• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Any high end DAC operating in voltage mode?

maxwell3e10

Member
Joined
Dec 29, 2020
Messages
42
Likes
23
Most high end DACs (eg. ES9038Pro-based) operate in current mode. The outputs of the DAC are held at virtual ground and the current flowing out of them is converted to a voltage using an I/V converter (eg. using OPA1612). This gives a lower distortion (according to ESS datasheet). But it also increases the power dissipation in the DAC so that its temperature changes a bit for different output levels. One can see that if a high level output is suddenly turned on and the RMS amplitude is monitored with a high-resolution AC voltmeter it will drift over the first minute or so. This of course is not audible, but still...

ES9038 also allows voltage mode operation, which would use a voltage follower at the DAC output without drawing any current. So my question is if there are any DACs implementing this topology? Perhaps THD would be somewhat worse but the output stability would be better and total power dissipation in the DAC would be smaller.
 
Last edited:

Darkman

Member
Joined
Sep 6, 2022
Messages
13
Likes
2
If you want to have natural sound only I / V, I recommend the J-Fet here.
GBW is too low for op-amps. To be stable, they must have NFB which castrates the sound.
 

Darkman

Member
Joined
Sep 6, 2022
Messages
13
Likes
2
Hi
The problem will always be in a DAC with a voltage output.
A current OPA series from / including 627 is looking for problems.
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
2,690
Likes
6,012
Location
Berlin, Germany
Most high end DACs (eg. ES9038Pro-based) operate in current mode. The outputs of the DAC are held at virtual ground and the current flowing out of them is converted to a voltage using an I/V converter (eg. using OPA1612). This gives a lower distortion (according to ESS datasheet). But it also increases the power dissipation in the DAC so that its temperature changes a bit for different output levels. One can see that if a high level output is suddenly turned on and the RMS amplitude is monitored with a high-resolution AC voltmeter it will drift over the first minute or so. This of course is not audible, but still...

ES9038 also allows voltage mode operation, which would use a voltage follower at the DAC output without drawing any current. So my question is if there are any DACs implementing this topology? Perhaps THD would be somewhat worse but the output stability would be better and total power dissipation in the DAC would be smaller.
I haven't seen the output voltage (after I/V) drift in any current-mode DAC. If anything, the voltage reference(s) to the chip will drift and settle.
Can you share some real data and how you measured it?

Voltage mode is not recommended for the ESS DACs as the internal bias resistor is very nonlinear vs voltage.

Many AKM DACs (notably 4490 and 4493) have voltage outputs (and measure and sound as well or even better than ESS competitor like 90382qm). Since those voltage outputs have to be filtered as well, the output is not unloaded actually, so much for that.
 

Darkman

Member
Joined
Sep 6, 2022
Messages
13
Likes
2
I'm old ... take a fix for it :)
By design, DACs with voltage outputs are problematic.
I have not heard any to please me.

After many, many years of fun ... I still use the Philips TDA1545A.
With jfet I / V Out balanced.

16 bit caters for everyone, we are not dolphins or bats.

Of course, everyone has different needs, but let's think about what we are able to consume in our own room.
Without it, the "arms race, come on, higher, BETTER?" Will force a lot of money.
I know a lot of people who spent a lot of money and often after 10..30 years they said that they had a good sound EVERYWHERE.
Somewhere they made a mistake and they WAS PADDED.
 

boXem

Major Contributor
Audio Company
Joined
Jun 19, 2019
Messages
2,014
Likes
4,852
Location
Europe
Say what now??
torture.jpg
 
OP
M

maxwell3e10

Member
Joined
Dec 29, 2020
Messages
42
Likes
23
Many AKM DACs (notably 4490 and 4493) have voltage outputs (and measure and sound as well or even better than ESS competitor like 90382qm). Since those voltage outputs have to be filtered as well, the output is not unloaded actually, so much for that.
Thanks, I probably should look at Topping E30 II DAC or SMSL D-6, which use AK4493S. It's true that output filters typically use pretty low value resistors. Would be interesting if one added a simple follower and used high-stability resistors in the filters or maybe just use larger resistor values to reduce heat dissipation.
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
2,690
Likes
6,012
Location
Berlin, Germany
Would be interesting if one added a simple follower and used high-stability resistors in the filters or maybe just use larger resistor values to reduce heat dissipation.
Reducing dissipation (to have less thermal modulation -- of what exactly?) doesn't buy you anything in term of performance, IME. That is, if any relevant at all, given the overall dissipation. I actually tried to unload the AK4493's outputs for audio frequencies with an active circuit in order to keep the filter as is and it did not improve anything.
Any follower would need to handle RF noise well, and better be very low distortion and low noise. Large resistors only increase noise and susceptibility to interference.
 

Human Bass

Addicted to Fun and Learning
Joined
Nov 25, 2018
Messages
653
Likes
660
I'm old ... take a fix for it :)
By design, DACs with voltage outputs are problematic.
I have not heard any to please me.

After many, many years of fun ... I still use the Philips TDA1545A.
With jfet I / V Out balanced.

16 bit caters for everyone, we are not dolphins or bats.

Of course, everyone has different needs, but let's think about what we are able to consume in our own room.
Without it, the "arms race, come on, higher, BETTER?" Will force a lot of money.
I know a lot of people who spent a lot of money and often after 10..30 years they said that they had a good sound EVERYWHERE.
Somewhere they made a mistake and they WAS PADDED.
You mean 44.1khz right? The 16 bits deals with another matter, the khz that determines the highest frequency being reproduced.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,834
Likes
16,496
Location
Monument, CO
Voltage-mode has its own set of problems, such as nonlinear output device capacitance and resistance change with signal voltage, bias modulation, load-dependent bandwidth and distortion (unless you add a voltage-mode output buffer, natch), thermal modulation of the output stage with varying voltage (which requires also varying current so you don't really get around it either way), plus the output impedance comes into play), etc. etc. etc. Most of the digital-to-analog converters I designed during my career (not audio, higher frequencies) were current-mode as performance was higher. I think it not unreasonable to assume that if the audio DAC manufacturers could wring better performance out of them in voltage mode, they would.
 

Matias

Master Contributor
Forum Donor
Joined
Jan 1, 2019
Messages
5,029
Likes
10,794
Location
São Paulo, Brazil
Many AKM DACs (notably 4490 and 4493) have voltage outputs (and measure and sound as well or even better than ESS competitor like 90382qm).
RME disagrees as they managed to make them measure quite similarly.

 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
2,690
Likes
6,012
Location
Berlin, Germany
RME disagrees as they managed to make them measure quite similarly.

Nope, AKM shows none of the typical ESS modulation pattern around the -20dBFS levels (where it will cause sever ESS hump if not "masked out" by enough noise).
 
OP
M

maxwell3e10

Member
Joined
Dec 29, 2020
Messages
42
Likes
23
That would increase noise.
Yes, it would be a balancing act. The output noise of best DACs is about 10 nV/Hz^1/2. The Johnson noise of a 1 kOhm resistor is 4 nV/Hz^(1/2), for 10 kOhm it is already 13 nV/Hz^{1/2}. So most circuits use less than 1 kOhm resistors to squeeze every fraction of dB of noise. One can use slightly larger values.
 
Top Bottom