The best sounding audio integrated opamps

[leeperry]

the technician did check for oscillation when he ditched the PCB socket and it was perfectly stable...but good point, I will ask him to check again after he puts the socket back on

indeed, I also tried to add 0.1uf caps on the V-/V+ of that soundcard, and it killed the bass percussion as well...ah well, I'll stop worrying and go see him tomorrow to get the damn socket back on..."don't fix it if it ain't broken" never sounded so true.

 

[Joachim Gerhard]

A good way to find out that OPamps oscilate is to measure DC offset right after the Pin 6
output. If no DC is at the input offset at the output should be input offset times DC gain.
If it is consideraby higher the OP gets into "preoscilation". If the oscilation gets worth the chip gets hotter and hotter until you can not touch it any more provided the PSU is strong enough. I found that method quite sensitive and it helped me several times to find the smalest posible Miller capacitor in complicated multistage feedback structures.

 

[leeperry]

the Spitfire is not ventilated at all, and even after several days in a row I've never had any cracklings/pops and it's never been warmer than my finger when touching it....I think it's a bad case of 1)very sensitive opamp 2)that browndog playing tricks on me 3)impedance or whatever other voodoo magic

it was giving a thrilling sound on the stock socket, and now it's boring to death...not my imagination.

 

[jean-paul]

the technician did check for oscillation when he ditched the PCB socket and it was perfectly stable...but good point, I will ask him to check again after he puts the socket back on

indeed, I also tried to add 0.1uf caps on the V-/V+ of that soundcard, and it killed the bass percussion as well...ah well, I'll stop worrying and go see him tomorrow to get the damn socket back on..."don't fix it if it ain't broken" never sounded so true.

Did you decouple from + and - to ground and did you use a cap between + and - ?

No socket/browndog and the chips directly on the PCB should be best technically and sonically. I think you might want to decouple better before putting the sockets back in !

 

[leeperry]

you mean adding a 0.1uf mkp cap between V-/V+? I'm a noob, and that's exactly what I did...I didn't decouple any ground, and it sounded dull....I did what many ppl do on that Prodigy HD2 soundcard: ImageBam

if you check this link, this seller says that the browndogs should be avoided like the plague: 10pcs Dual SOIC to DIP-8 Convert PCB Adapter - eBay

I've always found the "2*singles SOIC8> dual DIP8" browndogs crappy sounding..

 

[jean-paul]

The best socket is no socket if you forget about service-friendly faultfinding... Browndogs are to be avoided indeed unless there really is no other solution.

One cap from + to GND as close to the chip as possible, one form the - to GND as close to the chip as possible AND one directly across + and - as close to the chip as possible. The latter also with short lead wires preferably on or under the chips directly on the supply pins. Use good quality small film caps with short wires. Bend the wires exactly that they fit before soldering (and also cut them before soldering).

These are pretty basic high speed opamp techniques !

Please leave the chips where they are now are chances are likely that you will ruin the chips and/or PCB for no valid reason.

BTW check the datasheet of the chip in question for optimal type of caps and/or values. I think in this case 0.1 uf Wima MKS2 is a good choice for all 3 caps.

 

[leeperry]

ok, thanks for the tips! but there isn't enough room under the PCB in the Firestone Spitfire enclosure...and I was very much enjoying the sound before tbh

the technician has a hot hair gun, he's 60yo and highly skilled to this kind of changes...all I want is the same sound as before..

 

[leeperry]

lol hot air, not hair...EDIT killed me

 

[jean-paul]

There is not room for 5 mm pitch Wima MKS2 ? These are less wide than the length of the soldered pins of other components sticking out at the underside of the board...

So you are changing it again to a technical less appealing solution with a high risk for damage while the solution can be to decouple the opamps (as they should be because they're high speed opamps ) without even having tried to properly decouple them ?? Just to have the same sound you had in less optimal conditions ? Why change opamps then if the basic requirements for using high speed opamps are not met ? So one just does something without any preparation and expects things to change for the better just like that ?

We call that "womens logic" over here

BTW I see on the link you posted that the DAC has OSCON caps very near the opamps. I hope they haven't used those for decoupling analog circuitry. It could very well be the cause of the device not sounding optimal when it should with better opamps. OSCON in audio = bad sound.

If the power lines need "impedance" small value resistors could be used but I would only think of that after I would have done the basics first.

 

[leeperry]

hi jean-paul, yes I apologize for my miserable chick logic here

anyway, the guy put back the socket and I'm nodding my head silly again it took him 30 secs, the guy runs au audio repair shop and has golden fingers

even on his test speakers, the sound was much better w/ the socket...w/o the socket it was dull and not much low end bass or trebles, and w/ the socket the mad LT1028AC low end bass was back and the trebles as clear as usual!

w/ the socket the SS is also very holographic and narrower than w/o the socket...it made the center channel very "empty", all the sound was happening in the extreme sides of the SS and not much in between

last time I tried to add 0.1uf MKP caps between V+/V- on a soundcard(I posted the pic above), it was a big failure...it was also killing the low end percussion..

here's a clearer picture of my current DAC: ImageBam

that's the old Spitfire w/ a rollable opamp, a CS8414 S/PDIF receiver and components on both sides of the PCB....here's the new Spitfire w/ a soldered LM4562MA, solid caps(bad for audio?), DIR9001(better jitter than CS8414?) and components only on top: ImageBam

they both use PCM1793...don't you think that the new one is a downgrade? granted I really can't stand the LM4562 sound(very limited and boring SS to my ears), but I really didn't like the sound of the new Spitfire.

that's what their PSU looks like(I use the top gray 220V transfo version), it's feeding 24VDC(transformed into ±12V apparently): DIY supply for Firestone Spitfire? - Head-Fi

well, if you can find/make me a properly designed DAC(AK4396?) w/ LT1028ACN8 opamps LPF on a linear regulated DPS + toslink input, I'm all ears! that's my current setup, fed w/ glass toslink: ImageBam

after the technician put back the socket in my DAC, he compared it(fed w/coax) to the built-in analog output of his Arcam cd player...even w/ his +60yo ears he could easily hear a difference, far more resolution and much more percussive low end bass(LT1028AC is amazing for this).

he also checked for oscillation(w/ an analog oscilloscope) and measured some very very slight oscillation in the MHz range...nothing to worry about he told me.

 

[jean-paul]

Now we're speaking of oscillation in the mHz range. It could be that is the cause for you liking the sound as it is now !

BTW 30 secs is good but I suppose the PCB is (old school style) good quality. Most modern manufactured PCB material does not last long when soldering 2 or 3 times. Tracks peel off easily, that is why I warned you for possible damage. I reworked 3 "hong kong dac's" a few months ago and 1 sec with the soldering gun was enough to peel off the tracks.

AFAIK there is not much wrong with solid caps. It is OSCON that, despite its good specs on paper, manages to screw up sound big time.

Just like you I like the older Spitfire better as seen from the pics for modability ( if that's an english word ?!?). I never heard either of them.

 

[leeperry]

well yes it's entirely possible! But I always thought that oscillation would add some "zing" to the trebles...like it's often the case on improperly implemented LM4562's...that nasty metallic treble color I cannot stand

w/o the socket, the low end bass was dull and felt like not much was happening <50Hz and >14kHz..so I don't think that oscillation could improve the SQ so much..besides even after several days, the 1028AC are not warmer than my finger when I touch them.

Well, he did a great job, and the socket stays...as I've got some chips to try anyway, like AD845KN and ADA4627-1B

The new firestone models look cheapo to me, but the old models look very nice and sound fantastic to my ears(it had an OPA2604 stock, so it had to go )...it's also thanks to their DPS I'd guess, as we're mostly listening to the AC mains power as I understand it...one thing's for sure: I won't be using an SMPS anymore, it'd be a major nocebo for me anyway..

oh ok, I always read that solid caps were a terrible idea for audio...and I tried this soundcard that had a very harsh and agressive sound I couldn't bear: http://c1.neweggimages.com/NeweggImage/productimage/29-271-005-04.jpg

anyway, I'm still in the market for an AK4396/DIR9001/LT1028AC combo if anyone knows of one

R2R is supposedly the highest grade of DAC you can buy: Mother of Tone - Conversion Techniques

AK4396 is said to have the same SQ as R2R: DCX2496

Richard Kulavik of AKM Semiconductors explained it this way : "This DAC is a large departure from other delta-sigma DACs designed by us and others like BurrBrown, Analog Devices and Cirrus Logic. The AK4396 is an entirely new modulator, pioneered and patented by AKM. It achieves something unique. In the past, many of the old Phillips and BurrBrown parts were R-2R* based products. These older products were looked upon as some of the best. One of the reasons was high frequency noise. In older R-2R parts, HF noise was not present. In all delta-sigma parts prior to the AK4396, everyone has fought HF noise caused from the delta-sigma modulator with the insertion of large filters and other parts to attempt to solve a problem created by the delta-sigma design. The AK4396 today effectively does not suffer any modulator-induced HF noise and is over 60dB better than the nearest Cirrus and BB devices. All of this HF noise can cause many audible artifacts downstream. That is the 'miracle' we believe is making the difference today. This part gives you the performance and linearity of a delta-sigma device with the noise performance of an R-2R part, something that was never previously available."

but this guy says that opamps color the sound far more than DAC's ever will: Heres some test results for the new ESI Juli@ card. [3] - RightMark Forums

Everybody -in different locations, at different times, without knowing from each other- told the same story, that they found the differences between opamps more important than the differences in dac chips.

ah well, audio is such a complicated matter...and I love my DAC w/ its socket, its two LT1028AC and glass toslink input...that little upgraditis whining voice needs to shut it

 

[leeperry]

I've asked someone else, who tells me that sockets do add impedance...and I've tried to add a gold plated socket on top of the Spitfire's, and it sounds just like when I had sockets on the browndog...sound is less clear and "muddier".

so a wild guess would be that the impedance added by the stock socket was taken in account in their design...chick logic I know

the Spitfire is not ventilated at all, and after several hours the 1028's were slightly warmer than my finger...something like 40/45C? I think that's fine.

anyway, I will also try to solder some 1028AC's on this adapter: http://shop.diyrealaudio.com/product_info.php?cPath=42_91&products_id=851

this ebay seller says that the browndog's are worthless: http://cgi.ebay.com/10pcs-Dual-SOIC-to-DIP-8-Convert-PCB-Adapter-SMD-OPA627_W0QQitemZ230379640581

 

[leeperry]

BTW, I was wondering, is there a single version of NJM4580? this one is really good to drive headphones

 

[drizdraz]

Just wanted to chime in on the OPA1611 that some people are raving about on this thread. We tried them in a critical listening test on a front-end all-diff discrete/IC gain stage (40dB) and rated them dead last. They shifted timbre far more than the other 5 devices under test. Another poor performer was AD8671.

That said, the 1611 might be a stellar performer in a different topology. As always, it depends on myriad variables, designer's goals, power options, etc....

 

[zebra100]

I have just stumbled across this thread, so appologies if these have already been mentioned.

I like the LM4562 dual op amp. As a drop in replacement for my NE5532's (Phillips branded) in my NAD C520 CD player, they seemed to unearth a little more detail from the music.

I have the NE5532's tucked away for a future project, as I still think they are a good op amp when implemented correctly.

Chris.

 

[danvee]

Hi Chris!

If you like the LM4562s, you really need to try the OPA2141. You'll need an SOIC to DIP adapter and a bit of soldering skill, but well worth it!
Another goodie to try is the LM49860. That one comes as a regular 8 pin DIP and will be a straight drop in.
The 5532s are still an industry standard, and, as you said, work great in the proper application. They're dirt-cheap too.

Take Care,

Dan

 

[Ola]

LM series 61XX are worth trying.

 

[danvee]

Interesting....Which ones have you tried, and in what application?

Dan

 

[Melon Head]

Well, let's see, you've got two 4s in there and that's an 8. The word "eight" starts with "e," which is the fifth letter of the alphabet. 5 x 8 = 40, which is a 4, so we're back to too many 4s. This is circular. The symbol for a tube is based on a circle, tube sockets are circular. So therefore, you need to use a tube to replace the AD845.

Now, what kind of tube? A naive person, not steeped in the arcana of true numerology, would immediately think "845." But that's bad thinking, because it brings five 4s into the picture, the 4 in 845, the two 4s in the "8," and because 845 is numerologically an 8, there's two more 4s. This is NOT good.

So clearly, you need a CCa. C is 3, a is 1, so CCa is 7. This is neither a 4 nor divisible by a 4.

Don't listen to self-appointed hacks who don't really understand numerology properly.

This has to be one of the most enetertaining posts I have come across so far.
LMAO

Conclusion: You need a tube