Which Op-Amp? (Afraid to ask really...)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am a long time tube guy that has some sand experience. I recently built a couple op amp RIAA preamps with the BA4560 and the TLO82CP. The TLO82 definitely has an edge over the 4560 -- there is a lack of harshness on the high end that is clearly apparent. That has made me want to explore more sand amps while I build some tube RIAA preamps. I've looked on the site for well over a week and these are the op amps i've found highly praised on DIYAUDIO. I am looking for input on these (they have current digikey cost). I want to do some swapping, but I don't want them all... I only need a couple really... if you could get 4 or 5 of these, which would you get? Or would you get something else. PLEASE NOTE, I am using 8 pin DIP sockets, so no SOIC, etc... DIP available only please!! Thanks!


LME49720 -- $4.53

LM4562 -- $4.53

AD826 -- $5.60

AD823 -- $5.90

AD797 -- $8.76

AD711 -- $2.66

OPA228 -- $5.02

OPA227 -- $3.53

OPA37 -- $2.57

TLE2071 -- $1.75
 
Which are designed for audio? Check the data sheets, if it was designed for audio it will say so, no point in making things more difficult. I would explore the ones specifically designed for audio first, like the dinosaur, but still usable TL072 and 5532, and some newer parts from national and Texas Instruments designed for superior audio quality. Pick something intended for audio use, slew rate matters, noise matters, distortion matters, whereas input current, DC offsets and drift don't matter a damn!

Ted
 
Hi,

I disagree in part.
Distortion values (2nd and 3rd) and slewrate values as printed in the datasheets are very unimportant for audio usage like high level analog amplification, filtering and phono (except for ADCs and DACs).
Noise figures do matter as do Offset and input currents, especially for Phono where the highest gains will occur. High input bias currents of bipolar lownoise OPamps can create output offset voltages of several hundreds of mV! So if You use such OPamps You will have to keep the Offset-issue in mind.
Depending on the toplogy of the circuit You choose the OPs.
My favourites are 2-stage circuits with splitted passive/active RIAA-equalization. A first linear amplification stage with dc-servo or dc-blocking cap (20Hz after RIAA/IEC), followed by a passive 2120Hz filter and a second gain stage with active 50Hz-500Hz-equalization. This gives the best SNR-figures, the most precise equalization, very high flexibility and low distortion (with senible effort regarding cost and complexity of the circuit).
When You use the stage for MMs only You can choose bipolar as well as JFET-OPs for the first stage, when using MCs only a few bipolar OPs meet the noise requirements. For the second stage I´d always choose a JFET OP (high source impedances from the filterstage before).
Something like this one: http://www.matuschek.net/platina/

jauu
Calvin
 
Administrator
Joined 2007
Paid Member
OPA604 ( single )and OPA2604 (dual ) can sound very tube like with sympathetic design. Even harmonics predominate... still my first choice. Add a 0.1mfd on the supply pins directly if poss... very carefully solder to the top of the pins. You won't be dissappointed.
 
Serge66, thanks for the link will read it in its entirety later this evening.

I should add, I do not need help with schematics, etc at this point. Merely labeling a part AUDIO doesn't sell it to me either.

I have a really nice table, stereo and speakers, so it's easy to tell things apart. So to kind of restate what I was wanting -- I am looking for SUBJECTIVE input on these opamps. Which have you used and LOVED because it sounds great. I am not wanting to debate specs or topologies in the least.

Again, looking for the "best sound" -- subjective opinions are most welcome.
 
Sorry for the bad news, but no one op amp has the “best sound”. They all have different subjective strengths and weaknesses. What’s chosen depends on the application, surrounding circuit, and personal taste. Unfortunately there’s no substitute for personal experience regarding which one that’s preferred. If you really want to find what you like, you’ll have to make a longer list than just a couple based on other’s recommendations. I’ve tried over 25 different types myself, and still remain to be pleased when comparing them to a good discrete Lin design.
 
Personally I find I care about the sound of the music that moves me. I want the opposite from my equipment, that has to be as neutral as possible. Trying to select opamps for subjective qualities is a fools errand because it all depends how you use them. Chose them because they do the job of creating a neutral channel, that is something that can be done.

Ted
 
LME49720 -- $4.53 = "Dual High Performance, High Fidelity Audio" ( http://www.national.com/pf/LM/LME49720.html ) ... low THD+N and a good slew rate.

LM4562 -- $4.53 = Ferd's likes 'em enough to be a part of a "mod" upgrade to NAD CD player ( http://www.ferdsaudio.com/CDmods/C542.htm )

AD826 -- $5.60 = Op-amp input as part of A to D converter ... not for the playful or faint of heart. ( http://focus.ti.com/docs/prod/folders/print/ads826.html )

AD823 -- $5.90 = "" ""

AD797 -- $8.76 = Single channel for Instrumentation use ... very good specs but at least 2 req. for audio ( http://www.analog.com/en/other/militaryaerospace/ad797/products/product.html )

AD711 -- $2.66 = single channel instrumentation amp, again, best to look at the AD712 ( http://www.analog.com/en/prod/0,2877,AD712,00.html ). IMOP Analog makes some of the best, but steal your circuit suggestions from TI or National = they cheat and the overall results may be equal or better.

OPA228 -- $5.02 = TI / Burr-Brown :D = http://focus.ti.com/docs/prod/folders/print/opa228.html ... always a good choice, can't go wrong ( http://www.ti.com/lit/gpn/opa228 )

OPA227 -- $3.53 = """ """

OPA37 -- $2.57 = "Ultra-Low Noise, Precision ..." = http://www.ti.com/lit/gpn/opa37.pdf or search

TLE2071 -- $1.75 = "Excalibur Low-Noise High-Speed JFET-Input ..." JFET is GOOD, but a bugger to get perfect

Single channels for great separation, dual channels for easy circuit design ... :smash:
 
LME49720 and LM4562 are the same part.

Do you need singles or duals? If you want singles, it's very hard to go past OPA627 (the new OPA827 is good, but only available in SMD).

In duals, my personal favourite, or rather, my personal favourite which is available in a dual, is OPA2107 - very slightly warm, not so much dark as it is "not bright", and quite precise. LM4562 is also very good, but somewhat more analytical - a very good choice if you don't need to drive low impedance loads. If you're after something a bit cheaper, OPA2227 is good, and easy to work with, unlike its brother, OPA2228.

In singles, there's a fair amount of choice - while OPA627 is basically the undisputed best, AD843 is a good opamp, and it has a "smooth" sound, which may be preferable if you like your amps tube-y. The single version of OPA2227, OPA227, is also a nice, inexpensive choice. AD797 is also well worth of a look.
 
If I'm selecting, I look for low noise, decent bandwidth (slew rates ~ 20 or so for audio) and best THD+N => 0.0000.....000000 (four or more place to the right of the decimal is going to be close to undetectable to us mortals) ... a 115 db noise floor is nice as well (PSRR).

I also pay close attention to Texas Instruments and National's recommendations for power supply bypassing caps. (How do they get those great curves and specs? Plastic caps right across the chip power pins of the eight and sixteen DIP chips = they cheat = solder them right to the pin legs ... AND ... they use their own linear regulators for +/0/- 15 volts ala '7815 et al.)

Working with dual channel and four channel op-amps, where the signal is being massaged through the same chip, like in a band pass filter or a balanced to unbalanced line driver, but don't use dual channel (or 4 channel) op-amps for both sides of a stereo signal = the cross talk inside the chip ruins the distortion numbers and separation.

Best modern advise = read Bob Pease's columns and his white papers at both TI and National web sites. (Best ancient history advise = Don Lancaster's "Op-Amp CookBooks".) :smash: ... but you can't beat Analog Devices for as good as they get ...
 
Hi,

I am not wanting to debate specs or topologies in the least.
When You´re familiar with schematic design than You should know that the sonic footprint of a circuit depends more on the circuit topology in which it is used than anything else :att'n:
Example. There´s a hype towards ever faster and extrem low distortion OPs. But it becomes increasingly difficult with normal DIY-methods to design the circuit stable. Unfortunate PCB-Design and power supply decoupling and the thingies oscillate. How will they sound then?
You want to discuss sonics of a device without stating for which usage it is intended? Not really, or? :rolleyes:
Just as a rough personal guide. I prefer JFET-OPs in any but extreme lownoise/very low source-impedance circuits over bipolars.
And if I just need a buffer with gain 1, than a discrete JFET-stage is my way to go.

jauu
Calvin
 
You're looking for subjective listening expierence? Well, I only used opamps with higher gain circuits so far. I seem to be the only one at this site who doens't like the OPA627/OPA637: to smooth allover and fluffy in the bass. An ordinary OP27 sounds better to me, as does the ancient NE5534. My personal favorite ist the AD797, which can be used in real unsane high gain applications as well: accurate bass, clarity all over and crisp&clean upper freq. range. I'm having troubles using the LT1028 regulary (low level oscillations whatever I do). I did not check newer ones as the LME audio series however.

Rüdiger
 
Calvin: " ... I prefer JFET-OPs in any but extreme lownoise/very low source-impedance circuits over bipolars. ... And if I just need a buffer with gain 1, than a discrete JFET-stage is my way to go. ..."

What he said ... :D

Onvinyl: http://www.analog.com/en/OP27/productSearch.html for a variety of this Analog '27x series of op-amps. This is a decent audio dual op-amp ... http://www.analog.com/en/amplifiers...mplifiers-op-amps/op279/products/product.html = I would use this for a balanced input to gain stage pre-amp with input buffering and protection added (gain = 1 as per Calvin above) for audio only as the slew rate/bandwidth is a bit low for HF work.

Onvinyl: http://www.analog.com/en/other/militaryaerospace/ad797/products/product.html ... = much better specs = better noise rejection and better "clarity" in the higher freq. end of audio = slew rates ~ 20 ;) = This would handle everything coming out of the best DVD-A players (120 db bandwidth = "... low total harmonic distortion of -120 dB at audio bandwidths ...").
 
Sorry for the bad news, but no one op amp has the “best sound”. They all have different subjective strengths and weaknesses. What’s chosen depends on the application, surrounding circuit, and personal taste.

Well, wasn't that the question though? And I'd bet that 90% of op-amp RIAA circuits are a couple of passive EQ designs or the common feedback active eq design -- most with very little difference in topology, only certain selected values changing. The RIAA curve is fixed, and for the vast majority of circuits, it's pretty "fixed" too. Certainly in the realms of the pre Im building at this point.

LME49720 and LM4562 are the same part.

I actually wondered that, since they are the same price and info at digikey. What's the deal with the different numbers?

are you looking into moving coil or moving magnet designs? the answer has impact on the best choices ...

I have several MM carts now and I am definitely going to get a couple of the Denon MCs to try in the future. So both eventually... but MM for now.

Trying to select opamps for subjective qualities is a fools errand because it all depends how you use them.

Nah. Sorry. I've been building high-end equipment far too long to believe that. I'll take someone's opinion with a good ear over a technical argument any day. If the tech argument held, I and others, wouldn't still be building SS killing tube gear. I can guarantee you - it's not all theory. At least not until we understand it completely, and we aren't there yet.

When You´re familiar with schematic design than You should know that the sonic footprint of a circuit depends more on the circuit topology in which it is used than anything else.

As above. It's the sum of all the parts and all the choices you make. You give the circuit itself far too much credit.

Now, I want to thank everyone who has posted and will post. I do appreciate your input even if I don't agree.

So let me say it again, I want your subjective input on these things. We don't need to discuss MM or MC, or active or passive eq. I want opinions like this, "man, back in 1999, I built a MM pre using PART XX and it just sounded great. i haven't stopped using it."

While some say you can't select based on subjective input, that is; however, exactly what I intend to do. I want to know what moved you. All tubes sound different, even in the same brand. People discuss that all the time, without having to have an established schematic. Im looking for that same thing... like, "I ALWAYS use ACME OP#7s cause they rocked me and I've not found anything better."

It's that simple! Thanks all! -Sean
 
Onvinyl said:
You're looking for subjective listening expierence? Well, I only used opamps with higher gain circuits so far. I seem to be the only one at this site who doens't like the OPA627/OPA637: to smooth allover and fluffy in the bass. An ordinary OP27 sounds better to me, as does the ancient NE5534. My personal favorite ist the AD797, which can be used in real unsane high gain applications as well: accurate bass, clarity all over and crisp&clean upper freq. range. I'm having troubles using the LT1028 regulary (low level oscillations whatever I do). I did not check newer ones as the LME audio series however. Rüdiger

I missed this in my quotes, but this is the kind of info... general but accurate descriptions of what the music is like. Thanks Onvinyl!
 
" ... this is the kind of info... general but accurate descriptions of what the music is like. ..."

It is of course highly subjective. I have always been of the opinion that the electronics should make no contribution to the quality of the music, the musicians being the ultimate authority ...

However after listening over the last year or so to digital sources played through a vacuum tube pre-amp (out to solid state amps) that the equipment can indeed contribute, at least to the extent of "curing" or otherwise reducing the unintended, accidental (or not) irregularities of the engineers, producers and manufacturers. (You know the type: "Oh, by the way, which one's Pink ...")

Whether a particular op-amp can justly be called better or not, compared to the others, the spec being relatively equal, is a good question. Many have discovered that the capacitor and resistor types and qualities can easily make a more significant contribution to the listening than the actual op-amp. This I believe as well, and I have tried, like Cycline3, to find the "best" choice ... unsuccessfully, IMOP. I would defy anyone to determine the listening differences between TI's ultra low noise precision op-amp and the comparable one from Analog, Linear or which ever ... and hang the price. Opinions may continue to vary here, but they become like the question of which is better, the Ford or the Chevy.

The Q&A really is usability: If the particular op-amp is not available, as I usually desire, in the 8-pin mini-DIP package, use the comparable makers version. Note that those multi legged, mil-spec "critters in the can" can be placed into a guide socket that turns them into DIPs for easy DUT experiments.

Short story: build your circuit with a socket so's you can try out several different op-amps and see for yourself ... 'cause the slight variations in capacitor manufacturers can easily make more auditory differences than the op-amp variations. :smash:

A crudely built, 2 in, 1 out, mixer + low pass filter for 1 channel subwoofer ...
 

Attachments

  • jusfoolin3.jpg
    jusfoolin3.jpg
    82.3 KB · Views: 872
Hi,

"man, back in 1986 I built a MM/MC pre like the one I mentioned before (http://www.matuschek.net/platina, I assisted in designing this one too) and it just sounded great. I haven´t stopped using it. ;) It´s extremely flexible and virtually noise- and distorton-free, even with low output MCs and it sounds just great in every aspect"

There is basically no replacement part for the input device apart from the newer SMD-Version. You just can tweak with different OPamps in the second stage. But I tell You what...different PCB-Layouts, different power supplies and the decoupling and especially the precision of the equalization curve were the most sound influencing parameters.
Wether a OPA2604, a 2134, a AD712, AD8020 -all JFET-OPs- was used.....not much difference. Changing one resistor or cap of the eq-network and the sonics differered comparatively remarkable.

jauu
Calvin
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.