Why, oh why??I'm not going back to dual chips
You've got one dual opamp per each channel in your card, that is, no L/R crosstalk is involved. You won't lose anything.
No, it doesn't sound any better at all. I can compare them with the push of a button. And the DAC with the CS8 (powered with less than half the total V) sounds slightly better
well, the SS is just wider w/ single chips...there's more stuff happening in the center, it's been dead obvious each time I've rolled dual <> single
it's also been dead obvious that all the discrete chips as final buffer give a slight stereo imbalance on my card..maybe due to the low voltage? they are said to require at least ±12V and I feed them a bare ±8.8V.
anyway! the 1028CS8 final buffer increases the bass response but distorts the trebles..don't really like it, I will take them off in a minute. I might just need 6*AD797BN, or maybe an OPA1611/LT1677 final buffer
it's also been dead obvious that all the discrete chips as final buffer give a slight stereo imbalance on my card..maybe due to the low voltage? they are said to require at least ±12V and I feed them a bare ±8.8V.
anyway! the 1028CS8 final buffer increases the bass response but distorts the trebles..don't really like it, I will take them off in a minute. I might just need 6*AD797BN, or maybe an OPA1611/LT1677 final buffer
Last edited:
well, it seems that many ppl also agree that 2*singles are better...even on the OPA1612 they emphasise that the separation is better handled. Clearly, it gives more width to the center channel on headphones(better phase control?)
indeed the 1028 doesn't appear to be a drop-in replacement...but Slash47 said the same for the 797B, he told me it's oscillate like mad on my soundcard...it wasn't the case, there's no hard rule in op-amp rolling apparently
indeed the 1028 doesn't appear to be a drop-in replacement...but Slash47 said the same for the 797B, he told me it's oscillate like mad on my soundcard...it wasn't the case, there's no hard rule in op-amp rolling apparently
You keep not understanding... Those two dual opamps in between DAC chip and output buffer [which is supposed to be a dual opamp, funnily enough] are each of them only in one audio channel's signal path. Two channels of a dual opamp are in the left channel's signal path; and two channels of the other dual opamp are in the right channel's signal path.
Therefore, for that stage, you have no reason to worry about stereo crosstalk, phase problems, or anything like that... so a dual version of a given opamp (e.g. OPA1612) will be in no way a compromise as opposed to two singles on Browndog.
Sure, if you want to complicate your life just because you refuse any dual opamp by an uninformed prejudice, you're free to do so...
Whereas, regarding the dual opamp that buffers your card's line out... well, there it makes sense to use an adapter with two singles, because the two channels of the dual chip would be used to buffer the L and R stereo channels respectively.
Therefore, for that stage, you have no reason to worry about stereo crosstalk, phase problems, or anything like that... so a dual version of a given opamp (e.g. OPA1612) will be in no way a compromise as opposed to two singles on Browndog.
Sure, if you want to complicate your life just because you refuse any dual opamp by an uninformed prejudice, you're free to do so...
Whereas, regarding the dual opamp that buffers your card's line out... well, there it makes sense to use an adapter with two singles, because the two channels of the dual chip would be used to buffer the L and R stereo channels respectively.
Last edited:
I am well aware of what you're saying...yet, I get much wider center channel w/ singles as DAC LPF...there's obviously much to be gained by processing each polarity on its own.You keep not understanding...[..]
Whereas, regarding the dual opamp that buffers your card's line out... well, there it makes sense to use an adapter with two singles, because the two channels of the dual chip would be used to buffer the L and R stereo channels respectively.
browndog says : Single-to-dual Op-AMP Adapter - DIP version (p/n 021001)
Precision Amplifier - Wide Bandwidth - OPA1612 - TI.com
dual chips are more than likely made of compromises....an IC op-amp is already damn small, so let's not ask for even smaller.
when reading the old AD797BN datasheet, they clearly explain that there's all kind of protections and so on built in...maybe there's a very good reason where there's no AD799
used as what? final buffer? I/V? LPF? at what voltage? which ones did you try already?
Last edited:
I didn't see it like that... maybe you're right... there might be some unmeasurable degree of interference in processing the + and - halves of audio signals with two channels of the same opamp.
The OPA1612 "features completely independent circuitry for each channel" though. Sorry if I sound like I'm trying to sell something.
Last edited:
LT1028 (if working well in the given application). Secondly the OPA1611, and the AD797. Try the LT1028... you'll get exactly what you want
The AD797 has enough output currentoh yes! so much better w/o a final buffer, pure untouched AD797B goodness
Actually it's such a waste of simplicity to have separated filtering & buffering in your soundcard. I guess the 3 opamp thingy is a nice selling point in the eyes of the uninformed buyer of a soundcard. Funnily enough, they didn't even took full advantage of this unnecessary multiplication of chips (since the DAC chip is V-out), as they kept using a dual opamp for the final stage.
another company used the same DAC but put 4 op-amps: Why Claro halo needs 4 opamps for front output?
but the TI chip they used for the HP out is terribly harsh(I'll take my NJM4580V any day over this POS), and the sound was very agressive(I suspect the DSP chip, the same one as on the Asus STX and the SQ is pretty much identical...very agressive and harsh upper spectrum).
anyway, I'll try 6*AD797BN, 4*OPA1611/4*LT1677 + LT1028ACN8 as final buffer...and all these possible combinations, maybe OPA827 would work as final buffer. I love that rolling game
some songs were simply better on the burson, but it's a lot less versatile than the AD797BN..it only sings one song as majkel would say, AD797BN doesn't sing some of them as well..but its SS is constantly changing and not limited in any way(Slash47 agreed on that).
but the TI chip they used for the HP out is terribly harsh(I'll take my NJM4580V any day over this POS), and the sound was very agressive(I suspect the DSP chip, the same one as on the Asus STX and the SQ is pretty much identical...very agressive and harsh upper spectrum).
anyway, I'll try 6*AD797BN, 4*OPA1611/4*LT1677 + LT1028ACN8 as final buffer...and all these possible combinations, maybe OPA827 would work as final buffer. I love that rolling game
some songs were simply better on the burson, but it's a lot less versatile than the AD797BN..it only sings one song as majkel would say, AD797BN doesn't sing some of them as well..but its SS is constantly changing and not limited in any way(Slash47 agreed on that).
Last edited:
I'll also try to source some AD8597 for the hell of it...most ppl who love the 797 say that the 8597/99 are mind blowing too, that could make for a nice final buffer
damn, that AD797B is so colorless that it's hard to get used to..and that god**** SS, totally head spinning!
The AD797 is anything but colorless. If it were just a little like you described it, it would be an LME49710.
The AD797 which you're worshipping lately, is an old and very well-known opamp, and as I found is not free from +'s and -'s like everything else... including my favorite LT1028ACN8/CS8.
BTW, personally I'd expect the ADA4627-1 and possibly also the AD8597 to sound better.
Last edited:
I did mean 4*OPA1611
You might as well try! Also 6x OPA1611, for that matter
The OPA1611 would make a great ouput buffer. The LT1028 is one fine opamp for buffering too, but especially for intermediate signal conditioning stages (if stable in low gains in the particular circuit).
leeperry, why don't you try and get yourself a HA INFO NG98, with CS4398 or AD1852 (or both, for more rolling pleasure than ever) at your discretion?
It's got everything to sound great... including a great integrated headphone amp. It features the TI DIR9001 digital receiver, selected parts, very nice engineering... looks like a bargain
I'm sure that soon enough (there's no hurry...not at all!) I'll manage to get either a NG94 or a NG98 since it's christmas (= money) time 
Both would be great fun to play with... with no need for modifications at all (particularly the NG94, but the NG98 looks fine too with its Nichicon Muse ES capacitors in the signal path etc.).
I'm already beginning to think of what to use in each of them, among the LME49725, LME49720HA, OPA1611, LT1028ACN8, OPA211ID, OPA827, AD797ANZ, OPA627BP, AD845KN, etc. that I've got ready-to-use. Of course this mainly means that in my mind I'm playing with numbers & combinations of them...so much fun
It's got everything to sound great... including a great integrated headphone amp. It features the TI DIR9001 digital receiver, selected parts, very nice engineering... looks like a bargain
I'm sure that soon enough (there's no hurry...not at all!) I'll manage to get either a NG94 or a NG98 since it's christmas (= money
) time Both would be great fun to play with... with no need for modifications at all (particularly the NG94, but the NG98 looks fine too with its Nichicon Muse ES capacitors in the signal path etc.).
I'm already beginning to think of what to use in each of them, among the LME49725, LME49720HA, OPA1611, LT1028ACN8, OPA211ID, OPA827, AD797ANZ, OPA627BP, AD845KN, etc. that I've got ready-to-use. Of course this mainly means that in my mind I'm playing with numbers & combinations of them...so much fun
Last edited: