Hi Claude, Thanks for the input.There aren't here many people not liking "quite a lot" OPA1656, and that in various configurations suited to his specs (say mainly what people consider standard FET equivalent use, be it pre amp, amp, RIAA amp... quite versatile and robust little OPA as long as considering implementation as one should).
This application suits OPA1656 idealy. We had though already one person, out of a few hundreds, reporting not being happy. That means 2 things in my limited eyes and I won't go further as I do consulting for a living and that would mean charging you if starting a debate:
1- Something is wrong with your OPA1656 set up. SMD soldering, socket, PS... whatever...
2- You like it biased and non neutral re sonic performance, funky or loudness effet hello, or whatever your taste... and I respect that. Admittely there are more 'typed' OPAs around that can give you the sound YOU like and the beauty of the socket is you can try many provided they are compatible for this location.
As said, I won't expand further, did that already and opened a can of worms with someone that had limited tech knowledge. Further, at the end, my comments don't matter : this is Audio DIY, your are tweaking, important is YOU are happy and comments from others, including mine, are just what they are.
Have fun
Claude
PS: hint ...have you ever soldered SMDs before and did you consider a quick reflow of the OPA's pins?
I AM OUT, REALLY LOL!
I have been soldering smd components since last couple of years when I started playing with class d amps. However this is the first time I have soldered a 8 leg chip. Also I cross checked every solder joint just to make sure they aren't cold joints. The sockets and pins are pretty snug.
As for my like-ness for non neutral sonic, I am not sure. I feel NE5532 is much more neutral (i also see almost every site mention them as being neutral though a bit clinical) & the OPA 1656 which are certainly much evolved chip on paper lacks that level of neutrality. It sounds more warm, but that's fine as long as I am not loosing on details and soundstage. I can assure if you listen to my A07 with OPA 1656 , you will feel the loss of detail and that the soundstage became very narrow as compared to when it plays with NE 5532.
Also somebody said in OPA 1656 thread that what I am hearing as more details in Ne 5532 is the more noise !!! And that bcoz 1656 cuts those noises I am feeling it's less detailed! I have an argument against calling the NE 5532 less neutral and noisy though. How come the A07 with NE 5532 sounds closest to my other amps without any opamp ???? I guess best opamp is no opamp. And if NE 5532 sounds as close as those amps without an OPA, that means something. I am not bashing 1656 as that has been widely proved as a really good opamp but I really don't feel the Opa 1656 is a good choice in A07, might be the accompanying components around it doesn't synergize well with it. The difference is huge in a bad way. But I will still give it some time to mature.
Quite possible. Though almost all of the components seems laid out in a similar fashion and values (presuming) . But it's quite possible that OPA won't play equally well for both A04 & A07.
Ahhh, that's not an easy one to answer, isn't it?
Firt of all I congrat you for your openess, sharing your findings in always a positive way, not taking whatever bad mood regarding answers that are only meant to help. Not everyone is that neutral, congrats again.
Second, as music lover, I feel very sorry that you are not happy with your current sound, and we can't leave you like this, cant' we?
The sound you like may not me the one I like, who knows, the absolute sound, well, I am a 6S black belt coming from F1 and I learnt that absolute doesn't mean a thing in physics on this Earth LOL!
So...
I have been soldering smd components since last couple of years when I started playing with class d amps. However this is the first time I have soldered a 8 leg chip. Also I cross checked every solder joint just to make sure they aren't cold joints. The sockets and pins are pretty snug.
=> OK, sounds good. Just in case of, and that isn't a lot of work and doesn't mean wasting time, may I humbly suggest, based on the pix you posted, that you simply pull out the OPA1656 together WITH their sockets, put a lot of flux on the OPA1656 legs and simply put the soldering iron across the pins / legs legs of said OPA 1656. Just across all of them with a say a 350°C iron, touching quickly. Flux means the pins only will reflow, job done. And then clean the flux. May not change anything, but at least I will feel more confident LOL.
OK, there is reason for that : quick chips like OPA 1656... and much faster ones I dealt with can be quite sensitive to micro failures we may not see...
You can do that, or not, I realise cleaning the flux is a pain, but then you have flux I suppose already on hand and you paid for the OPA 1656s, so if these can be "saved" cheaply".... win-win
As for my like-ness for non neutral sonic, I am not sure.
=> Important is indeed what you like.
NE 5532 isn't bad, but in my ears re details, response shapness, 3D image, quite in another league than amps as AD825AD 20 years ago... which is also slightly lacking IMHO vs OPA1656 everywhere, a bit, bare in the (exagerated) bass department. So at the end NE5532 vs OPA1656 in the right application (04 is such one) and with the correct OPA PS (admittely modern OPAs are more sensitive to a correct PS) should be a no brainer. Where I come from a nice Tyrell Ford from the early 70s vs a Williams of the mid 90s LOL.
Not saying neither cans be beaten nowadays, but either way it ius true 2 balanced F1 cars somehow, not lacking anything if not compared directly ;-)
I feel NE5532 is much more neutral (i also see almost every site mention them as being neutral though a bit clinical) & the OPA 1656 which are certainly much evolved chip on paper lacks that level of neutrality.
=> I found OPA1656 rather more neutral in that application and others. Sometimes you could say some bass distorsion is missing. Ok, it has some very low second harmonic that brains usualy like (need?).
But NE5532 has so many limitations... something must be wrong somewhere.
Saying that, if not comparing directly, on a balanced amp, one could be happy with NE5532, it isn't that bad that something obvious pops as a problem.
But your case is different, you can compare, and assuming the right power feed (again, OPA 1656 needs a good PS, NE 5532 probably can do with lesser ones), you should hear a difference.
Oh, and the you are! Either a right one, which means your 07 needs some compensation, or a non conventional one.
Let's reflow the board / chip, and then it leaves you with what you hear. If it has changed a lot and you like it, bingo, welcome on board. If not, then we can assume you really prefer NE5532 to 0PA1656 in your set up.
If that's the case, either you are happy that way, or you wish more flavour, in which case let us know what you would like. As things are, I have always my JFET flavour box with me, various OPAs of the best quality, all measuring the same à la ASR... and all sounding so different. Gilles knows that box, called flavour box as all OPAs are quite interchangeable (similar parameters so can be used for same application). As posted on the OPA1656 thread, we played with it, and ended up in all cases with OPA1656. Use the search function to find the other OPAs and "our " findings, just our ears in a given situation.
It sounds more warm, but that's fine as long as I am not loosing on details and soundstage. I can assure if you listen to my A07 with OPA 1656 , you will feel the loss of detail and that the soundstage became very narrow as compared to when it plays with NE 5532.
=> Please reflow, just in case, just to make sure it is your taste and not whatever else...
Also somebody said in OPA 1656 thread that what I am hearing as more details in Ne 5532 is the more noise !!!
=> Can be, indeed, happens. Is often very noticeable in the bass. If listening with small speakers and old DACs I must confess AD825 biased in class A nicely blends out (= lack of definition!!) the DACs limits in terms of resolution while adding some very (=too) solid deep bass which complements non 25Hz FC LS LOL
The world is wide...
How come the A07 with NE 5532 sounds closest to my other amps without any opamp ????
=> May I humbly ask what these systems are and what amps are your reference. Well noting: I am not criticising here, I have various systems for various uses, money is not the key, I am only trying to understand in hope, hopefully, to help somewhat going into the direction YOU SEEK. Whatever that direction may be, even if of course not mine, no issue mate!
I am sorry that the biggest unknown here, and that's a majour one, is that you are tweaking 07 and we have just twekaed some 04 and have no clue on 07 or how close that one may be or things may be transfered. Many IFs, ouch!
Good luck whatever you way, as you deserve to have a sound that suits YOU perfectly, and I am quite confident that can be reached a way or another...
Claude
Firt of all I congrat you for your openess, sharing your findings in always a positive way, not taking whatever bad mood regarding answers that are only meant to help. Not everyone is that neutral, congrats again.
Second, as music lover, I feel very sorry that you are not happy with your current sound, and we can't leave you like this, cant' we?
The sound you like may not me the one I like, who knows, the absolute sound, well, I am a 6S black belt coming from F1 and I learnt that absolute doesn't mean a thing in physics on this Earth LOL!
So...
I have been soldering smd components since last couple of years when I started playing with class d amps. However this is the first time I have soldered a 8 leg chip. Also I cross checked every solder joint just to make sure they aren't cold joints. The sockets and pins are pretty snug.
=> OK, sounds good. Just in case of, and that isn't a lot of work and doesn't mean wasting time, may I humbly suggest, based on the pix you posted, that you simply pull out the OPA1656 together WITH their sockets, put a lot of flux on the OPA1656 legs and simply put the soldering iron across the pins / legs legs of said OPA 1656. Just across all of them with a say a 350°C iron, touching quickly. Flux means the pins only will reflow, job done. And then clean the flux. May not change anything, but at least I will feel more confident LOL.
OK, there is reason for that : quick chips like OPA 1656... and much faster ones I dealt with can be quite sensitive to micro failures we may not see...
You can do that, or not, I realise cleaning the flux is a pain, but then you have flux I suppose already on hand and you paid for the OPA 1656s, so if these can be "saved" cheaply".... win-win
As for my like-ness for non neutral sonic, I am not sure.
=> Important is indeed what you like.
NE 5532 isn't bad, but in my ears re details, response shapness, 3D image, quite in another league than amps as AD825AD 20 years ago... which is also slightly lacking IMHO vs OPA1656 everywhere, a bit, bare in the (exagerated) bass department. So at the end NE5532 vs OPA1656 in the right application (04 is such one) and with the correct OPA PS (admittely modern OPAs are more sensitive to a correct PS) should be a no brainer. Where I come from a nice Tyrell Ford from the early 70s vs a Williams of the mid 90s LOL.
Not saying neither cans be beaten nowadays, but either way it ius true 2 balanced F1 cars somehow, not lacking anything if not compared directly ;-)
I feel NE5532 is much more neutral (i also see almost every site mention them as being neutral though a bit clinical) & the OPA 1656 which are certainly much evolved chip on paper lacks that level of neutrality.
=> I found OPA1656 rather more neutral in that application and others. Sometimes you could say some bass distorsion is missing. Ok, it has some very low second harmonic that brains usualy like (need?).
But NE5532 has so many limitations... something must be wrong somewhere.
Saying that, if not comparing directly, on a balanced amp, one could be happy with NE5532, it isn't that bad that something obvious pops as a problem.
But your case is different, you can compare, and assuming the right power feed (again, OPA 1656 needs a good PS, NE 5532 probably can do with lesser ones), you should hear a difference.
Oh, and the you are! Either a right one, which means your 07 needs some compensation, or a non conventional one.
Let's reflow the board / chip, and then it leaves you with what you hear. If it has changed a lot and you like it, bingo, welcome on board. If not, then we can assume you really prefer NE5532 to 0PA1656 in your set up.
If that's the case, either you are happy that way, or you wish more flavour, in which case let us know what you would like. As things are, I have always my JFET flavour box with me, various OPAs of the best quality, all measuring the same à la ASR... and all sounding so different. Gilles knows that box, called flavour box as all OPAs are quite interchangeable (similar parameters so can be used for same application). As posted on the OPA1656 thread, we played with it, and ended up in all cases with OPA1656. Use the search function to find the other OPAs and "our " findings, just our ears in a given situation.
It sounds more warm, but that's fine as long as I am not loosing on details and soundstage. I can assure if you listen to my A07 with OPA 1656 , you will feel the loss of detail and that the soundstage became very narrow as compared to when it plays with NE 5532.
=> Please reflow, just in case, just to make sure it is your taste and not whatever else...
Also somebody said in OPA 1656 thread that what I am hearing as more details in Ne 5532 is the more noise !!!
=> Can be, indeed, happens. Is often very noticeable in the bass. If listening with small speakers and old DACs I must confess AD825 biased in class A nicely blends out (= lack of definition!!) the DACs limits in terms of resolution while adding some very (=too) solid deep bass which complements non 25Hz FC LS LOL
The world is wide...
How come the A07 with NE 5532 sounds closest to my other amps without any opamp ????
=> May I humbly ask what these systems are and what amps are your reference. Well noting: I am not criticising here, I have various systems for various uses, money is not the key, I am only trying to understand in hope, hopefully, to help somewhat going into the direction YOU SEEK. Whatever that direction may be, even if of course not mine, no issue mate!
I am sorry that the biggest unknown here, and that's a majour one, is that you are tweaking 07 and we have just twekaed some 04 and have no clue on 07 or how close that one may be or things may be transfered. Many IFs, ouch!
Good luck whatever you way, as you deserve to have a sound that suits YOU perfectly, and I am quite confident that can be reached a way or another...
Claude
Last edited:
Thanks. I will reflow the chips legs with flux again if it changes. Initially i had the chip VCC & GND bypassed with 4.7uF Nichicon + Wima FKS. Tried Vishay MKP 1837 & Vishay MKT 370 in place as well but didnt helped, so have removed them. My setup is pretty humble. Overnight Sensation speakers (with crossover upgraded parts) + Fiio K3 dac. I usually switch between my Allo Volt + D , heavily modded dual TPA 3116 chip (red board with Coilcraft flat wire inductors, Nichicon KZ decoupling close to chip, Nichicon FW at buffer position, snubber mod + Bevenbi MKPA - S input coupling caps bypassed with Vishay 1837) & A07. My heavily modded dual chip sounds the best (dynamics, decay, extension, midrange clarity, timbre) but lacks the powerful bass A07 has. I usually pair it with a diy tube pre/buffer with Voskhod 6J1P-EV which helps in making the midrange and bass sound a bit fuller. The combo sounds very very good & lively. While Aiyima A07 is most effortless in terms of the sheer amount of power and bass it puts, it simply doesn't cut across Allo Volt + D & my modded amp from fidelity viewpoint.
Last edited:
Ah, sadly you venture in a field I don't know.
We are back to 07 seeming at the end being quite a different fish than 04.
Regarding the rest of the system, can't comment either, other than a quick look on the internet on what makes IMHO the most of the sound, the LS, reveals that these are quite small speakers, defo not full range. Whatever the upgrade and how good they might sound (I don't know them), they are for sure very far from full range especialy in the bass department. Who knows what they like to be compnsated with...
The secondary system I used for this moderate 04 cheap amp testings had bass reflex speakers (quick bass loss) with 32Hz at -3dB... my very own system, the one I cherich a lot and who sees much more elaborated amps than these Class D toys, has closed cabinet speakers (slow roll off!) with 25Hz at -3dB...
Bass is not the last word of course, but I find that bass contains... not only bass when it is recorded, but a lot of room information regarding echo and location, spatialization etc.
I can't really say how your system sounds, how coudl I! but based on tour product description I would say it either needs a big boost from a subwoffer or a some kind of compensation in the low range. After decades of experience, I would say that eventhough prople tell us that walls etc. boost the bvass, it doesn't come close to the real thing.
Bottom line, and that's a pure conjecture, only in hope to be helpfull: it could weel be that you prefer for your set up an OPA with say quite a bass boost, and even some distorsions might be perceived as benefical in some cases.
If you feel you need more ommph for your taste, then you might indeed look for modern but somewhat bass heavier OPAs
If all this is not the case, please just ignore my post, sorry for interfering!
Claude
We are back to 07 seeming at the end being quite a different fish than 04.
Regarding the rest of the system, can't comment either, other than a quick look on the internet on what makes IMHO the most of the sound, the LS, reveals that these are quite small speakers, defo not full range. Whatever the upgrade and how good they might sound (I don't know them), they are for sure very far from full range especialy in the bass department. Who knows what they like to be compnsated with...
The secondary system I used for this moderate 04 cheap amp testings had bass reflex speakers (quick bass loss) with 32Hz at -3dB... my very own system, the one I cherich a lot and who sees much more elaborated amps than these Class D toys, has closed cabinet speakers (slow roll off!) with 25Hz at -3dB...
Bass is not the last word of course, but I find that bass contains... not only bass when it is recorded, but a lot of room information regarding echo and location, spatialization etc.
I can't really say how your system sounds, how coudl I! but based on tour product description I would say it either needs a big boost from a subwoffer or a some kind of compensation in the low range. After decades of experience, I would say that eventhough prople tell us that walls etc. boost the bvass, it doesn't come close to the real thing.
Bottom line, and that's a pure conjecture, only in hope to be helpfull: it could weel be that you prefer for your set up an OPA with say quite a bass boost, and even some distorsions might be perceived as benefical in some cases.
If you feel you need more ommph for your taste, then you might indeed look for modern but somewhat bass heavier OPAs
If all this is not the case, please just ignore my post, sorry for interfering!
Claude
Can't remember the shootout I did back then, search function is your friend, BUT IF that applies to your case then some LTs were maybe more in your direction / favour and for sure far superior to any 5532...
Good luck and whatever the OPA : be HAPPY with this little unit and set up, sounds promising
Claude
Good luck and whatever the OPA : be HAPPY with this little unit and set up, sounds promising
Claude
I have attached the frequency response of my speaker (Placement 1). Its pretty flat from 100hz onwards. The dip you see at 70 & 120 are room modes. Also the hump at 60 hz is room mode. It might be due to the depression in 120-125Hz that's causing a feel of thin sounding bass. But am pretty content with it. These are 4 inch speakers in a very small cabinet. The bass they do is unbelievable for their size & also my room is roughly 15x10x12 ft. I can feel the bass from 50hz onwards very prominently. I have managed to eliminate the crest & trough in bass region via some careful planning ( Placement 2). Still the trough at 120-125 hz persists(attribute thin bass which i dislike, lack of boominess that i like). I am still okay with that untill i do proper damping.
My concern with the A07 is mostly soundstage and loss of detail related. My speakers sound very very good with my diy amp & Allo Volt + D but they dont have the brute power to drive my inefficient 8 ohm speakers to insane level.
I wish i had a flavour box like yours & i could swap opamps in & out till i find the right synergy & call it a day. But since i dont have, i want to make an informed decision on which opamp should i go with. Since OPA 1656 (which presumably is best not counting burson & sparkos etc) didnt worked out for me, i guess this time i will just buy few low cost ones just to see if they are my flavour. Based on the PDF i found somwhere on the internet, i feel i will like the TLE2072ACP most. But that too is costly & i dont want to get dissappointed after waiting weeks for it.
Attachments
Last edited:
What is it you would like?
Starting from OPA1656, assuming it works correctly we should have a baseline. So what would you like to change over OPA1656 in a perfect world? What is it you like from this OPA (eg don't whish to be changed) and what is it you would change / improve?
on the other hand
"A07 is mostly soundstage and loss of detail related"
I understand from your post that your speakers are doing wonders despite their physical limits and thanks to your optimization.
It is true that there aren't many notes / instruments below 50Hz. But then most of the room information is there!
As things are, my main system enables me to switch woofer, rest of the LS, or both. It always amazes friends when they listen to the "rest of the LS" (without bass information) and then I switch only the bass which souds often as a quiet burble impossible to understand and then sum of both makes such a music. Another interesting test is the a capella song from Tracy Chapman's first album. Switch on the "rest of the LS" without bass, it sounds credible, perhaps a tad thin but I am confident many people would hear it like that. Switch on the bass onmly and, oh, there is music there!! How can that be? Switch both for the full system and... what a music. And I am dead sure she can't sing below 50hz LOL
I digress, sorry, just name what flavours you are after!
Claude
Starting from OPA1656, assuming it works correctly we should have a baseline. So what would you like to change over OPA1656 in a perfect world? What is it you like from this OPA (eg don't whish to be changed) and what is it you would change / improve?
on the other hand
"A07 is mostly soundstage and loss of detail related"
I understand from your post that your speakers are doing wonders despite their physical limits and thanks to your optimization.
It is true that there aren't many notes / instruments below 50Hz. But then most of the room information is there!
As things are, my main system enables me to switch woofer, rest of the LS, or both. It always amazes friends when they listen to the "rest of the LS" (without bass information) and then I switch only the bass which souds often as a quiet burble impossible to understand and then sum of both makes such a music. Another interesting test is the a capella song from Tracy Chapman's first album. Switch on the "rest of the LS" without bass, it sounds credible, perhaps a tad thin but I am confident many people would hear it like that. Switch on the bass onmly and, oh, there is music there!! How can that be? Switch both for the full system and... what a music. And I am dead sure she can't sing below 50hz LOL
I digress, sorry, just name what flavours you are after!
Claude
Thanks again. As to what i would like - Even i don't know. It's abstract !! 😉
But i will try to summarize. What i am looking for is that if my system can replicate the wholesome feel of the music that can be replicated again & again on various other systems. If one of the components clearly (& largely) sounds different from other bunch of equipments, its probably that either it's doing it wrong or it's just too good. And since i don't have a reference point of how the same music sounds on your system or best say it sounded in recording studio, i don't know what to expect from them. But being used to listening most of my amps without an opamp chip its quite clear that there are few things that i feel different with OPA 1656.
1. Lack of Extreme highs (the sparkle in highs i get with other amps have gone, highs are clean & clear but lacks musicality)
2. The seperation between instruments lessened (the sound feels more congested, the notes would fly all over the place with NE 5532, with OPA its very disciplined & it makes it sound boring)
3. Narrow soundstage (music seems to come out in between the speakers only, mostly from center, previously it used to be spread even outside of speakers on both end)
Apart from these, bass is godly. Very powerful & punchy after OPAMP upgrade but the lively feel has gone. I will see if i can record a few audio clips just to demonstrate the effect i have explained here, but i have just a basic lapel mic. I dont know if that will be a good comparison.
With all said I still feel 1656 sounds good but in comparison to NE 5532 it falls sort in my Aiyima A07.
I can't live with the stuffy small soundstage. I want my sparkling airy highs back but not at the cost of harshness in mids or highs. I also don't care if its noise or whatever.
See, I don't know what faithful reproduction of music is (as a listener we never know). My belief is no matter how ruler flat our speakers are, it will still sound different if we put it in a different room (especially when a lot of good music are mixed and mastered on barebones speakers, i honestly don't think they are all flat, i work in movies so telling you from my personal experience). So unless we can mimic the same enviroment settings, from furnitures to equipments in the signal chain, material of driver cones to even the height of recording studio from sea level it won't be the same. So, I am not chasing for that unicorn accuracy, that can never be achieved.
All i am looking for is musicality !!! I told you it's abstract 😉
Last edited:
It is quite odd TBH, as what you consider strong points are rather points where the OPA1656 is "just good", whereas what you seem to dislike more are exactly OPA1656s forte!
Very odd...
The only case where I had your findings was ages ago when the OPA's socket wasn't engaged properly in the board socket... I would defo check that and also reflow those legs.
NOW, should you wish what you stated, and perhaps it is what your system or your ears need, you could try other OPAs.
Very odd...
The only case where I had your findings was ages ago when the OPA's socket wasn't engaged properly in the board socket... I would defo check that and also reflow those legs.
NOW, should you wish what you stated, and perhaps it is what your system or your ears need, you could try other OPAs.
Based on your last post, I had a quick look at my conclusions of my last shootout on my test bench. Not saying these OPA would give the same sonic results in your case, but based on the similar implementation it is very likely (unless you have a defective OPA PS etc. of course).
Did you give ADA4627-1 a try? Could suit your taste, or if you should want even more in direction speed and treble, move to ADA4625-1?
ABSTRACT...
5532
Just for reference, good old one I have since decades. Noisy, bass has impact and reaches low but feels slow and not well defined, clarity is poor, airyness acceptable, highs seems rounded off and metalic, music doesn’t flow well and integration of all music registers is poor, as is imaging. It shows its age, wouldn’t recommend it today…I didn’t 20 years ago either !
LT 1169
Supposed to be kind of a modern upgrade. Bass is faster, treble much better, integration as a hole better, more details, flow is still average, still metalic sounding, bass reaches deep and is predominant. Wouldn’t really recommend it, there are better choices.
LT 1113
That’s the first unit that is acceptable. Integration and treble are better, bass is more differentiated / faster / more impulsive, it sounds less metalic, still mainly 2D imaging but it starts to get 3D, music flows and it is quite expressive around a somewhat rounded off nature. It is quite enjoyable but flow remains average, as are details. A possible option, easy fit in bi gold size op amps, well noting one can do better in all registers with other parts. Based on that I recognised the usual performance and sonic signature LT units had in the past, I could try a double LT1792 (SMD) but fear it won’t get me nowhere vs hassle so stopped there my LT exprience. This is a fit and forget 1 :1 part, just doing the job as a good work horse… no race horse though.
ADA4625-1
I had a lot of expectations on that one, as its spreadsheet reads as one of the best FET op amps around. And indeed it shone with less noise, excellent clarity, a lot of speed, a good integration of all registers and music instruments, treble seemed a bit pronounced and bass a bit dry and less diffenciated but still low reaching and very impulsive / fast. Music flows, spacialty is good, airiness aswell with lot of atmosphere, it is enjoyable ! It though sounds a bit too clean/dry, not metalic though. Listening long term revealed it could be less relaxed and less enjoyable than his smaller (warmer) brother, the ADA4627-1. At the end this is one that could fit if one needs or wants a very fast and clean sound while not sounding too sterile. Still on the dry side for me.
ADA4627-1
I had very low expectations on that one as basicaly it is supposed to be a lower grade ADA4625-1. Well, it is not in audio terms IMHO ! It does nearly all the the ADA4625-1 does… but better, much better in terms of musicality ! The bass is not as dry while still fast, details are still present, integration and tonality are excellent, it doen’t sound as dry and is rather on the neutral / warm side with lot of expressivity. Speed, flow, end of notes, great atmosphere, 3D spaciality, fun factor, body and flesh : all very good to excellent. Treble could perhaps be a tad more precise. This is a top op amp, one of the best I ever heard and TBH it has my recommandation.
OPA1656
In the past I have to say I prefered widely the AD precise sound vs the softer traditional TI sound. I used though the better TIs in output stages of middle class DACs as these benefitted from a a more round /warm presentation with less details, as that masked their weaknesses and in fact improved the situation. It did not to my ears when having higher quality stuff and there AD’s speed, airiness, transparency shone, provided you avoided the bright ones. I tried this TI though because it was recommended here. And I must say that it differs from past TIs ! While not having that traditional « what people call old good tube sound », it is still a bit on the warm and musical side while being more more « open » than the TIs I knew. I found it had weight, impact, some airiness, a very good 3D soundstage, lot of details, noise is really very low, body of the instruments so enjoyable, an open treble that though tends a very small tad to the metalic side. This is also a keeper, probably one of the very best high quality all rounder provided you like it a tad warm rather than bright.
.....
I hope this helps
Good luck with your quest
Claude
Did you give ADA4627-1 a try? Could suit your taste, or if you should want even more in direction speed and treble, move to ADA4625-1?
ABSTRACT...
5532
Just for reference, good old one I have since decades. Noisy, bass has impact and reaches low but feels slow and not well defined, clarity is poor, airyness acceptable, highs seems rounded off and metalic, music doesn’t flow well and integration of all music registers is poor, as is imaging. It shows its age, wouldn’t recommend it today…I didn’t 20 years ago either !
LT 1169
Supposed to be kind of a modern upgrade. Bass is faster, treble much better, integration as a hole better, more details, flow is still average, still metalic sounding, bass reaches deep and is predominant. Wouldn’t really recommend it, there are better choices.
LT 1113
That’s the first unit that is acceptable. Integration and treble are better, bass is more differentiated / faster / more impulsive, it sounds less metalic, still mainly 2D imaging but it starts to get 3D, music flows and it is quite expressive around a somewhat rounded off nature. It is quite enjoyable but flow remains average, as are details. A possible option, easy fit in bi gold size op amps, well noting one can do better in all registers with other parts. Based on that I recognised the usual performance and sonic signature LT units had in the past, I could try a double LT1792 (SMD) but fear it won’t get me nowhere vs hassle so stopped there my LT exprience. This is a fit and forget 1 :1 part, just doing the job as a good work horse… no race horse though.
ADA4625-1
I had a lot of expectations on that one, as its spreadsheet reads as one of the best FET op amps around. And indeed it shone with less noise, excellent clarity, a lot of speed, a good integration of all registers and music instruments, treble seemed a bit pronounced and bass a bit dry and less diffenciated but still low reaching and very impulsive / fast. Music flows, spacialty is good, airiness aswell with lot of atmosphere, it is enjoyable ! It though sounds a bit too clean/dry, not metalic though. Listening long term revealed it could be less relaxed and less enjoyable than his smaller (warmer) brother, the ADA4627-1. At the end this is one that could fit if one needs or wants a very fast and clean sound while not sounding too sterile. Still on the dry side for me.
ADA4627-1
I had very low expectations on that one as basicaly it is supposed to be a lower grade ADA4625-1. Well, it is not in audio terms IMHO ! It does nearly all the the ADA4625-1 does… but better, much better in terms of musicality ! The bass is not as dry while still fast, details are still present, integration and tonality are excellent, it doen’t sound as dry and is rather on the neutral / warm side with lot of expressivity. Speed, flow, end of notes, great atmosphere, 3D spaciality, fun factor, body and flesh : all very good to excellent. Treble could perhaps be a tad more precise. This is a top op amp, one of the best I ever heard and TBH it has my recommandation.
OPA1656
In the past I have to say I prefered widely the AD precise sound vs the softer traditional TI sound. I used though the better TIs in output stages of middle class DACs as these benefitted from a a more round /warm presentation with less details, as that masked their weaknesses and in fact improved the situation. It did not to my ears when having higher quality stuff and there AD’s speed, airiness, transparency shone, provided you avoided the bright ones. I tried this TI though because it was recommended here. And I must say that it differs from past TIs ! While not having that traditional « what people call old good tube sound », it is still a bit on the warm and musical side while being more more « open » than the TIs I knew. I found it had weight, impact, some airiness, a very good 3D soundstage, lot of details, noise is really very low, body of the instruments so enjoyable, an open treble that though tends a very small tad to the metalic side. This is also a keeper, probably one of the very best high quality all rounder provided you like it a tad warm rather than bright.
.....
I hope this helps
Good luck with your quest
Claude
Claude, could you clarify few points, please?
- this OPA bypass with 0.1uF is recommended for any OPA, or specific for 1656?
- if it suits to any OPA, maybe under-board position will be more convenient?
- is placing it on the DIP8 to make it closer to legs so critical?
Was thinking to put 10uF Elna+10uF X7R on the DIP8, since it specific for 1656, and put 0.1uF under the board, if it suitable for any OPA (in case of changing OPA).
Is it acceptable solution?
Thanks for great tutorial, btw 🤝
Hi,
In a few words...
It adds contacts, parasitic capacitances and resistances etc., Google is your friend, many threads on that. this is no black science, just logic, again please refer to spec sheets and papers from TI, AD etc., very useful IMHO
Was thinking to put 10uF Elna+10uF X7R on the DIP8, since it specific for 1656, and put 0.1uF under the board, if it suitable for any OPA (in case of changing OPA).
Is it acceptable solution?
=> Let me be more precse on that.
quote "10uF Elna+10uF X7R on the DIP8, since it specific for 1656".
Nope, not really. It is NOT specific to OPA1656. And in a perfect world, if the Aiyima was properly designed, all this wouldn't be required probably,. It all depends on the power supply for the OPA. In fact, these additional caps are what we found sounded best, in the case of OPA1656 indeed as the fine tuning was done with it, but it is somewhere more Aiyima specific (and somewhat related to OPA1656 FOR THIS APPLICATION).
What is suitable for another OPA, what is suitable to your ears, I can't tell "sadly". We can only share that we liked what we ended up with. All is related. If we would have tweaked the PS differently, perhaps we would have ended up with another PS bypass combo, who knows. Same with another OPA.
As far as the second part of your question is concerned, again, the shorter the better and you can put the caps where you want based on this logic, be it under or above... or remote. A further hint: the smaller the cap's value, the clostest it has to be to the OPA legs. And that is regardless the OPA, just common sense and electronic. So in summary:
I hope this helps
Enjoy tweaking
Claude
In a few words...
- this OPA bypass with 0.1uF is recommended for any OPA, or specific for 1656?
- if it suits to any OPA, maybe under-board position will be more convenient?
- is placing it on the DIP8 to make it closer to legs so critical?
It adds contacts, parasitic capacitances and resistances etc., Google is your friend, many threads on that. this is no black science, just logic, again please refer to spec sheets and papers from TI, AD etc., very useful IMHO
Was thinking to put 10uF Elna+10uF X7R on the DIP8, since it specific for 1656, and put 0.1uF under the board, if it suitable for any OPA (in case of changing OPA).
Is it acceptable solution?
=> Let me be more precse on that.
quote "10uF Elna+10uF X7R on the DIP8, since it specific for 1656".
Nope, not really. It is NOT specific to OPA1656. And in a perfect world, if the Aiyima was properly designed, all this wouldn't be required probably,. It all depends on the power supply for the OPA. In fact, these additional caps are what we found sounded best, in the case of OPA1656 indeed as the fine tuning was done with it, but it is somewhere more Aiyima specific (and somewhat related to OPA1656 FOR THIS APPLICATION).
What is suitable for another OPA, what is suitable to your ears, I can't tell "sadly". We can only share that we liked what we ended up with. All is related. If we would have tweaked the PS differently, perhaps we would have ended up with another PS bypass combo, who knows. Same with another OPA.
As far as the second part of your question is concerned, again, the shorter the better and you can put the caps where you want based on this logic, be it under or above... or remote. A further hint: the smaller the cap's value, the clostest it has to be to the OPA legs. And that is regardless the OPA, just common sense and electronic. So in summary:
- do what you want / what you can do
- go for what you like best
- but keep legs and wires as short as possible and if in doubt always favour the smallest caps so that it is closer to its destination.
- Personal addition: try to do reversible mods, so you can go back if needed without big hassle
I hope this helps
Enjoy tweaking
Claude
Last edited:
The "Short as possible" is a very real thing. Once upon a time I was working on creating some wideband analog PCB assemblies. The VIAS connecting one side of the PCB to the other had enough inductance to form little parasitic resonance circuits. I learned the Solder-caps-across-Opamp-power-legs technique from the folks at Museatex. Ed Meitner and his team applied a whole lot of little tweaks (Like pulling current from op-amp outputs to the negative rail) long before they became generally accepted. BTW. That last one may be useful in this situation. Just connect a ~1.8k resistor from the opamp output to the negative rail. It forces the amp to counter the DC current pull. Depending on topology, sometimes puling to the positive rail is more beneficial.
Quote: "Ed Meitner and his team applied a whole lot of little tweaks (Like pulling current from op-amp outputs to the negative rail) long before they became generally accepted. BTW. That last one may be useful in this situation. Just connect a ~1.8k resistor from the opamp output to the negative rail. It forces the amp to counter the DC current pull. Depending on topology, sometimes puling to the positive rail is more beneficial."
Ah, been there, done that
Perhaps I didn't mention here what I did re OPAs...
That interesting trick is called Class A biasing, and it can be done with a resistor but with some negatives, it is better (for the sound) to use resistors AND transistors to do that. Google is your friend, admittely more on papers from the 90s and subsequent Audio forums (as of me I didn't have internet in the early 90s and Audio disussions where a bit later more groups on Yahoo - someone though kept most of ours!)
My favourite output stage, for decades, has been AD825 biased in Class A with (not kidding) genuine 2SK170. Inexpensive then. A bit more complicated though to fit around an OPA than our caps LOL. It is still the output stage of my DAC since 2001, although it would nowadays benefit from an OPA1656 (apart perhaps from bass punch, TBD). One day, time permitting, argh!
Long story short, I tried this trick on OPA1656 and it does NOT improve its sound. I posted it here on the OPA1656 thread 2y ago to its creator and the short answer is I should have known or asked first, as it was very foreseeable. OPA1656 does not respond to that because it has something else included in its schematic that does a similar trick and that isn't compatible with this trick.
As as side note, I also posted sonic results, the post above is an abstract from it. To my ears and in various applications (see original thread) OPA1656 stand alone sounded both better than AD825 stand alone and still had the edge over AD825 biased in Class A. As said, OPA1656 sounded slightly better without Class A bias.
Bottom line: IMHO not worth explorating, been there and dead end (for me). But of course this is DIY and no harm trying. That result is quite OPA1656 and topology depending: some OPAs respond well - or not - to this trick.
Have fun tweaking
Claude
Ah, been there, done that
Perhaps I didn't mention here what I did re OPAs...
That interesting trick is called Class A biasing, and it can be done with a resistor but with some negatives, it is better (for the sound) to use resistors AND transistors to do that. Google is your friend, admittely more on papers from the 90s and subsequent Audio forums (as of me I didn't have internet in the early 90s and Audio disussions where a bit later more groups on Yahoo - someone though kept most of ours!)
My favourite output stage, for decades, has been AD825 biased in Class A with (not kidding) genuine 2SK170. Inexpensive then. A bit more complicated though to fit around an OPA than our caps LOL. It is still the output stage of my DAC since 2001, although it would nowadays benefit from an OPA1656 (apart perhaps from bass punch, TBD). One day, time permitting, argh!
Long story short, I tried this trick on OPA1656 and it does NOT improve its sound. I posted it here on the OPA1656 thread 2y ago to its creator and the short answer is I should have known or asked first, as it was very foreseeable. OPA1656 does not respond to that because it has something else included in its schematic that does a similar trick and that isn't compatible with this trick.
As as side note, I also posted sonic results, the post above is an abstract from it. To my ears and in various applications (see original thread) OPA1656 stand alone sounded both better than AD825 stand alone and still had the edge over AD825 biased in Class A. As said, OPA1656 sounded slightly better without Class A bias.
Bottom line: IMHO not worth explorating, been there and dead end (for me). But of course this is DIY and no harm trying. That result is quite OPA1656 and topology depending: some OPAs respond well - or not - to this trick.
Have fun tweaking
Claude
Last edited:
These tricks were still pretty new in the 80's when I worked with Ed. A lot of the museatex stuff used siliconix CR390's On AD829's and AD845's. All the 2SK147BL's that didn't meet binning requirements ended up pulling current source duties. I still have a few dozen of them (with 2SJ72's) We cheated and used resistors on the line of ADS car amps Ed designed the topology for. Even a resistor biasing provided benefits on 5532's.Quote: "Ed Meitner and his team applied a whole lot of little tweaks (Like pulling current from op-amp outputs to the negative rail) long before they became generally accepted. BTW. That last one may be useful in this situation. Just connect a ~1.8k resistor from the opamp output to the negative rail. It forces the amp to counter the DC current pull. Depending on topology, sometimes puling to the positive rail is more beneficial."
Ah, been there, done that
Perhaps I didn't mention here what I did re OPAs...
That interesting trick is called Class A biasing, and it can be done with a resistor but with some negatives, it is better (for the sound) to use resistors AND transistors to do that. Google is your friend, admittely more on papers from the 90s and subsequent Audio forums (as of me I didn't have internet in the early 90s and Audio disussions where a bit later more groups on Yahoo - someone though kept most of ours!)
My favourite output stage, for decades, has been AD825 biased in Class A with (not kidding) genuine 2SK170. Inexpensive then. A bit more complicated though to fit around an OPA than our caps LOL. It is still the output stage of my DAC since 2001, although it would nowadays benefit from an OPA1656 (apart perhaps from bass punch, TBD). One day, time permitting, argh!
Long story short, I tried this trick on OPA1656 and it does NOT improve its sound. I posted it here on the OPA1656 thread 2y ago to its creator and the short answer is I should have known or asked first, as it was very foreseeable. OPA1656 does not respond to that because it has something else included in its schematic that does a similar trick and that isn't compatible with this trick.
As as side note, I also posted sonic results, the post above is an abstract from it. To my ears and in various applications (see original thread) OPA1656 stand alone sounded both better than AD825 stand alone and still had the edge over AD825 biased in Class A. As said, OPA1656 sounded slightly better without Class A bias.
Bottom line: IMHO not worth explorating, been there and dead end (for me). But of course this is DIY and no harm trying. That result is quite OPA1656 and topology depending: some OPAs respond well - or not - to this trick.
Have fun tweaking
Claude
Interesting... and indeed it must have been quite unknown to the main public in the 80s!
I was too young back then, got into audio (as follower and reader, not tweaker!) in 84... and improved only slightly since then
The important bit I noticed with this biasing is that you need to determine what level you want (up to a limit the part can cope with). That depends a lot on your needs, the input impedance of the next section of your component etc., as all you need is to stay in Class A for the max needed voltage swing so not running out of mA at the output of the OPA. In the late 90s I had a rig which enabled me "outboard tuning of the Class A bias with a variable resistor", very handy.
Time flies, you must have had a tremendous experience working with those guys, lucky you
Claude
I was too young back then, got into audio (as follower and reader, not tweaker!) in 84... and improved only slightly since then
The important bit I noticed with this biasing is that you need to determine what level you want (up to a limit the part can cope with). That depends a lot on your needs, the input impedance of the next section of your component etc., as all you need is to stay in Class A for the max needed voltage swing so not running out of mA at the output of the OPA. In the late 90s I had a rig which enabled me "outboard tuning of the Class A bias with a variable resistor", very handy.
Time flies, you must have had a tremendous experience working with those guys, lucky you
Claude
Thanks for such detailed answer! 🤝
***All OPAs are not interchangeable
Of course. I meant only compatible ones. Actually, there's only few OPAs in my list to check: 1469, 2114, Muse2. So this 0.1uF bypass should fit them all in this Aiyima, right?
***Critical is the short distance, not the DIP8
Sure, thats what I meant. Question is "Is under-board placement short enough?" =)
***Nope, not really. It is NOT specific to OPA1656. What is suitable for another OPA, what is suitable to your ears, I can't tell "sadly".
Indeed! I meant "Your solutition specific for 1656 in Aiyima design based on your experience"
***As far as the second part of your question is concerned, again, the shorter the better and you can put the caps where you want based on this logic, be it under or above... or remote.
I was thinking to make it as you did it with that 'lytic bypasses (under the board with shortest legs as possible).
Question is, is it short enough and length is acceptable? As you mentioned before, you put it under mainboard during the tests, and moved it on DIP adapter later. So this is not so critical in this case, right?
Again, thanks for such detailed answer and all previously shared experience!
You inspired me for this tweaking adventure)
***All OPAs are not interchangeable
Of course. I meant only compatible ones. Actually, there's only few OPAs in my list to check: 1469, 2114, Muse2. So this 0.1uF bypass should fit them all in this Aiyima, right?
***Critical is the short distance, not the DIP8
Sure, thats what I meant. Question is "Is under-board placement short enough?" =)
***Nope, not really. It is NOT specific to OPA1656. What is suitable for another OPA, what is suitable to your ears, I can't tell "sadly".
Indeed! I meant "Your solutition specific for 1656 in Aiyima design based on your experience"
***As far as the second part of your question is concerned, again, the shorter the better and you can put the caps where you want based on this logic, be it under or above... or remote.
I was thinking to make it as you did it with that 'lytic bypasses (under the board with shortest legs as possible).
Question is, is it short enough and length is acceptable? As you mentioned before, you put it under mainboard during the tests, and moved it on DIP adapter later. So this is not so critical in this case, right?
Again, thanks for such detailed answer and all previously shared experience!
You inspired me for this tweaking adventure)
Hello, I hope someone can help me!
I a couple of weeks ago I got a couple of Aiyima A07 amps to build a surround system.
For the main speakers I use an amp which I just built using 60W Ultra-low distortion composite amp kits from Kaltecs.
I read a good reviews about Aiyima so I thought it would be a nice fit.
I was absolutely amazed how small they were but this was the only good thing!
First, I tried them with my main speakers. The sound was not good at all - hum, no clarity, very poor dynamics, feeble bass, terrible highs.
I read that op amps rolling is very good mod so I got OPA1656 and OPA1612 from an eBay seller with 100% positive feedbags.
I installed the new op amps. First OPA1656, the sound got little bit better, highs got improved, but still not clarity and hum in the speakers (hearable 10" away from the mid range), poor bass. Then I tried OPA1612. The sound got little bit better than OPA1656 but still far away from the quality that I expected.
Second, output power. The amps are rated 2x300W. Ha, ha, ha, what a joke. If this is 2x300W, then my 60W kits must be 600W!!!
Third, I tried them with my surround speakers. In this case the difference in the sound quality compared to my main amp wasn't so huge, but still very long way to go.
So, as I already modded the amps I can't return them and I was wondering are there any other mods that I can try? Maybe my speakers are not good match for Aiyima amps. Maybe the power supply is not good.
Any help will be highly appreciated!!!
I a couple of weeks ago I got a couple of Aiyima A07 amps to build a surround system.
For the main speakers I use an amp which I just built using 60W Ultra-low distortion composite amp kits from Kaltecs.
I read a good reviews about Aiyima so I thought it would be a nice fit.
I was absolutely amazed how small they were but this was the only good thing!
First, I tried them with my main speakers. The sound was not good at all - hum, no clarity, very poor dynamics, feeble bass, terrible highs.
I read that op amps rolling is very good mod so I got OPA1656 and OPA1612 from an eBay seller with 100% positive feedbags.
I installed the new op amps. First OPA1656, the sound got little bit better, highs got improved, but still not clarity and hum in the speakers (hearable 10" away from the mid range), poor bass. Then I tried OPA1612. The sound got little bit better than OPA1656 but still far away from the quality that I expected.
Second, output power. The amps are rated 2x300W. Ha, ha, ha, what a joke. If this is 2x300W, then my 60W kits must be 600W!!!
Third, I tried them with my surround speakers. In this case the difference in the sound quality compared to my main amp wasn't so huge, but still very long way to go.
So, as I already modded the amps I can't return them and I was wondering are there any other mods that I can try? Maybe my speakers are not good match for Aiyima amps. Maybe the power supply is not good.
Any help will be highly appreciated!!!
