Someone expressing his subjective listening impression doesn't mean it's actually the case, unless he posts measurements along with it as supporting evidence.
Indeed, when I was young I had a Yamaha amp with extreme low THD, then someone demonstrated a Mission amp (it could have been a Creek, Musical Fidelity, Nytech or a tube amp). The Mission amp was very ugly but it destroyed my Yamaha so easily... The Yamaha did sound totally awful and unbearable in comparison. That was the shock of the century for me...
Last edited:
Very stupid argument, really!
I'm not your butler get your lazy *** up and look at the TI datasheets. Jesus...
Did anyone ever correlate a measurement to explain why they sounded bad?
HOW slow?....in terms of V/uSec slew rate.......
More V/uSec does not mean automatically that it sounds better, but there is a minimum V/uSec for SQ. If you get below the minimum you will hear the SQ dropping.
Last edited:
That would depend on the bandwidth required and voltage out (sine ) .
If you limit it to 20kHz and 10V out that would be a bit over 1V/uS.
"Slewing " isn't nice to listen to as it can introduce distortion in the waveform.
If you limit it to 20kHz and 10V out that would be a bit over 1V/uS.
"Slewing " isn't nice to listen to as it can introduce distortion in the waveform.
I think that's a bit on the slow side. Slow opamps are definitely inferior audibly, but they can slew at 1-2 V/uSec. IME, slew rate improvements of up to 10 V/uSec are audible, but not beyond.
I'll look at it if you know of one amp using OPA1612 sounding audibly better than another amp using cheaper op amp in objective listening comparison.
This is a direct quote from D.Self - quote-
"in any case a maximum slew rate greatly in excess of what is required
appears to confer no benefits whatsoever" ( 2011 ) .
Nothing changes in life the same arguments in the 80,s/90.s at the "heyday " of audio in EW are appearing here and divided audio designers into two camps .
Those were the D.Self proponents and the JLH proponents it certainly was interesting times then and very heated views were exchanged .
EW is pretty staid now so I stopped buying it.
Every month I looked forward to the letters page ,the writers trying to "nuke " each other .
"in any case a maximum slew rate greatly in excess of what is required
appears to confer no benefits whatsoever" ( 2011 ) .
Nothing changes in life the same arguments in the 80,s/90.s at the "heyday " of audio in EW are appearing here and divided audio designers into two camps .
Those were the D.Self proponents and the JLH proponents it certainly was interesting times then and very heated views were exchanged .
EW is pretty staid now so I stopped buying it.
Every month I looked forward to the letters page ,the writers trying to "nuke " each other .
Calculate required slew rate as
SR(signal) = Vamplitude*2*pi*frequency.
Most +/-15V op amps have max Vamplitude = 14V. The max SR(signal) = 1.26 V/us at 20 kHz. Some designers say the amp SR should be at least 5x the signal SR, others say 10x. So, a slew rate of 9 V/us is probably the minimum SR. Beyond that, spend your money on lower noise and distortion.
SR(signal) = Vamplitude*2*pi*frequency.
Most +/-15V op amps have max Vamplitude = 14V. The max SR(signal) = 1.26 V/us at 20 kHz. Some designers say the amp SR should be at least 5x the signal SR, others say 10x. So, a slew rate of 9 V/us is probably the minimum SR. Beyond that, spend your money on lower noise and distortion.
I think so too, good opamps start at 4 V/uSec better opamps start at approx. 8-10 V/uSec. A 40 V/uSec opamp can sound much worse than a 10 V/uSec opamp.
Last edited:
I think there are also other factors. I don't know if the more modern opamps do this, but I remember that one thing that made the LM318 sound less than ideal was that its slew rate was ASYMMETRICAL----it was about 30% faster going up than down.
Oops, I had a typo. That should be 1.76 V/us. The rest of the numbers are correct.
Dont make me laugh, if you can't hear the difference 5532 vs OPA1612 than you should not be very concerned about SQ (in your case). Even AD797 or OP627 cant hold a candle vs OPA1612 in terms of SQ - the difference is HUGE!
My preamp sounds best with OPA1612 for gain and OPA2192 as buffer. And this preamp sounds MUCH better than Audio Research, Burmester and Accuphase preamps costing above €10.000. Heard it all here with different LS and poweramps. Despite my opinion even the owners of the mentioned preamps have clearly stated that. BTW, the Audio Research is sold because of that! My own Nova preamp was a Stereophile (the US magazine) reference, sounds more or less terrible in comparison. My opamp preamp has a mediocre PSU, good quality parts but standard regulators, imagine the SQ with Jung regs...
Last edited:
I agree about the OPA1612 for gain. I would tend to use the OPA1642 for buffering, filtering and EQ. JFET inputs let you work with a larger range of input resistors.
The differences you've heard, was it from level matched double blind listening?
Do you also hear a difference in your preamp with different power cords?
It's not a stretch to be able to hear a difference in a 5532 and an OPA1612, depending on the gain, and how many are put together in a row.
I have a fully parametric EQ design that I would like to build using OPA1642 op amps, because even with about 6 of them cascaded, I expect their noise and distortion to still be inaudible. I last tried the circuit with OPA2134 in year 2000, and with TL074 in 1992. That was the circuit where the OPA2134 sounded better. Double blind AB, the TL074 circuit had audible hiss from the speaker, and the 2134 circuit did not. Double-blind AB we also preferred the OPA2134 at the listening position, especially with the EQ in heavy use.
By "we" I mean the design and marketing staff at Stillwater Designs, where I used to work, 20 years ago. I left the PCB at SD, so I'll have to build a new one. At least this one won't have to be powered from a 12V car battery.
I have a fully parametric EQ design that I would like to build using OPA1642 op amps, because even with about 6 of them cascaded, I expect their noise and distortion to still be inaudible. I last tried the circuit with OPA2134 in year 2000, and with TL074 in 1992. That was the circuit where the OPA2134 sounded better. Double blind AB, the TL074 circuit had audible hiss from the speaker, and the 2134 circuit did not. Double-blind AB we also preferred the OPA2134 at the listening position, especially with the EQ in heavy use.
By "we" I mean the design and marketing staff at Stillwater Designs, where I used to work, 20 years ago. I left the PCB at SD, so I'll have to build a new one. At least this one won't have to be powered from a 12V car battery.