oh my, LT1211ACN8...I'm in love 
I'm always luckier with slow chips, God knows why #_#
definitely an improvement over LT1213, I'll try 1215 later.
instead of mocking us, if someone with technical background could explain what specs matter the most in datasheets..now that'd be a lot more interesting than trolling around IMHO.
apparently LT1364 has a huge DC offset and can kill headphones, it's really not clear from the datasheet..
I'm always luckier with slow chips, God knows why #_#
definitely an improvement over LT1213, I'll try 1215 later.
instead of mocking us, if someone with technical background could explain what specs matter the most in datasheets..now that'd be a lot more interesting than trolling around IMHO.
apparently LT1364 has a huge DC offset and can kill headphones, it's really not clear from the datasheet..
Nothing in the datasheet can tell you how the opamp will sound... If the slowest of the series (LT1211/1213/1215) sounds better, it's because of its type of distortion I guess... as slower = lower biased, which definitely will change the harmonic content of its distortion in the audio spectrum. What you can do is judge by ear, nothing else...
BTW, it's also the case with the LT1364 and the slower LT1358. Generally, for sound quality, I'd take the LT1358 which is the same thing (same circuit) but lower biased thus slower. Remember that digital audio has a max. slew rate of around .7 V/uS
The LT1364 is a bipolar-input opamp with 600 nA of "input bias current", there's no mystery as to why it produces DC offset if placed in circuits designed strictly for FET-input opamps. Instead in both positions in your soundcard it will be fine.
Last edited:
BTW that LT1211 looks nice, since (from the datasheet) it has distortion below .001% up to 20 KHz. The distortion rises a lot when gain is higher than unity, though. For higher gains it would be better to use its faster brothers...or something else.
It has a not so low voltage noise at 1 KHz, though... although it remains the same even at 10 Hz.
Were it me, I'd just use an LME49725 or LME49723
It has a not so low voltage noise at 1 KHz, though... although it remains the same even at 10 Hz.
Were it me, I'd just use an LME49725 or LME49723
Last edited:
instead of mocking us, if someone with technical background could explain what specs matter the most in datasheets..now that'd be a lot more interesting than trolling around It's not that easy.
It's as I keep saying, it's how that opamp interacts with the circuitry around it.
I too have a suspicion that the nature of the distortion produced (the spread of harmonics 2nd, 3rd 4th etc) has quite a bit to do with "sound" although at the levels we are talking about it's hard to believe... nevertheless, it seems to have some basis. I tend to go for devices where even harmonics predominate which if you have to have distortion then even is the most pleasing to the ear.
As for slew rates, bandwidth etc when you look at what CD is capable of (or more correctly not capable of) then you understand that is much less of an issue.
In one of the links I put on here there was mention of opamps used as I/V convertors in a DAC and that bipolar devices may "react" to the high frequency hash present in a way that FET's do not. That's the first time I have seen that argument mentioned and it sounds to have some basis of truth about it.
With all opamps you MUST correctly deploy them... and that means making sure the compensation etc is optimised... otherwise the tests are meaningless.
You could take 10 opamps and drop them in a simple test circuit and aply a 5 khz square wave and all would look different on the scope... you MUST compensate the devices. Even the stray circuit board capacitance has a huge effect on the overall result and must be allowed for in the final design.
As to offset voltages. You just don't understand the way opamps work to say that the LT1364 has a huge offset and will kill headphones...
Noise is probably the least important parameter for anything other than head amps/mic amps etc and even then the lowest noise devices aren't always what they seem in audio. Once again it comes down to optimising the design and something like the NE5532 or LM833 will give just about the lowest noise possible at the impedances we use for audio.
Andrea... you mention slew rates... have you seen what a 1khz squarewave from CD looks like... any player will show this the same. By the time you get to 6 khz or so it's triangular.
This is 1khz off disc
Attachments
OK, let me reformulate then: I find your experience with the LT1028 very odd. I wonder how many good opamps you've tried and in how many different places. That said, I do believe it's the bypassing of the output buffering stage of the Zhaolu what brought you the effect you described, becase the LT1028 per se doesn't have anything like poor bass, poor treble, poor dynamics... nothing such really, instead, those are the very reasons why it's my favorite single opamp, besides its midrange resolution & its tonal transparency.Here you are whining about how people are not taking your non-standard opamp judgement well, when you completely reject others' opinions when they're not what you like.
Clergy what? Did you notice that I mentioned 4 different opamps of 3 different brands in the post that started this thread? Have my other favorites been accepted any better, possibly...? I haven't noticed.
Ah...you're really shooting in the air with me, as I never expect anyone to make some use of my posting. I write mostly to clear things up for myself.
Ah, wwenze, at any rate... the LT1028 has better bass, better treble, better midrange, and better dynamics, than the OPA827, OPA211... and thus many others as well.
Do I appear as a freak? Maybe only the freaks have good ears? I hope not.
I do surely think that many people in forums overrate their discernment of good sound, given what kind of sound they end up favouring. Not talking of opamps only; also transducers (e.g. headphones), and other stuff
Do I appear as a freak? Maybe only the freaks have good ears? I hope not.
I do surely think that many people in forums overrate their discernment of good sound, given what kind of sound they end up favouring. Not talking of opamps only; also transducers (e.g. headphones), and other stuff
Last edited:
Hm, so you're saying that whenever you see DIP sockets, they are but a design mistake... I take note.
Hey, when the opamp is unity gain stable, and when it's configured so as to act as a low-pass filter in such a heavy way as happens after a DAC chip, I doubt you have reason to split hairs with regard to the compensation thing... I seriously doubt.
Agreed.
They told me that the LME49723 is two times noisier than the LME49720... what do I care, when I don't hear the noise anyway, while I do hear the sound as subjectively better?
Good to know... but this doesn't change the fact that a faster opamp (all the rest ideally being equal) tends to let itself be appreciated even with CDs.
For instance... the LT1363 has 1000 V/us and definitely gives the impression of bringing out some very small details better than the 20 V/us LME49710. So much for the microscopic level; on the macroscopic side, though, things do seem to sound a bit clearer with the LME49710...
Soo many things we fail to explain
Last edited:
Interesting... I guess this might explain why leeperry prefers the slower bipolar opamps just after the AK4396 (with its modest 75 dB of stopband rejection).
Last edited:
Apparently it has to do w/ "2µA Maximum Input Bias Current"..meaning that if not decoupled w/ caps it can easily output more than 20mV of DC voltage...which will KILL headphones.
the AK4396 uses two LPF, no I/V like the PCM1972...but well, it's still a DAC output anyhow...and on both those DAC's slower op-amps have always sounded better IME.
Last edited:
I'll let you work it.
The power dissipated in the voice coil at DC is just ohms law.
Volts squared/resistance
Even if the coil were 1 ohm what do you get ?From an audio point of view though, it's far from ideal as it will displace the diaphram from it's rest position.
77(8.8V*8.8V)/32Ω=2.42?
well some ppl measured VERY high DC offset on that chip, it has killed many headphones on small cmoys..
anyway, AD797BN has CRAZY low end bass, trumpets sound amazing, so do vocals..just like majkel said: Audio-gd discrete op-amps reviewed
feels like too much bass right now tbh..haha, and some background small details that were sounding holographic on the 1124/1211 sound dull now..but well, when there's no low end bass it's easy to make mids shine(the same trick the 4562/49720 use). Jazz is really mind blowing on this chip, and the Burson spans its SS into 270 degrees of goodness
I'll try it w/o the burson, see what it has to say on its own.
well some ppl measured VERY high DC offset on that chip, it has killed many headphones on small cmoys..
anyway, AD797BN has CRAZY low end bass, trumpets sound amazing, so do vocals..just like majkel said: Audio-gd discrete op-amps reviewed
feels like too much bass right now tbh..haha, and some background small details that were sounding holographic on the 1124/1211 sound dull now..but well, when there's no low end bass it's easy to make mids shine(the same trick the 4562/49720 use). Jazz is really mind blowing on this chip, and the Burson spans its SS into 270 degrees of goodness
I'll try it w/o the burson, see what it has to say on its own.
If this is the circuit,
HeadWize - Project: A Pocket Headphone Amplifier by Chu Moy
then the reason for high DC offset with some opamps is that the input impedances (inverting/noninverting) are wildly different. Fine for FET's not for Bipolar.
Even so 8 volts sounds extreme and I suspect that again this is down to poor implementation, perhaps the device was oscillating at HF which would give strange DC levels. This is a problem with widebandwidth devices that are not used correctly.
HeadWize - Project: A Pocket Headphone Amplifier by Chu Moy
then the reason for high DC offset with some opamps is that the input impedances (inverting/noninverting) are wildly different. Fine for FET's not for Bipolar.
Even so 8 volts sounds extreme and I suspect that again this is down to poor implementation, perhaps the device was oscillating at HF which would give strange DC levels. This is a problem with widebandwidth devices that are not used correctly.
Not precisely that, but evolutions of that elementary circuit that often (I don't know why?) maintain the near-exclusive compatibility with FET opamps.
oh ok, well my soundcard has many decoupling caps to kill DC offset anyway
I agree that AD797BN has a vynil sound(as opposed to CD sound)...for the good and the bad, low end bass is very percussive and detailed but small details in the trebles are not at the front. I'm not sure what I prefer
that's a bummer that there's no op-amp that does it all well, a chip w/ great low end bass, detailed mids, very clear trebles and wide SS...ah well, I'll keep digging. I don't think I like how male vocals/mids sound on this AD797BN...low end bass is not everything
I agree that AD797BN has a vynil sound(as opposed to CD sound)...for the good and the bad, low end bass is very percussive and detailed but small details in the trebles are not at the front. I'm not sure what I prefer
that's a bummer that there's no op-amp that does it all well, a chip w/ great low end bass, detailed mids, very clear trebles and wide SS...ah well, I'll keep digging. I don't think I like how male vocals/mids sound on this AD797BN...low end bass is not everything
Last edited: