Two separate topics discussed by Martin Mallinson of ESS from RMAF:
RMAF11: Digital Jitter and Volume Controls, Martin Mallison, CTO, ESS Technology - YouTube
RMAF11: Digital Jitter and Volume Controls, Martin Mallison, CTO, ESS Technology - YouTube
You mean -135db, right?
As it becomes obvious from the video and confirmed by many listener's ears, in the context of a Sabre based dac it is quite pointless to add an analogue volume control. In case of an active, low output impedance IV stage, the best solution is to forget about analogue VC or any kind of additional preamp circuitry. In some systems this may even be possible with passive I/V or voltage out solutions.
Thanks Jack, that's a nice presentation.
The digital volume control part fits nicely with my experience (so I like it) in that high bit digital volume controls are hard to beat. As I tend to use no more than 10dB attenuation and 24 bits, it works very well.
The jitter part of the presentation was good, but more complex. I have built DACs with ESS chips and they do have a "different" sound to them.
BTW, is that Jan Didden asking a question?
The digital volume control part fits nicely with my experience (so I like it) in that high bit digital volume controls are hard to beat. As I tend to use no more than 10dB attenuation and 24 bits, it works very well.
The jitter part of the presentation was good, but more complex. I have built DACs with ESS chips and they do have a "different" sound to them.
BTW, is that Jan Didden asking a question?
Yes
I know the volume control in a Sabre DAC is awesome but I still want an analog volume control for analog sources
As far as I can tell this method uses potentiometers? Unless I am missing something? I don't think this is what Martin is talking about in the video at 12:00 - 13:30
BTW, not very on topic here, but here: RMAF11: Founding Fathers of High-End Audio - YouTube you'll find Bob Cordell (2nd from left), Jeff Rowland (right) and others. Not very well moderated though.
jan
jan
you should see the Accuphase AAVA volume control.
imagine 16 current sources multiplied by 4 (x2 for balanced, x2 for stereo).
each is an improved Howland topology implemented with a precision opamp. none of the 'usual' suspects but precision, low-noise types nevertheless. the positive feedback path contains another opamp configured as repeater. each of the two opamps has a buffer at the output (2 transistors). and we should remember that the Howland current source is sensitive to resistor values. so we have hundreds (yes, do the math) of precision, low-noise resistors in the signal path.
somehow it achieves 120+ dB SNR. and the AAVA is only a portion of the preamp.
kinda makes me laugh thinking at the passive volume controls some are in love with. it also makes me wonder if sonic quality has to do with this choice at all. pretty hard to DIY something similar to AAVA. better to talk about how passive sounds better.
just sayin'.
Somehow an overcomplicated volume control circuit doesn't make me laugh at all. Without getting into the specifics of the circuit and examining the potential pitfalls it's very hard to pass judgement.
Is it actually better sounding than a Shallco ladder switch populated with Texas Components naked resistors using the same output buffer? It's obviously much cheaper and easier to remote control. But actually better?
See my thread here, no one replied though. You can see some interesting work by bpp here: http://www.diyaudio.com/forums/vend...control-module-ultimate-attenuator-audio.html
See here: http://www.diyaudio.com/forums/anal...aava-accuphase-analog-vari-gain-ampli-er.html
I have the complete schematic, not sure if I'm allowed to post it (it's not patented etc).
I will post it if someone explains the policy for posting schematics of commercial production gear.
Last edited:
upon further consideration I really don't think the detailed schematic is something that belongs in the public domain.
the basic idea is obvious from the principle diagram.
what's notable is that they went to some lengths with the implementation in order to maximize linerarity. far from what one would expect and the absolute opposite of the typical internet wisdom, short signal path etc. and also very different from the "let's have tons of that, that and that" approach (where "that" typically stands for insane slew-rate, insane bandwidth, insanely priced exotic caps etc). it looks like a very "worked out" implementation where everything is there for a very specific and clearly defined reason (which seems to be linearity above all).
but it's DIY unfriendly as it gets.
the basic idea is obvious from the principle diagram.
what's notable is that they went to some lengths with the implementation in order to maximize linerarity. far from what one would expect and the absolute opposite of the typical internet wisdom, short signal path etc. and also very different from the "let's have tons of that, that and that" approach (where "that" typically stands for insane slew-rate, insane bandwidth, insanely priced exotic caps etc). it looks like a very "worked out" implementation where everything is there for a very specific and clearly defined reason (which seems to be linearity above all).
but it's DIY unfriendly as it gets.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.