Not at all, know what real blind tests are.
Not knowing what is in there, and not seeing any data. Actually blind.
Because data is no different than story telling. Well, obviously its different.
It just creates bias, and the outcome always changes.
In reality the sound isn't much different or even different at all.
I dont have to lie to myself and have a rather large collection of opamps.
I know what they sound like, and have heard Audiophile and Musician yaba yaba for 30 years or more now.
Been in hundreds of recording studios and know very elaborate signal chains. Full of 4558 or TLO7x or 5532/4
Yes. Its about about how these devices are implemented and the supporting circuitry. Decent capacitors, attention to input offset, power supply filtering, etc. While alot of pro gear has these op amps, they usually use the better ones in critical areas requiring more precision and lower distortion.
The NE5523/34 is a great op amp family. The 1458 stuff is the lowest spec you'll find in a signal path on a pro console. They're also supported by decent power supplies and proper grounding schemes.
I used to repair and modify alot of vintage Neve, Soundcraft, Midas, PPI, SSL, etc. They do use alot of NE55xx and TL0xx in specific places. They also use some of the best coupling and input transformers, discrete inductors, etc.
The consoles are designed to have a specific "flavor" of sound to enhance the music. They aren't designed to be perfect fidelity, laboratory instruments. You want the coloration in specific places, especially in the gain and EQ.
In CD players and other playback devices, you want low coloration and distortion, so the music comes out (in theory) the same as it was recorded. That is a harder thing to achieve, especially if your building to a budget, especially on a high gain MC phono stage.
There is less concern or focus on cost in the case of a high end Neve or SSL console, where the price tag reflects it and some flavor is desired. If the console didn't have any personality, it wouldn't make a difference which console you used to record with. In that case you'd choose the one with the most precise, lowest distortion gain and eq stages. The biggest concern here is headroom before clipping. This is the magic of an analog console.
Most music is recorded in a studio, where its given the character it needs as an art form to draw attention to itself and pull you in as a listener. For a high end playback device, its the accuracy, resolution and fidelity which matter. You dont want much of any flavor in this application.
The NE5523/34 is a great op amp family. The 1458 stuff is the lowest spec you'll find in a signal path on a pro console. They're also supported by decent power supplies and proper grounding schemes.
I used to repair and modify alot of vintage Neve, Soundcraft, Midas, PPI, SSL, etc. They do use alot of NE55xx and TL0xx in specific places. They also use some of the best coupling and input transformers, discrete inductors, etc.
The consoles are designed to have a specific "flavor" of sound to enhance the music. They aren't designed to be perfect fidelity, laboratory instruments. You want the coloration in specific places, especially in the gain and EQ.
In CD players and other playback devices, you want low coloration and distortion, so the music comes out (in theory) the same as it was recorded. That is a harder thing to achieve, especially if your building to a budget, especially on a high gain MC phono stage.
There is less concern or focus on cost in the case of a high end Neve or SSL console, where the price tag reflects it and some flavor is desired. If the console didn't have any personality, it wouldn't make a difference which console you used to record with. In that case you'd choose the one with the most precise, lowest distortion gain and eq stages. The biggest concern here is headroom before clipping. This is the magic of an analog console.
Most music is recorded in a studio, where its given the character it needs as an art form to draw attention to itself and pull you in as a listener. For a high end playback device, its the accuracy, resolution and fidelity which matter. You dont want much of any flavor in this application.
And after all is said and done, 95% of the op-amps are set to bypass, and the purist drums and husky vocal recorded with 2 mics carefully suspended on booms are sent to the A 2 D. The vocalist wanted to sound more like "The Animals", so the entire vocal track is passed through an overdrive VST giving the loud bits 30% distortion or more.😀
You will only have 0.6V across the inputs in clipping and slewing, so already distorting. Their purpose is to avoid reverse breakdown of the input transistors spoiling the noise figure permanently.
Back in 1992, Gary Galo wrote that the NE5534(5532) could not be used in a CD player or DAC that had high-end performance.
It's true that I used that NE5532 from Signetics in my first DACs about fifteen years ago. It was better than the classic TI 5532, but for me, nothing special for the I/V stage. After that I switched to tubes and CFA op amps.
It's true that I used that NE5532 from Signetics in my first DACs about fifteen years ago. It was better than the classic TI 5532, but for me, nothing special for the I/V stage. After that I switched to tubes and CFA op amps.
Gary Galo stood in 1992, arms crossed, declaring the NE5534 (and its sibling, the 5532) unworthy of high-end CD players and DACs. But Signetics had a secret: their NE5534 had already infiltrated the Cambridge CD3—a player known for its chef’s-kiss sound. This wasn’t just a “meh” op-amp; it was the quiet hero, balancing precision and musicality like a tightrope walker in a tuxedo, until Lukasz Fikus appeared and transformed it into a LampizatOr. 😛
It's not exactly true that Gary Galo sat with his arms crossed. He replaced the NE5534 in the Philips DAC960. I guess you read about it, The Audio Amateur 2/92 😉 .
Of course, I read it. But reading theory and listening are not the same. Recently, I've been listening to a friend's Cambridge Audio CD3 as a sound source compared to a turntable.
(The CD3 features 16-bit x 16-times oversampling, a Philips CDM 1 MK II transport, four Philips TDA 1541 A DACs, and eight Signetics NE5534. It's built like a battleship.)
The setup includes:
VPI/SME M2 - 9/Grace F9/E
Sindric Pointe Analogia MK V - phono MM
Bent Audio TVC Preamp Model 102M: A passive preamp with two Stevens & Billington transformers.
Softone Model 7 Amplifier: A tube amplifier with four KT88 tubes.
Magna Acoustica Speakers: Full-range cabinets with two custom 10" Ciare PA drivers.
Whether playing the turntable or the CD player, the difference ultimately comes down to personal taste😉
(The CD3 features 16-bit x 16-times oversampling, a Philips CDM 1 MK II transport, four Philips TDA 1541 A DACs, and eight Signetics NE5534. It's built like a battleship.)
The setup includes:
VPI/SME M2 - 9/Grace F9/E
Sindric Pointe Analogia MK V - phono MM
Bent Audio TVC Preamp Model 102M: A passive preamp with two Stevens & Billington transformers.
Softone Model 7 Amplifier: A tube amplifier with four KT88 tubes.
Magna Acoustica Speakers: Full-range cabinets with two custom 10" Ciare PA drivers.
Whether playing the turntable or the CD player, the difference ultimately comes down to personal taste😉
Last edited:
This is true in theory, however it will depend on the specific characteristics of these diodes. The junction saturation voltage is somewhat gradual and will always, at a minimum, have a partial effect on the signal. The in-circuit configuration of the op amp will determine how much actual distortion the diodes generate. The circuit's input impedance is obviously defined by the effective resistor values to ground and neg feedback loop. Bipolar op amps perform best in low impedance applications, which makes the diodes at the input less of an influence but that doesn't completely discount their influence.
Its true in practice as the difference in voltage between the inputs in normal operation is a few mV at most and varies even less (a few microvolts typically). No Si pn junction does anything at these low levels but behave like a small linear capacitance. And its a differential capacitance at that. You can safely assume the noise protection diodes do not degrade circuit function in any measurable way other than the capacitance increase. They are essential to the guaranteed noise performance of the devices (which is very measurable).
Linear Technology opamp datasheets carry a warning: these input protection diodes interfere with slewing (as mentioned by @davidsrsb in post 164 above).
_
_
Attachments
If you are massively slew-limited like this, you are already heavily distorting - this is nothing to do with linear audio use.
Slewing can be a component of settling time, which is an important opamp spec in a D-to-A application.
Interesting insights.
Rolling Op-Amps in and out of your test circuit got my attention. The listening tests sounded like a good Idea especially with Friend #1 doing some basic measurements before you started. I am sorry to see that the measurements never happened.
Op-Amp rolling is always good sport that nearly always deteriorates into arguments that will never be settled with gold plated tin ears.
Go ahead have fun. It is not my intention to dissuade anyone.
Lately I solder a OPA1656 directly to the PCB with the concept that the discrete output devices will dominate the output distortion signature. As in the plot of the WHAMMY that you posted.
Thanks DT
All opamps have slew rate limits. A well designed circuit should not let this happen by filtering. Hitting an opamp with picosecond rise time steps can only cause problems.
I saw this in an early Sony CD player that used 741s of all things, very obvious and audible slewing in the filter.
Thanks for dropping a line. I ended up moving to a new home almost immediately after this post was made, and my DIY work has been more oriented toward framing and drains for the last few years. I do have some basic test equipment that I intend to properly set up - perhaps I will double back and provide datasets.
I realize that this forum leans heavily toward data orientation and earned expertise - I am grateful for folks' patience with me here.
Gary Galo wrote a lot of stuff in Audio Amateur magazine, that won't stand up to 2025 technical inspection or A/BX testing.
I like Lf353 but i got a bunch of RC 4580 and they are really good. 5v/ us and more stable then 5532. Very low noise as well
There is less then a mV between the inputs in normal operation. The diodes play no part in audio distortion.
Except all the ones that have low slew rates you mean? Plenty of very slow micropower opamps out there...