[carlmart]

Forget about the slew rate, this is snake oil: https://www.radio-electronics.com/info/circuits/opamp_basics/operational-amplifier-slew-rate.php. I don’t know what do you understand by “line level”, but for quitar that’s 0.8-0.9V RMS and for a regular DAC that would be 2-2.2V RMS. That means any audio opamp can handle line-level voltages when used in output buffer.

A very serious and respected audio engineer, Walt Jung, considers slew rate a very important thing on an opamp, so I don't need to read anything that might claim that is snake oil.

It's related to the capacity for the DAC filtering being done as fast as possible, and the analog opamp being able to be fast. On his first mods in TAA (you should find them and read them) on Philips CD players he used a fast IC and a video buffer to modify the original DAC.

And yes, line level is about that voltage you mention, and yes, most opamps have the capacity to output that. Other things, like distortion, slew-rate, current capacity, etc. contribute to the chip sounding in a particular way, that some consider "better".

Those opinions, based on listening experience, is that I would like to have. Contrary to many audio engineers, certainly not Walt Jung, I believe in subjective opinions to evaluate audio, along with measurements.

 

[raoultrifan]

What I was trying to say is that for an opamp used as output buffer slew rate is definitely not important, feel free to read the above calculator. Mackie, JBL a.s.o. are building audio monitor speakers around NJM/JRC opamps with a slew rate of 3V/uS to 5V/uS and musicians and audiophiles appreciate these speakers as having a very good audio quality. So, in this case more slew rate is indeed snake oil and not even AP equipment will "see" any differences between a 2V/uS and a 20V/uS opamp if used within the audio bandwidth only.

A fast opamp could help when used in I/V stage, sometimes in gain stage (RIAA or mic or is higher gains are used), but not as unity gain buffer.

1. OPA1688
2. OPA2608 (good for gain stage, not so good for output buffer)

For all your 3 questions I suggest you to use the diyaudio forum instead. You'll definitely get more luck in there.

P.S.: A good output buffer might also be http://www.ti.com/product/LME49724. It has great specs and huge output current for such a small-case opamp (SOIC8 with power pad).

 

[dhruvmeena96]

What I was trying to say is that for an opamp used as output buffer slew rate is definitely not important, feel free to read the above calculator. Mackie, JBL a.s.o. are building audio monitor speakers around NJM/JRC opamps with a slew rate of 3V/uS to 5V/uS and musicians and audiophiles appreciate these speakers as having a very good audio quality. So, in this case more slew rate is indeed snake oil and not even AP equipment will "see" any differences between a 2V/uS and a 20V/uS opamp if used within the audio bandwidth only.

A fast opamp could help when used in I/V stage, sometimes in gain stage (RIAA or mic or is higher gains are used), but not as unity gain buffer.

1. OPA1688
2. OPA2608 (good for gain stage, not so good for output buffer)

For all your 3 questions I suggest you to use the diyaudio forum instead. You'll definitely get more luck in there.

P.S.: A good output buffer might also be http://www.ti.com/product/LME49724. It has great specs and huge output current for such a small-case opamp (SOIC8 with power pad).

Ohh lol...

And I interpreted it in different way...lol

 

[raoultrifan]

A very serious and respected audio engineer, Walt Jung, considers slew rate a very important thing on an opamp, so I don't need to read anything that might claim that is snake oil.

I'm sure most of us know about Walt Jung and about his work, he was like father of audio engineering. I remember I've read about some mods he did on a Philips CD player, but it was related to the DAC chip from inside, something related to the digital FIR filters, then he also biased an opamp to class-A, then he bypassed some opamps with additional capacitors as well, but don't think he used video opamps as output buffers...really don't see why to do that for.

Contrary to many audio engineers, certainly not Walt Jung, I believe in subjective opinions to evaluate audio, along with measurements.

Me too, but measurements first and if these fail then I care not about how it sounds. I've heard fast opamps sounding so great...with an interesting soundstage and vocals were so rounded and pleasant to my ears, but when I connected my scope I've only seen ringing and oscillations.

Feel free to read about the slew rate here: https://www.eetimes.com/document.asp?doc_id=1278805&page_number=2, actually all the pages from this article explain very well many well-know opamp specs.

 

[carlmart]

For that project he used the LH0002 as output buffer, and a very fast IC for the I/V. Jung also suggested a fast discrete buffer instead of the LH0002.

Later on he also modified a Philips DAC, but I do not remember the chips he used.

OTOS I don't think you need high slew rate on a power amp, so Mackie's or JBL's audio monitors serve very much as an example.

And Jung did suggest that video buffers were recommended for DACs. Other projects on that mag used them.

Also fast opamps are prone to oscillation if you don't implement it alright, particularly if they are current feedback types. So I'm not sure your ringing or oscillation observations should rule out fast buffers for an output DAC, which Jung did recommend.

 

[Rroff]

While generally you don't need massive slew rates to hit the threshold for reproducing audio one thing to watch out for is that many opamps either state a "typical" rate rather than a minimum and/or use slightly different scenarios for where/how the slew rate was measured. I'm pretty certain some opamps that on paper have a "typical" slew rate that is "good enough" for audio aren't consistently managing it especially in DIY installations and aiming for some reasonable overhead is a good idea if you don't have the experience and/or equipment to test for it.

From an earlier conversation I think some of the subjective experience that tends to show slew rates as mattering often boiled down to a mixture of signal step response and/or cable length causing enough capacitive load to reduce the opamp's performance causing the actual issue people were seeing rather than straight up being the slew rate itself.

 

[raoultrifan]

I thought we're speaking about output buffers of an audio equipment (line level), not DAC voltage-out or current-out buffers.
However, for DAC buffers I totally trust manufacturer's specs. For example, LM4562/LME49720 is quite OK for use in both I/V and LPF. I'm not saying it's the best, but it's a very good opamp.

L.E.: LH0002 was replaced by BUF634 and LME49600/49610 250mA output buffers. These are output buffers for driving 16-600 Ohms headphones, not for line level outputs (>10 KOhms impedance).

 

[selvakumar]

Actually MUSES8920 are quite a good match in IV and LPF on my PLAY, though I like more 3xAD8599 (IV&LPF) + 2xAD797 (gain).

im gona order it and how to put
Analog Devices Inc. AD8599 only SOIC-8 package available

 

[raoultrifan]

SOIC-8 are mines as well, but I was able to DIY adapters for them. You can probably find already soldered on SOIC-DIP8 adapters.
You can also find AD797 in DIP8 package; these are extremely low-noise opamps and very recommended in gain stages.

 

[CoiL]

Any ideas how I can feed Burson V5i-D opamp properly in my Aune T1 ? Build separate power feed circuit from power brick entrance? Or modify existing amp section board? I really wish to get it working but it won`t power up properly instead NE5532A chip.

 

[raoultrifan]

After taking a look to http://www.pt80.net/thread-451874-1-1.html and used Google translate I see that OPA2134 is used for DAC output, hence it could get replaced by LME49720 or MUSES8920, and NE5532 could get replaced by LME49720 or similar SO8/SOIC8 opamps.

If you do have enough space between the PCB and the outer case, then you should be able to solder the V5i directly on PCB to replace existing NE5532, but lot of care is needed and perhaps some 5mm long wires.

Given the complexity of this task I would honestly not recommend you to install V5i in Aune T1, I would vote instead with LME49720, AD8599 or MUSES8820/8920 with confidence.

 

[BoomBap08]

Has anybody already had the chance to compare ad8397 against muses8920? And which one's better? Thanks!

 

[dhruvmeena96]

Tell me the application

Output buffer with ad8397 is very detailed and clean. Muses is more musical with some good low end texture.

 

[yurt28]

I do like their high current and low impedance drive capabilities, hence these opamps could be a great output buffer. Do note the lower THD for THS4931 vs. THS4032: -96dB vs. -90dB.

That is True by 6DB, but the big difference is 1 channel Vs 2

 

[yurt28]

Has anybody already had the chance to compare ad8397 against muses8920? And which one's better? Thanks!

Ad8397 is fun power wise with flat details, but against Muses8920 loudness can increase up to 139DB and is balanced. I haven't heard any hissing when listening at full blast.