DAC AD1862: Almost THT, I2S input, NOS, R-2R

[Joseph K]

Before going crazy with decoupling.. precisely this why I had been suggesting the small smd on top. Can help, without further complications..

Ciao, George

 

[miro1360]

Also need to clarify, how did you use lt3015 with WM8740? it does not need a negative supply.

It was on my LT1963 + LT3015 dual powersupply, and opamps needs this negative supply and LT1963 was for DAC. I write it all out of my head so I probably have it all mixed up, but this is how I memorized the dual supply.

 

[miro1360]

Vunce, I had a look at your schematic.
Place a 'base stopper' resistor between the two sections if OPA860, right on the input of the output buffer. After the filter cap! Like 1kohm..

Try the 1k, it can help ...

Also, the input stage base would need some hundred ohm (called Rs in the data sheet). (if You look carefully, the datasheet prescribes these 'damping' resistors at both places (Rs - 20ohm-250ohm), at the OTA input stage base; at the input of the buffer stage.

Ciao, George

This resistor should be small, let the I/V to see as smallest impedance as possible ... the simulation is working good with these changes (the R7 is only for simulation)

 

[Joseph K]

Miro,

If we want splitting hairs, yes the input Rs contributes to the impedance into the emitter.. but as You know it gets divided by Beta of the diamond transistor..
An order of hundred ohm will result in less than 1ohm extra..

We have more losses otherwise. And instead the damping effect in the base results less inductive response in the emitter branch. (Which could then resonate with input stray capacitances)

Ciao, George

 

[miro1360]

Right George, it significantly adds stability which is more important and the impedance remains small.

 

[Vunce]

A picture is worth 1000 words, thanks Miro
In order to do this modification it’s a little more tricky without cutting a trace on the board. The Opa860 pin3 (Base) is directly connected to GND.
I will figure a way though

 

[miro1360]

bend up pin 3 on that OPA860

 

[diyiggy]

It turns out to be funny and cool thread

 

[Vunce]

Yes it is Diyiggy!!
Just a bunch of peeps with a common goal....great sound!!!

 

[miro1360]

The motivation comes from fun

 

[lasercut]

Has anyone compared OPA1612 or another common OPA to more exotic active IV choices mentioned here?
OPA1612 ranks best out of a few other common op amps ( LM4562, NE5532, OPA2134) but still has slightly unnatural and compressed sound compared to passive, on the other hand passive is too lean sounding and this has become difficult to ignore.
Considering some of the high speed AD OPAs but they are expensive...

 

[diyiggy]

Sen/Zen...

 

[Vunce]

Hi laserscrape,
I’ve been using the opa1611, single version of 1612, it’s not bad. I like the LM6171AI better though.
Your description of the opa1612 is similar to what I hear also. My biggest gripe is the soundstage seems to shrink.

 

[lasercut]

Thanks, Vunce.
I will give the Lm6171 a try.

Sen/Zen is not really possible with those JFETs, theres the BJT version but even still the 2 floating supplies are a big investment.

 

[Vunce]

Hi Diyiggy,
Do you have Rogic’s opa861 I/V circuit on the output of a TDA1541 or AD1862?

 

[diyiggy]

TDA1541A AYA2 with opa861and a discrete diamonds with another dac with AD 1862

 

[diyiggy]

Has anyone compared OPA1612 or another common OPA to more exotic active IV choices mentioned here?
OPA1612 ranks best out of a few other common op amps ( LM4562, NE5532, OPA2134) but still has slightly unnatural and compressed sound compared to passive, on the other hand passive is too lean sounding and this has become difficult to ignore.
Considering some of the high speed AD OPAs but they are expensive...

OPA1656: High-Performance CMOS Audio Op Amp


I really think opa861 is an non exotic choice and if not one of the best and known from 15 years for that I/V task. Buffer is an other story and one should always precise what either task.
For driving a pre/amp, opa861 is certainly good enough diret after its I/V job and don't need buffer. However why not try a simple good enough discrete as buffer ? Matched LSJ74/LSK170 and voilou, walla, valhalla - pronounce like you 're able to-

 

[Joseph K]

Hehe. 15 years.. Maybe You wanted to say >>20 years..
The time when Thorsten started to show around his solution with OPA660.. Definitely from 2000 I still have a board with OPA660.. )

 

[Vunce]

TDA1541A AYA2 with opa861and a discrete diamonds with another dac with AD 1862

Could you share the DD I/V circuit your using with the AD1862 dac?

The OPA828 looks good to try also.

 

[diyiggy]

Hehe. 15 years.. Maybe You wanted to say >>20 years..
The time when Thorsten started to show around his solution with OPA660.. Definitely from 2000 I still have a board with OPA660.. )

You know my memory, lol ! Was is not Pedja Rogic that opened the hostilities with the opa861 after he woke up the ad944 ??