I like the AD-8610 Over the AD-825 and OPA-627, But the 8610 is supply voltage limited to +/- 13 Volts rather than +/- 15 Volt supply rails of the other two Devices. However the AD-8610 swings to within 1.2 Volts from the Rails.
Sound quality of the AD-8610 and the Dual AD-8620 are amoung the Finest available from an opamp. None of the undesireable qualities that some atribute to Opamps are nott presentt with this device. Smoothness defines the sound with Imaging and resolving of detail first rate. none of the sibelense of the AD-825 and to a lesser extent the OPA-627. also the 8610 has full bodied Bass, another quality that all but the best Op Amps lack. IMHO the AD-8610 and the AD-843 are the Finest Op Amps for Audio you can get whitch one depends upon what kind of sound quality desired and what application the device is used. some circuits like simulated inductors are best with a High speed Video grade Op Amp. so are some active filters. but for general line level Gain the 8610 and 843 are my top choices. if a dual is required the AD-8620 is a Dual AD-8610. the Dual AD-8620 is good for equipment upgrades that used low performance duals like the 1458, 4558 and similar. Thay work wonders in I/V convertors.
RIAA and Mic gain stages are best with Ultra low noise BJT input devices like the LT-1028 and the AD-797
Sound quality of the AD-8610 and the Dual AD-8620 are amoung the Finest available from an opamp. None of the undesireable qualities that some atribute to Opamps are nott presentt with this device. Smoothness defines the sound with Imaging and resolving of detail first rate. none of the sibelense of the AD-825 and to a lesser extent the OPA-627. also the 8610 has full bodied Bass, another quality that all but the best Op Amps lack. IMHO the AD-8610 and the AD-843 are the Finest Op Amps for Audio you can get whitch one depends upon what kind of sound quality desired and what application the device is used. some circuits like simulated inductors are best with a High speed Video grade Op Amp. so are some active filters. but for general line level Gain the 8610 and 843 are my top choices. if a dual is required the AD-8620 is a Dual AD-8610. the Dual AD-8620 is good for equipment upgrades that used low performance duals like the 1458, 4558 and similar. Thay work wonders in I/V convertors.
RIAA and Mic gain stages are best with Ultra low noise BJT input devices like the LT-1028 and the AD-797
Op-amps for audio
Has anyone ever compared these two devices sonically - I've never had good results with the AD797, finding it soft, woolly and generally lacking in musicality.
I'd agree with ppl above, the AD86x0 are the best op-amps I've used, bar none in high level applications.
I also like the AD825, but it can be sibilant, with a bright sound, despite some very promising attributes for audio (wide open-loop bandwidth).
Andy.
Has anyone ever compared these two devices sonically - I've never had good results with the AD797, finding it soft, woolly and generally lacking in musicality.
I'd agree with ppl above, the AD86x0 are the best op-amps I've used, bar none in high level applications.
I also like the AD825, but it can be sibilant, with a bright sound, despite some very promising attributes for audio (wide open-loop bandwidth).
Andy.
AWL> I tend to agree with you on the AD-825 it is quite sibilant, to the Point of being irritating. the AD-797 is also somewhat sibilant. I don't What the Term Woolly is, But i liked the LT-1028 In my last Mag Phono Circuit Better than the AD-797. This was a standard jfet Follower driving each of the Op-Amp inputs. this circuit was to present a low impedance load to the Opamps for lowest noise while the Low input Bias currents of the Jfets allow for DC coupling. This Circuit was real quiet in real world testing like -90 dB below 500 mV. Hear the LT-1028 sounded alot more detailed and articulate than the AD-797. Both were quiet. The LT-1028 was in the Cer Dip Package and the AD-797 in the P-Dip If that matters?
Thanks ppl
By woolly, I mean soft sounding and lacking in dynamics - a bit 'muffled' I suppose.
I've never tried the LT1028, so I will secure some samples and have a play.
One thing I would say about the AD825, more than any other op-amp I've used, it seems t benefit from significant warm-up times.
No idea why, but cold they are horrible, but once warmed up (and this can seem to take hours, to days!) they can be very good. I think careful application and tailoring of circuitry can yield superb results, but it's not quite so 'drop-in' a replacement.
There's some intersting TI devices coming out now that may be worthy of investigation, some offering very high speed and extremely low noise. All SMD though, and tiny packages.
Andy.
By woolly, I mean soft sounding and lacking in dynamics - a bit 'muffled' I suppose.
I've never tried the LT1028, so I will secure some samples and have a play.
One thing I would say about the AD825, more than any other op-amp I've used, it seems t benefit from significant warm-up times.
No idea why, but cold they are horrible, but once warmed up (and this can seem to take hours, to days!) they can be very good. I think careful application and tailoring of circuitry can yield superb results, but it's not quite so 'drop-in' a replacement.
There's some intersting TI devices coming out now that may be worthy of investigation, some offering very high speed and extremely low noise. All SMD though, and tiny packages.
Andy.
ALW> thanks for the tip on TI devices any part Numbers? Yes the AD-825 is ultra critical of the application and will require a long warmup like you say at least hours. Also a Current source is a great help with this and Other Op-Amps thanks for the Termanology inlightenment, yes the AD-797 is Kind of Bland sounding But i gess that is what attracts so manny people to it. Me i like the LT-1028 alot Better and i am not that much of an LTC fan. Except for the LT-1122,LT-1010,LT1085,LT-1033,LT-1210 and for High gains the LT-1028 and maybe the LT-1115.
.
What I menat was after pluggin it in it sounds very bad, that is when it is cold.
After it gets warm it sounds veeeery good. It takes about 30-60 minutes or so.
Haven´t used it for some time now because I have a SCD-XB940 with LClock and Zapfilter which is sooo nice, especially with SACD.
Before I forgett about the Zapdac completely, I will try to improve upon it with on board LClcok and sync wire to transport.
Also do something with the analog section. Either use a Zapfilter, or try to do something "innovative" such as feeding the 1702K into a Jensen transformer to get a balanced signal, then use BUF634 with batteries on the output. I don´t know if it would work with a transformer though, instead of the AD811 which is the I/V stage now.
/Peter
After it gets warm it sounds veeeery good. It takes about 30-60 minutes or so.
Haven´t used it for some time now because I have a SCD-XB940 with LClock and Zapfilter which is sooo nice, especially with SACD.
Before I forgett about the Zapdac completely, I will try to improve upon it with on board LClcok and sync wire to transport.
Also do something with the analog section. Either use a Zapfilter, or try to do something "innovative" such as feeding the 1702K into a Jensen transformer to get a balanced signal, then use BUF634 with batteries on the output. I don´t know if it would work with a transformer though, instead of the AD811 which is the I/V stage now.
/Peter
ppl: What was the general range of resistance that you use to buffer the output capacitance of the current source/sink?
BTW, note that if you use current-regulator diodes, in practice the amount of current will change as the voltage across the CR diodes varies. With Zener diodes also, the voltage output will change according to the amount of current. If possible, I recommend incorporating measures to prevent the voltage across CR diodes and the current through Zener diodes from changing.
Re the AD797, I have also tried it, but must admit that I don't find much wrong with it, other than noting that it needs good bypassing and prefers a guard ring on the inverting input node.
Currently I am working on a prototype for a commercial phono stage with a 797 in the front end, and it sounds surprisingly good. Good bass and rhythms, good ability to decipher complex harmonies, good dynamics, quite relaxed and stays out of the way of the music. Granted, I have put in some interesting tricks (and there is still more that I want to do), but this circuit is definitely opamp-based.
I will be bringing a prototype to the Las Vegas, so anyone who visits the Immedia/Scantech room at "The Show" can check for themselves what it (Lyra Erodion) sounds like.
I don't think that opamps are the be-all and end-all of analog amplification, but I suspect that in many cases the perceived problems with opamp circuitry have much more to do with implementation issues than the opamp chips themselves.
regards, jonathan carr
BTW, note that if you use current-regulator diodes, in practice the amount of current will change as the voltage across the CR diodes varies. With Zener diodes also, the voltage output will change according to the amount of current. If possible, I recommend incorporating measures to prevent the voltage across CR diodes and the current through Zener diodes from changing.
Re the AD797, I have also tried it, but must admit that I don't find much wrong with it, other than noting that it needs good bypassing and prefers a guard ring on the inverting input node.
Currently I am working on a prototype for a commercial phono stage with a 797 in the front end, and it sounds surprisingly good. Good bass and rhythms, good ability to decipher complex harmonies, good dynamics, quite relaxed and stays out of the way of the music. Granted, I have put in some interesting tricks (and there is still more that I want to do), but this circuit is definitely opamp-based.
I will be bringing a prototype to the Las Vegas, so anyone who visits the Immedia/Scantech room at "The Show" can check for themselves what it (Lyra Erodion) sounds like.
I don't think that opamps are the be-all and end-all of analog amplification, but I suspect that in many cases the perceived problems with opamp circuitry have much more to do with implementation issues than the opamp chips themselves.
regards, jonathan carr
Jonathan
Thanks for that post, my experience of the AD797 was as a drop-in listening test to a simple headphone amplifier.
I use this as in general it's a highly revealing and simple way to evaluate sonics of different devices, but I'd freely admit it's optimized for different devices (JFET i/p) and layout would be different were I to do it now (I built it 7 years ago, and still use it daily).
Will re-visit the AD797, as I fancy doing a simple phono stage myself and the noise figure of the AD797 is tempting.
I'll reserve further judgement until I've looked in some more detail.
Andy.
Thanks for that post, my experience of the AD797 was as a drop-in listening test to a simple headphone amplifier.
I use this as in general it's a highly revealing and simple way to evaluate sonics of different devices, but I'd freely admit it's optimized for different devices (JFET i/p) and layout would be different were I to do it now (I built it 7 years ago, and still use it daily).
Will re-visit the AD797, as I fancy doing a simple phono stage myself and the noise figure of the AD797 is tempting.
I'll reserve further judgement until I've looked in some more detail.
Andy.
Andy:
Note that I am using the AD797 as the input stage of a two-section design. For applications that require lots of voltage swing into lowish impedances, I believe that the TI/BB OPA551 or OPA552 may be more appropriate than the AD797. Or perhaps you could consider adding a unity-gain buffer chip after the 797.
regards, jonathan carr
Note that I am using the AD797 as the input stage of a two-section design. For applications that require lots of voltage swing into lowish impedances, I believe that the TI/BB OPA551 or OPA552 may be more appropriate than the AD797. Or perhaps you could consider adding a unity-gain buffer chip after the 797.
regards, jonathan carr
Andy:
Passive RIAA is better for overload margin, and tends to do a better job of maintaining the proper RIAA response at all frequencies (active commonly results in a rising HF response, and although this can be corrected via another network, there will be some sonic penalties). Active RIAA has better sn ratios, but the complex feedback load may on occasion require the use of phase compensation to avoid ringing or instability.
In my experience, active can sound at least as good as passive when implemented without phase compensation and with careful design techniques, but passive is the easiest to get good sound from (usually requires less design know-how).
Hybrid is also good, but because passive and active tend to have different "sounds", unless the RIAA curve is spot-on, the circuit can sound somewhat discontinuous. Passive or active only seems more tolerant of RIAA error than hybrid, and passive (subjectively) seems the most tolerant of all.
hth, jonathan carr
Passive RIAA is better for overload margin, and tends to do a better job of maintaining the proper RIAA response at all frequencies (active commonly results in a rising HF response, and although this can be corrected via another network, there will be some sonic penalties). Active RIAA has better sn ratios, but the complex feedback load may on occasion require the use of phase compensation to avoid ringing or instability.
In my experience, active can sound at least as good as passive when implemented without phase compensation and with careful design techniques, but passive is the easiest to get good sound from (usually requires less design know-how).
Hybrid is also good, but because passive and active tend to have different "sounds", unless the RIAA curve is spot-on, the circuit can sound somewhat discontinuous. Passive or active only seems more tolerant of RIAA error than hybrid, and passive (subjectively) seems the most tolerant of all.
hth, jonathan carr
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.