it wrong and the sound is flat and lifeless
It not a good idea to control volume from the computer is not ideal

[Reffc]

Hi,

My C-J pre-amp left for a service today, and while it is away I have been forced to plug my CD/DAC (with digital volume) straight into the power amp.

I am suprised to find that the sound is detailed with good stereo imaging and timbre, but a bit rolled off in the treble, and the bass is quite a bit less weighty and well defined and sounds somewhat muddy and opaque. Given that a pre-amp essentially buffers the signal and isn't supposed to add-much of itself, why should this be?.

One of the features of my pre-amp is a very low output impedance and lots of voltage drive - is the output stage of a DAC going to struggle driving the high-impedance input of a power amp? I wouldn't have thought it would be working any harder then when driving the pre-amp itself.

I suppose it could be that the digital volume control has to be significant attentuated to keep the volume sensible (-40Db).

Any ideas what is happening here?

Alex

In theory, MOST modern DAC chipsets these days are perfectly capable of supplying a low distortion signal, at between 1.5 and 2V rms on full gain with less losses than analogue attenuation stages providing attenuation is within say -30dB and 32 bit digital attenuation devices are utilised. The problem with some is the amount of gain or where lower than 32 bit is used into high impedance passive attenuation circuit. That though is a bit of an over simplification so someone with more in depth knowledge may be better able to explain further. There's a good post on the subject (post#223) HERE (http://www.pinkfishmedia.net/forum/showthread.php?t=130437&page=15).

In general though, I have always preferred digital sources played via an active amp stage even though in theory, S/N is going to be lower overall than direct from the DAC into power amp. This isn't usually noticed much as the power amp tends to have the lowest S/N ratio within the source/amplification stage. My CDP has an output level control, ostensibly for recording purposes, but can drive the power amps without any loss of resolution or apparent dynamics on its own. From what I've read on the subject, digital sources tend to work best into low impedance active pathways and not into passive high impedance pathways so I'm not surprised that some report very mixed results connecting DACs straight into passive pots.

[Yomanze]

Yes a lot of variables Paul as my DAC is around 60 Ohms output impedance (including output transformers) specified to drive 'any load you could care to think of' according to the manufacturer. Still, I do prefer active pres even with this DAC. Having said this something like a Bushmaster with a passive pre opens up more matching issues (even if this combo can work well with some systems).

[Yomanze]

Passive pre can work fine in the right system No one would describe my system of lacking resolution but get it wrong and the sound is flat and lifeless
It not a good idea to control volume from the computer is not ideal

Am sure those huge open baffles are plenty dynamic as it is so yeah as usual comes back to synergy.

[AlexM]

In theory, MOST modern DAC chipsets these days are perfectly capable of supplying a low distortion signal, at between 1.5 and 2V rms on full gain with less losses than analogue attenuation stages providing attenuation is within say -30dB and 32 bit digital attenuation devices are utilised. The problem with some is the amount of gain or where lower than 32 bit is used into high impedance passive attenuation circuit. That though is a bit of an over simplification so someone with more in depth knowledge may be better able to explain further. There's a good post on the subject (post#223) HERE (http://www.pinkfishmedia.net/forum/showthread.php?t=130437&page=15).



Interesting. John Westlake seems to be saying that using a digital volume control on a source component running directly into a power amp is the lowest noise/highest performance solution. That as may be (from a theoretical perspective), but in practice an active preamp/buffer stage seems to be preferable to most. I thought that most preamp inputs were high impedance, and that it was important for the output impedance of the source to be low - is this not the case?. In what way does a power amplifier input differ from a preamp input? I mean this in terms of the load that the source component output buffers 'see'.

Regards,
Alex

[nat8808]

The output stage of the source still had to be able to drive the power amp well though. Essentially there is no difference between an active line stage in a pre-amp and the output stage of a source component... other than how it is designed and what it is specifically designed to drive!

I guess the 851 isn't designed in a way that drives your Conrad-Johnson sufficiently well.

Of course, err.. opening up ALL possibilities, free thinking and all that.. perhaps you're experiencing something more neutral and your previous set up was slightly top and bottom heavy?

I remember a famous speaker designer saying that that was many people's reaction to his speakers that he'd put so much effort into being completely flat responsed and neutral. As they came from other speakers accentuating certain extremes, they were hearing lack of sparkle and bass. Sticking with that neutral goal however, over time people adjusted their ears.. and now regard the designs as being neutral.

[MartinT]

Essentially there is no difference between an active line stage in a pre-amp and the output stage of a source component... other than how it is designed and what it is specifically designed to drive!

The key difference is in output impedance. A passive stage uses the source component output stage and puts the signal through a volume pot. So the output impedance is increased and is variable depending on the volume setting. This could potentially give changes to the sound according to the volume setting.

An active stage sets the volume earlier in the gain stages and essentially has a low and constant output impedance.

Your final point is a good one: a power amplifier designed for passive input will perform better with a passive stage. Watch the screening on your interconnects, though, as the higher impedance could allow noise and hum into the system.

[realysm42]

Even with a tvc line stage, my system sounds significantly better withone in than without, I was surprised at how much!

[john dolan]

I run my M-dac full volume + 3 through a Audio Synthesis Passion passive preamp because to my ears that sounds better than M-dac direct to the amps.

Running the M-dac full volume you are hearing it at its best and going through the passive the signal is only going through a single Zfoil resistor per channel.

[AlexM]

The output stage of the source still had to be able to drive the power amp well though. Essentially there is no difference between an active line stage in a pre-amp and the output stage of a source component... other than how it is designed and what it is specifically designed to drive!

I guess the 851 isn't designed in a way that drives your Conrad-Johnson sufficiently well.

Of course, err.. opening up ALL possibilities, free thinking and all that.. perhaps you're experiencing something more neutral and your previous set up was slightly top and bottom heavy?

I remember a famous speaker designer saying that that was many people's reaction to his speakers that he'd put so much effort into being completely flat responsed and neutral. As they came from other speakers accentuating certain extremes, they were hearing lack of sparkle and bass. Sticking with that neutral goal however, over time people adjusted their ears.. and now regard the designs as being neutral.

Hi Nat,

It well known that many speaker designers deliberately engineer a rising treble characteristic or exaggerated bass response and that some people have aclimatised to that, and so a speaker with flat response curve may sound dull or bass-light at first aquaintance and that time is needed to get used to the 'non-loudness' version :)

My pre-amp has a pretty much ruler flat frequency response out to about 50k, so I don't think it is simple as that. The line-output impedance is 95 Ohms, which is pretty low, and will be able to drive the 100kOhm input of the power amp without any problems whatsoever. The output impedance of the 851c is <50ohms, which also suggests plenty of ability to drive a high impedance input too. I have read that this is typical in domestic audio, i.e. low output impedance/high input impedance - the load is essentially a bridging load and doens't actually load the source at all, and this maximises the voltage swing across the signal and ground of the power amp's input, with very little current flow.

One area where the specs don't guide is highlighting at what frequency the output impedance is measured - if it varied significantly with frequency and between the source and pre-amp, then that could account for differences in frequency response. To be honest I don't think that is happening here as the 100KOhm power amp input resistance is so much greater than the source output impedance.

... which is all well and good, but I am left wondering why a pre-amp with a flat response sounds so different to a directly connected source given that the interface to the power amp is 'ideal' in terms of impedance. It actually sounds 'fine', but a bit bloodless, lacking in 'oomph'. I am just curious as to why this is... this is just a temporary arrangement while my preamp is away for a service. It will certainly be interesiting to see what I think when it is back in the system - I hope I like it!.

Regards,
Alex

[AlexM]

Ok, now I'm seriously confused... I added some Rothwell 10Db attenuators at the power amp input, and it is sounding quite different again (and quite a bit better).

Mid range has opened up, and the treble is cleaner and better defined. Mid-bass is also less opaque - bass textures seem much cleaner.

Happy now, but haven't got a clue as to what is going on!

[Macca]

Play some varied material and see if those Rothwells give you any issues with distortion. Not always noticeable with some music but I have always had issues when I have tried to use them even though at first things sounded superficially better.

[AlexM]

Martin,

I've checked for any form of distortion with some recordings that I know well and I'm sure that it's as clean as a whistle. What kit were you using that caused that issue?.

Regards,
Alex

[Macca]

Martin,

I've checked for any form of distortion with some recordings that I know well and I'm sure that it's as clean as a whistle. What kit were you using that caused that issue?.

Regards,
Alex

I've tried them with all sorts - too much to list - to try to resolve gain issues and never got a good result. Some people swear by them, I know. If they are working for you that's great it is an easy fix to what can be a big problem.

[DSJR]

Older amps and some modern ones are DESIGNED to just have a pot in front with no line buffer. Many modern ss amps NEED an active line buffer to feed them. TEC do a very cheap one (well under a hundred quid) which would easily do as a spare, although it won't have the cachet or 'personality' of a more expensive model.
This is one area where some technical expertise is required if very variable results aren't to be experienced I fear.

[AlexM]

Thanks All - it was just my curiosity more than anything.. The direct connection into the power amp works well enough for now. Hopefully my preamp will be turned around quickly and normal service will be resumed...

[StanleyB]

I had the same problem with the variable output on the TC-7520 and Caiman. It is one of the reason why I use the fixed output of my Bushmaster into the variable input of my power amp. Works a treat that way. I even sold my preamp afterwards to someone on AoS.