[kiteki]

 
IV stages are a bit of a different case, here I would champion the use of a very fast current feedback opamp designed for video mostly, as the glitch energy of delta sigma type dacs in particular is at very high speed, but you cant just drop current FB chips in place of voltage FB.

 
I'll try some video op-amps soon, after XX hours of listening the LT1028, OPA827 and MUSES are the current favorites in my DAC I/V stage.  It's difficult to know which one is the most revealing, and if I was blind evaluating the MUSES I suspect I'd completely overlook it.
 
 

 
I think you are misunderstanding the term line out. line out simply means line level voltage output, just a reference level voltage so that engineers know they arent going to blow or clip the input of their amps or mixers as they know what to expect. it has nothing to do with whether there is a circuit in the output and an IV convertor is simply there to give that standard level from a current output. this could be a simple resistor, a discrete jfet IV, or transformers along with many other types.
 
the above quoted specs I wouldnt take anything from the above, the 3 dacs have just taken the ideal datasheet performance from the dac chip, nothing to do with its performance in the application.
 
the below is a table for the line level single ended standard. in reality dacs are all over the place, the sabres balanced out for instance is about 3vrms @ 3.9ma which is about double the +6db SE standard, so pretty hot voltage, but lower than normal current.
 
Some current out dacs will handle passive resistive IV well, depending on their voltage compliance and current, others will not handle it well at all, with the sabre the THD+N is not great with passive IV, but some still prefer the sound. I personally dont like passive IV all that much; its a bit musical for my taste, top end lacking sparkle, easy to listen to, but for me dynamics are compromised, at least with the dacs i've tried it with.
 

Use	Nominal level	Nominal level, VRMS	Peak Amplitude, VPK
ARD, Germany	+6 dBu	1.550 (approximate)	2.192 (approximate)
USA professional audio	+4 dBu	1.228 (approximate)	1.737 (approximate)
Consumer audio	−10 dBV	0.316	0.447

Line levels and their nominal voltage levels.

 
the voltage out dacs, like the one you mention and wolfstons, should be buffered, not connected directly, they should not be loaded with capacitive cables etc and usually have an internal opamp anyway. so if you just have a well designed transparent IV stage that you leave constant during the tests, then this is just as valid as a baseline.
 
I generally prefer current out dacs as I like maintaining control of the IV stage, but i'm quite intrigued by the AK4399, so have picked up a couple of chips and a suitable proto board

 
I don't think the PCM5102 has an internal op-amp like the ES9023, or it wouldn't be called voltage-out, or am I missing something?
 
While your repy was helpful, thanks, my wave of thought is on a purist perspective.  For example in my Teclast T51 the 'line-out' has OPA2604, I don't want that horrible chip there colouring the sound in my experiments, I need a 'pure' line-out, even if only in theory.  As you can see in my JVC FX500 modification I like removing filters from the signal, digital and acoustic.
 
In short I'm considering the Musiland Dragon, TeraDak 2.6 and Colorfly CK4 right now, I just don't have the technical experience to know if these are the right choices for my 'line-out' experiments and listening!
 
 
 

 
I think kiteki is playing with the output stage of his dac, arent you? but yeah everything else is covered above. i've not looked at the pelican PSU, I really dont see the need for +/-16v, but hey...
 
haha yeah batteries will do that, no current limiting

 
The toasted OPA627 was in my DAC.  The 9V batteries in 2x, 4x, 6x are connected to a DIY portable amplifier.

 

[qusp]

Quote:

I don't think the PCM5102 has an internal op-amp like the ES9023, or it wouldn't be called voltage-out, or am I missing something?

yep, youre missing something es9023 is a 'voltage out' chip too. you havent mentioned a single current out dac. still dont think you get it, these things are all made from tiny little fets and transistors among other features and to provide a somewhat stiff output with decent voltage compliance and to set the correct voltage level, there is a circuit which is loosely called an opamp.
 
voltage is higher noise than current transfer, it would be pretty odd for a manufacturer to use voltage feedback in a dac rather than current feedback and the easiest and best measuring way to convert current to voltage and set the level is an opamp. this isnt a modern initiative either, its been that way for years.

 

[kiteki]

The common perception is to use LOD to bypass the amplifier, and that the signal is sent directly from the DAC.
 
The TeraDak has "passive I/V" and the Lite-Dac AH uses AD847, one uses op-amps - the other doesn't - so no I still don't get it...

 

[Avro_Arrow]

Your chasing a ghost...
Any DAC, whether current out or voltage out, still requires a low pass filter to keep the high frequency "junk" from
making it through to the amp. You might think that the fewer components after the DAC the better, but only to a point.
 
I think the quote from Albert Einstein says it best..."Make things as simple as possible, but not simpler."
 
Even the PCM5102 (which puts the I/V stage on the chip) recommends a low pass filter on the output.
 
Quote:

The common perception is to use LOD to bypass the amplifier, and that the signal is sent directly from the DAC.
 
The TeraDak has "passive I/V" and the Lite-Dac AH uses AD847, one uses op-amps - the other doesn't - so no I still don't get it...

 

[AmarokCZ]

Quote:

Your chasing a ghost...
Any DAC, whether current out or voltage out, still requires a low pass filter to keep the high frequency "junk" from making it through to the amp....

No one DAC strictly requires low pass filter, but apparently for most designers (and especially for "analog people") it's more acceptable when sine wave looks really like sine wave and not like that "sine" on my avatar (which actually is a sine wave from DAC without low pass).
For example NOS DACs usally don't use low pass filters (or just some simple RC filter to round off the edges, but it still don't look like a sine).

 

[qusp]

Quote:

The common perception is to use LOD to bypass the amplifier, and that the signal is sent directly from the DAC.
 
The TeraDak has "passive I/V" and the Lite-Dac AH uses AD847, one uses op-amps - the other doesn't - so no I still don't get it...

 
the common perception is wrong...
 
the teradak is compromised IMO, it is designed for tweakers, with plenty of space so they can install huge film caps and boutique resistors where a proper output stage should be =) i'm of the opinion that while you can make something half decent if you basically replace everything in it, there is no water into wine. as TDA's go, this chip is entry level only and the PCB layout, particularly the digital and power supply section is ....not ideal.
 
you cannot build that type of thing point to point with patch wires and expect high performance, if you want a decent dac, build or buy a decent dac from the ground up. some designs are basically good designs that could use a bit of a helping hand, but the layout, parts choice and grounding of this simply cannot be undone. sure its a bit of fun and some people have done some interesting things with these dacs, but there is a point where the layout itself hampers efforts.
 
yes i'm opinionated, some will disagree with my view, but dacs, particularly these days of higher speeds are more akin to RF than standard analogue, so are quite sensitive to layout issues /rant
 
Quote:

No one DAC strictly requires low pass filter, but apparently for most designers (and especially for "analog people") it's more acceptable when sine wave looks really like sine wave and not like that "sine" on my avatar (which actually is a sine wave from DAC without low pass).
For example NOS DACs usally don't use low pass filters (or just some simple RC filter to round off the edges, but it still don't look like a sine).

dont tell me you are a nosser too? cant get away from you damn people lol NOS simply uses the ear to act as LPF, but given how wideband many amps are nowdays i'm not sure this is enough

 

[AmarokCZ]

Nope, I'm not NOSser. But I don't think that it's bad idea to use ear as LPF, because it works and it's certainly better than oversampling combined with LPF set to -3dB@25kHz to 30kHz (which sounds terrible). I prefer high oversampling and then some LPF with low delay.

 

[qusp]

Quote:

Nope, I'm not NOSser. But I don't think that it's bad idea to use ear as LPF, because it works and it's certainly better than oversampling combined with LPF set to -3dB@25kHz to 30kHz (which sounds terrible). I prefer high oversampling and then some LPF with low delay.

 


indeed I agree putting a filter right in the middle of the HF spacial information is dumb and all too common. your preference sounds about the same as me, I use software apodising OSF (8 x 8 x FS) which feeds a synchronously clocked ESS. that way a very high quality cap can be used pretty cheaply too.
 
I just fret about room modes bringing some of that info that should be rolled off down into audible FR, so I prefer an actual filter

 

[Headdie]

Hi,
 
I'm looking for advice or suggestions.
 
My new desktop headamp is the Firestone Cute Curve,
 
http://www.firestoneaudioproducts.com/cute-series/cute-curve/
 
The PSU is 24V and could be raised to 30V.
 
From the specs :
 
Main OPAmp : TI – OPA2604
Servo OPAmp : TI – TL072
 
Both are in sockets.
 
Do you think I could improve it a little with new OPAmp ?
 
I don't know anything about it. I've read enough in this thread and elsewhere to understand that the best I can do is to ask for a little help.
 
My phones range from Denon D2000-25ohms to Beyerdynamic DT990-600ohms. If it's relevant, I'd prefer to optimize it for mid/low impedance... I already have what it takes for high impedance cans.
 
I listen to jazz, classical, progrock and many thing else.
 
I don't listen loud with headphones, if it matters to op amp selection...
 
So what do you think guys : Am I a good candidate for new opamps ? And what would you suggest...
 
Thanks a lot,
 
Headdie

 

[Headdie]

I've begun by ordering LME49860. I guess you have to begin somewhere...
 
Now a basic question : Do I simply grab and pull out the opa2604 from the socket, and then push the new one in it ?

 

[kiteki]

I've begun by ordering LME49860. I guess you have to begin somewhere...
 
Now a basic question : Do I simply grab and pull out the opa2604 from the socket, and then push the new one in it ?

 
I'm not a fan of the OPA2604.  I've listened to it for over a year and tried it in my DAC, it has a little colour, it's like a neutral colour, it's deceiving you into thinking it's a nice natural sound, I saw someone call it "like red wine" once, anyway.
 
You need to use a tool to extract the op-amp, ideally.  It's called extractor something, looks like a polygon shape with a ~1cm wide tweezer on the end.
 
You should order more than just the LME49860.  What type of sound are you after?
 
Will comment on the other posts later.

 

[Headdie]

Bonjour Kiteki,
 
I'm after better sound : less noise, better fidelity and details. I'm not looking for a way color the sound.
 
I intend to roll once with the LME49860. If it's a clear change, then I may be tempted to continue. If not, I'll probably stop there.
 
May I roll it without the extracting tool ? I mean, is it locked in the socket somehow...
 
Thanks,

 

[zilch0md]

Quote:

Bonjour Kiteki,
 
I'm after better sound : less noise, better fidelity and details. I'm not looking for a way color the sound.
 
I intend to roll once with the LME49860. If it's a clear change, then I may be tempted to continue. If not, I'll probably stop there.
 
May I roll it without the extracting tool ? I mean, is it locked in the socket somehow...
 
Thanks,

 
With everything turned off, I just insert a narrow, flat-bladed screwdriver under the op-amp and gently twist it a little bit, going to the opposite side to do the same, then back to where I started, until I've lifted it away from the socket as high as the screwdriver blade is wide, then I just grab the package between thumb and forefinger and gently rock it the rest of the way out of the socket - being careful not to move it sideways until it is well clear of the socket - to avoid bending pins.
 
Joy!
 
Mike

 

[kiteki]

 
Did you buy a dual like this one?  http://www.ebay.com.au/itm/1PCS-Dual-Mono-LME49990MA-replace-OPA2604-NE5532-TL072-/320863640326
 
What Mike said with a screwdriver will work.
 
Hope you like the new sound, you won't more details, however you should find a cleaner character.

 

[Headdie]

LME49860 is already two channels...
I've considered an opa627 dual of that sort...