The common-mode distortion of 5534/5532 is documented (this is one of the properties of opamps that Doug Self has researched and provided in several of his books on analog circuit design - basically non-inverting topology shows higher distortion than inverting at high signal levels and frequencies, and especially with high source impedances).
If you use an inverting topology to avoid common-mode voltage swings then your stage is necessarily low input impedance (compared to a bare opamp input pin) as the non-inverting input pin is at virtual ground and usually there's a resistor of a few k loading the input into that virtual ground.
If you use an inverting topology to avoid common-mode voltage swings then your stage is necessarily low input impedance (compared to a bare opamp input pin) as the non-inverting input pin is at virtual ground and usually there's a resistor of a few k loading the input into that virtual ground.
Last edited:
The Harrison was indeed their first console(not sure of its true series though...very confusing intel from google on that), but it is said that the console was modded by Leif Mases...We'll never probably know what he did ...Maybe he replaced all op-amps with ne5532 
Can you share a schematic for 5632 or it's just having more channels than 4032?
D Self measured cm distortions on ne5534 at unity gain if i remember right as he recommended some circuits to counter that.
Can you share a schematic for 5632 or it's just having more channels than 4032?
D Self measured cm distortions on ne5534 at unity gain if i remember right as he recommended some circuits to counter that.
Thanks Mark, good info. With regard to Doug Self's research, is that specific to 5534/5532 or op-amp topology in general?
Apologies to the op for the tangent.
Lower common mode distortion, and lower input impedance, apply to all op amps when in the inverting topology,
compared to non-inverting.
I like the the sound of EQ built around TL072, but I find the consoles with full of TL072/074 sound a little murky (slow) in general. Middle class consoles with full of TL07x from 80s-90's such as SoundCraft, SoundTracs, Neotek, all share the similar sound signature. I'm sure that some consoles such as Neotek and Trident are still considered to be very desirable by many for their warm analog sound. The consoles built around 5534 usually sound cleaner and punchier. It's not uncommon that the consoles with the mixture of 07x and 553x, and 07x are usually found at EQ parts.
V/μs has no effect on the 'speed' of the sound, I agree with Markw4 that the term 'slow' used by the plasnu was used more as a sound signature for the TL07...
Once I did an I/V conversion with the AD811 which has 2500V/μs. How would you describe the sound then?
Ok...is it about settlement time then? Is it measurable? 14 years ago i made some Mesa boogie tube guitar preamp clones and I remember that the most resolute and "speedy" one if i'm allowed to use simillar terms was one named Triaxis which had 5 ecc83 tubes , user friendly tone eq followed by an internal 5 band fixed eq amplified with tl072 op amps stage ironically called "dynamic voicing "and a tl072 line out buffer at the end .I've seen all anti-high end techniques used there like using tantalum electrolitics and I can't remember a more reveered preamp at the time than that one.Dream Theater used it everywhere( hence my nickname...).Anyone in the cloning bussiness knew that without that 5 band fixed eq that preamp's sound was all over the place and that was the real main ingredient of its character . Do you know anyone more subjectively sensitive to "dynamics" than a tube guitar player?I'm talking about John Petrucci tastes here folks! There's not a single famous tune where he didn't use Triaxis in his chain and he's probably one of those guys who used at least 5 different boxes in series to get a tone. When i hear the word "speed" in anyonetalk about audio i suddenly remrmber my 120v/us kenwood amp whose sound couldn't be more boring and "slow" ...Most of the time mid-high frequency distortions give the impression of huge dynamics and speed unlike really fast amplifiers who's sound are usually found in a very clear and uninvolved register.
Besides any tl072 ever used can be replaced by opa2134 or 2132 ...try swap them and hear the difference cause their parameters are really different especially output impedance and you'll quickly realise that the impedance of the next stage plays the most important role in how a tl072 sounds.
Slew rate and settlement time is always about slewing a certain current into a certain load.If you overload the output of an op-amp there's no slew rate while you get absolute best settlement time...as your output signal is quickly " settled " to... nothing.
Besides any tl072 ever used can be replaced by opa2134 or 2132 ...try swap them and hear the difference cause their parameters are really different especially output impedance and you'll quickly realise that the impedance of the next stage plays the most important role in how a tl072 sounds.
Slew rate and settlement time is always about slewing a certain current into a certain load.If you overload the output of an op-amp there's no slew rate while you get absolute best settlement time...as your output signal is quickly " settled " to... nothing.
I can't seem to find a link, but there's a discussion (probably 'Gearspace') featuring at last 2 engineers who worked for Neve in the 70s. As some of you might know, Rupert Neve worked on the predecessor to the NE5534 as a consultant, and was one - if not the first - of the big console designers to implement it. The story goes that once RN could get his hands on the 5534, the BBC were first in the queue to have some of the discrete modules in their consoles changed out, on account of there being a particular BBC 600R broadcast requirement that the discrete modules couldn't quite attain. According to the Neve engineers, the BBC engineers came back remarking how excellent the modified NE5534 consoles sounded If you find the right search string, you'll find the thread.
TL072s have poor overload recovery and are hissy as hell. If you think they sound 'murky', it's probably because of the blanket of hiss they leave everywhere. They are great outside of the signal path as drivers for metering or simple DC duties etc. I'd hate to see them discontinued.
If you find the right search string, you'll find the thread. TL072s have poor overload recovery and are hissy as hell. If you think they sound 'murky', it's probably because of the blanket of hiss they leave everywhere. They are great outside of the signal path as drivers for metering or simple DC duties etc. I'd hate to see them discontinued.
Some decades ago I was shown a paper schematic of a Neve opamp-based console. IIRC the opamps were biased into class A with an offset voltage, and the audio path was capacitor coupled using electrolytic caps (apparently because of the amplifier offset). Never had a chance to listen to one.
The discrete single-ended Neve circuits use gapped transformers, precisely because of the massive DC offset. The transformers also step up, in order to get decent headroom off a single +24VDC supply. There are many different eras of Neve circuitry, and Rupert himself left the firm in the early 80s. They also used different transformers, not just the famous Marinair. I only mentioned the story in the previous post as I was amused to see more than one ex-engineer corroborate the sentiment that the Beeb were happy to go IC. I think I read that Neve didn't let on that they'd fitted ICs, and the BBC were going purely on their own impressions.
I know guys that 10 years ago were able to pay real money to clone the 1073 and 1081 discrete preamps although they already had the more modern op amp based Neves and ssl's...Two years ago a guy having a recording studio in Germany asked me if i can diy a few 1073 channel for 1000 $ each ...I had to decline as I wasn't sure how I'd make a suitable output transformer with a simillar sound to the original one.
Not really sure what you're talking about honestly.The descreete versions are way more reveered by the recording engineers than the new ones.
Not really sure what you're talking about honestly.The descreete versions are way more reveered by the recording engineers than the new ones.
Last edited:
From the 8078 onwards (possibly the most revered Neve of all), RN started using the 5534. The Montserrat consoles were chock full of 5534s. George Massenburg has stated that his favourite Neve is the 8078. Pretty much every Chic record was recorded on an 8078. If you find the discussion I referred to, you'll see that the Neve engineers themselves are bemused by the religion surrounding the fully discrete consoles. They saw proof for themselves that the likes of the BBC were just as happy to use NE5534 driver stages. They know fully well that 'good sound' is about overall design, not simply discrete BJTs vs IC.
The Jensen 990 opamp uses the LM394. If discrete is somehow superior, how do you explain that?
The great Arif Mardin was asked in his last interview (SOS magazine - easy to find) whether he preferred a Neve or SSL. He replied that he was agnostic and would use either, provided it was well maintained and of pro quality.
No A/B/X test shows that discrete has a quality that monolithic doesn't have, only folklore and Ebay prices. Analogue is analogue
I don't really get it:
they're not only using discretes, but also the 2n3055 as driver for the output transformers in both new spx and dpx versions of 1073...
https://www.google.com/amp/s/www.soundonsound.com/reviews/neve-1073-spx?ampThis is a 5000 $ reissue using 2n3055...
https://www.google.com/amp/s/www.soundonsound.com/reviews/neve-1073-dpx?ampAnd some history:
https://www.proaudiodesign.com/blogs/news/a-brief-history-of-the-legendary-neve-1073https://www.sweetwater.com/insync/neve-1073-history-and-legacy/
they're not only using discretes, but also the 2n3055 as driver for the output transformers in both new spx and dpx versions of 1073...
https://www.google.com/amp/s/www.soundonsound.com/reviews/neve-1073-spx?ampThis is a 5000 $ reissue using 2n3055...
https://www.google.com/amp/s/www.soundonsound.com/reviews/neve-1073-dpx?ampAnd some history:
https://www.proaudiodesign.com/blogs/news/a-brief-history-of-the-legendary-neve-1073https://www.sweetwater.com/insync/neve-1073-history-and-legacy/
Last edited:
The above is written by Geoff Tanner, who worked at Neve during the time when the mythical discrete desks were released. You can read the thread here: https://gearspace.com/board/high-end/346229-neve-8068-8078-a.html
Neve employed a lot of engineers in the 70s and plenty of them are still around. I think you will find that they are quite academic and back Mr Tanner up.
NB - the drivers that use the 2N3055 are very much Class A. They definitely have a sound...because of the transformer.
Hi,
I think there is a misunderstanding:
Rupert Neve owned his Name as a brand until Ams bought it after the company had financial issues. Iirc this happened end ( or mid) 70's. Design were discrete based at that time with transformers in/out and interstage. Those are the Vintage 80XX. 1073,1081,....
From there all said Neve's products ( which should be named Ams/Neve) had only the name on it, Rupert Neve never been involved in the designs.
The next generation of Ams/Neve then used 553x. and mic transfo input (Belclere iirc) By design they are of same gneration than SSL ( but with significant differences in circuits and philosophy / routing is way more complex and elaborate on Ams/Neve than on SSL, automation (Flying Faders)is less intuitive than SSL, so you ussually seen Neve in A for recording/tracking, SSL in B for mixing). Those where the V series (Vr/Vx), SSL concurent was 4000E/G series. The next move was then 88r which was the concurent of SSL9000j/k/Xl.
Meanwhile, Rupert Neve designed the Focusrite console from which all Focusrite range ( Blue and Red) are the children ( they used 553x and same output transfo ( which was very different from the vintage Neve, from what i've seen those are Carnhill transformers, hence the misconception Carnhill did the Vintage one ( who were Marinair). Input transfo are Lundahl 1538.
He designed for Amek too during this period. At the end of 90's to offer an alternative design to 88r/9000 Amek hired him to design the 9098i. Mc33078, 553x and That vca used in signal path as well as the same output transformers than Focusrite's one ( but used differently).
It was the last 'big' inline analogue recording/mixing desk he designed.
Early 2000 he launched his brand RND, with portico range of modules and custom designed consoles ( split for record/mixing and some mastering dedicated one's).
Montserrat console was named. Those were owned by Martin yes but business being business Ams/Neve had a 88r bought by Air studio. The name stay the same, the technology and sound is not. Don't get me wrong, i love the 88r ( as i loved the 9000, 4000 and Vr/Vx ) but the sound as nothing to do with vintage Neve ( despite their was 1073/1081 modules availlable for it ( those where 'clone' but different transfos used than vintage ones... close but not equal imho)).
Last thing, Ams/Neve are in no way classA. Maybe the 88r ( i've not had my hands into one ( but on it i had) nor had access to schematics) but Vseries...nope.
I think there is a misunderstanding:
Rupert Neve owned his Name as a brand until Ams bought it after the company had financial issues. Iirc this happened end ( or mid) 70's. Design were discrete based at that time with transformers in/out and interstage. Those are the Vintage 80XX. 1073,1081,....
From there all said Neve's products ( which should be named Ams/Neve) had only the name on it, Rupert Neve never been involved in the designs.
The next generation of Ams/Neve then used 553x. and mic transfo input (Belclere iirc) By design they are of same gneration than SSL ( but with significant differences in circuits and philosophy / routing is way more complex and elaborate on Ams/Neve than on SSL, automation (Flying Faders)is less intuitive than SSL, so you ussually seen Neve in A for recording/tracking, SSL in B for mixing). Those where the V series (Vr/Vx), SSL concurent was 4000E/G series. The next move was then 88r which was the concurent of SSL9000j/k/Xl.
Meanwhile, Rupert Neve designed the Focusrite console from which all Focusrite range ( Blue and Red) are the children ( they used 553x and same output transfo ( which was very different from the vintage Neve, from what i've seen those are Carnhill transformers, hence the misconception Carnhill did the Vintage one ( who were Marinair). Input transfo are Lundahl 1538.
He designed for Amek too during this period. At the end of 90's to offer an alternative design to 88r/9000 Amek hired him to design the 9098i. Mc33078, 553x and That vca used in signal path as well as the same output transformers than Focusrite's one ( but used differently).
It was the last 'big' inline analogue recording/mixing desk he designed.
Early 2000 he launched his brand RND, with portico range of modules and custom designed consoles ( split for record/mixing and some mastering dedicated one's).
Montserrat console was named. Those were owned by Martin yes but business being business Ams/Neve had a 88r bought by Air studio. The name stay the same, the technology and sound is not. Don't get me wrong, i love the 88r ( as i loved the 9000, 4000 and Vr/Vx ) but the sound as nothing to do with vintage Neve ( despite their was 1073/1081 modules availlable for it ( those where 'clone' but different transfos used than vintage ones... close but not equal imho)).
Last thing, Ams/Neve are in no way classA. Maybe the 88r ( i've not had my hands into one ( but on it i had) nor had access to schematics) but Vseries...nope.
Last edited: