Au Revoir NE5534

[knobtwiddler]

This news is a from a few weeks back, but I am still struggling to come to terms with it: https://www.diyaudio.com/community/t...ounced.405443/

My understanding is that none other than Rupert Neve himself worked as a consultant on the TDA1034 (that preceded the 5534), and his first console to utilise it was the highly desirable 8078 (much of the front end is discrete, but the driver stages used the new IC).

One of our esteemed members worked at Neve and mentioned it a while back: https://www.vintage-radio.net/forum/...71&postcount=7

I recall reading recollections from an engineer who interacted with BBC engineers when working at Neve, where he stated that the newer cards which had the TAD1034 stages not only gave them less stress because of enhanced reliability, but the Beeb staff also felt they sounded better than the discrete output blocks they replaced! (could be due to their ears being more relaxed on account of knowing said mixer isn't about to drop a channel during a recording of a TV show!).

Another thread on its merits:
https://www.vintage-radio.net/forum/...hlight=tda1034

Although there are a select handful of newer ICs that come close in terms of the current noise requirement needed to beat the NE5534's noise performance for a moving-magnet preamp stage, I have yet to see actual measurements of a real world design. You can beat the 5534 for an MM stage if you go discrete, but only by a 1dB margin.

The 5534 has more parallel input BJTs than its 5532 sibling, hence its lower noise. It'll drive 600R all day long. It has pins for external compensation, making it more versatile than the 5532. Up until recently, you could get 1000 of them for less than 50p a pop. Where else do you get such functionality for such low outlay?

Douglas Self really can't be happy... It truly feels like the end of an era

[jamesperrett]

I have a Neve 33122a module which uses a TDA1034 just after the input transformer. I also have a 33122 which uses a discrete module after the input transformer.

[knobtwiddler]

I guess the fact that they were used as O/P drivers as well as I/P stages is testament to how versatile the IC is.

edit - I should also state that the BBC story was from memory, and it might have been parts other than the O/P stages that the 1034 replaced. The general sentiment about the beeb being happy about the 1034 was definitely accurate.

Mr Tanner (post 14):

the BBC would not accept consoles fitted with the discrete amps due to their poor overload recovery. Neve thought the IC amps were better, the BBC laboratories thought the IC amps were better and I share their point of view. It's the iron makes the sound more than choice of amplifier

https://gearspace.com/board/high-end...68-8078-a.html

(I'd give the other posts a wide berth at that group!)

There must be a thousand other OEMs who used the 1034 / 5534 to great effect. A landmark component, without doubt.

[Ted Kendall]

Is this all 5534s or merely DIL packages?

[knobtwiddler]

I am 99% sure that all are going. I have tried searching SOIC and all variants at Ti site and get an EoL message:

https://www.ti.com/product/NE5534?ke...secase=GPN-ALT

The confusing thing on their site is that if you search 'NE5534' it says 'active', but when you look at variants, they all say EoL.

[paulsherwin]

Presumably there are second source Chinese suppliers who will continue to supply this chip. Take away the history, and it's just a high quality audio op-amp.

I don't doubt that won't satisfy the golden ears brigade though.

[Jon_G4MDC]

That was a nice double inverting opamp.

[knobtwiddler]

It never has satisfied the golden ears brigade *when they've known that the device contains it* (or the other ubiquitous NE5532). The fact that (this is a quote from Self) almost every record made in the last 30 years has been through about a hundred NE5532.

I'm not aware of any Chinese supplier for the 5534. There were 2 other firms making it, OnSemi (ex-Motorola) and Nishimbo (Japanese Radio - they still make the 5532).

Ti were packaging it in China, I believe (apparently the wafers were made elsewhere).

I would put money on no firm in China getting the same specification. The Signetics NE5532 had better performance than Ti's 5532 (I thought this was an urban myth, but then tested it myself). These ICs may be 45 years old, but they are capable of sub -100dBV noise + THD performance. When you get to that level, everything that happens at the wafer fab plant matters.

[knobtwiddler]

Originally Posted by Jon_G4MDC

That was a nice double inverting opamp.

If you mean 'dual channel' by double, then fear not, the NE5532 is still made by Ti, as well as other firms such as Nishimbo (which allegedly outperforms the Ti part - I must see if this is true). It is almost the same design, but lacks the 5534's external compensation and isn't quite as quiet in terms of current noise.

[cmjones01]

There are an awful lot of fake NE5532s around. There are lots of audio modules advertised at low prices on the usual merchant sites which say they use NE5532s. I bought a tone control module recently described as such. The chips do say NE5532 on them, but they're not the real thing - the power supply current is much too low and the slew rate is hopeless, much more like a common-or-garden 1458 or similar. I haven't tried measuring noise because it would be pointless!

Chris

[Hartley118]

Regarding knobtwiddler’s post 1, I’m not sure that I’ve ever been called ‘esteemed’ before, but I do recognise myself in this description. It’s intriguing, and somewhat disturbing, how myths can develop in engineering as well as in more spiritual pursuits. No, Rupert Neve had nothing to do with the design of the TDA1034/NE5534, though I guess that its subsequent reputation may have been connected with the Neve company’s adoption of it to replace previous discrete component amplifier units.

The full story is too long for a forum contribution, but goes back to the long hot summer of 1976 at the Neve plant in Melbourn, one of the ‘necklace’ of high-tech villages encircling Cambridge that together made up the famous ‘Cambridge Phenomenon’.

I’d recently joined Neve as Technical Director and the company had won an order from the South Africa Broadcasting Corporation for a large number of portable mixers for OB use. These were pretty far outside the scope of Neve’s usual design experience. What’s more, our then CEO had somewhat mischievously assigned Rupert himself to the design of these products. This presented a management challenge because Rupert needed a couple of experienced design engineers working with him to complete the more routine aspects of the job. Rupert was at that time a consultant to the Neve company and no longer a full-time employee.

In those days, Neve circuits were strictly discrete component, generally using the BA438 low signal discrete voltage amplifier and BA440 discrete output amplifier modules. However, a compact portable mixer clearly cried out for IC amplifiers. But the professional audio world was at that time highly sceptical of the performance of ICs. What to do? Here, Rupert, originally one of the greatest sceptics of ICs came into his own because a friend of his was Alex Balster, designer of desks for Philips studios in Holland. I was then party to a phone call between Rupert and Alex in which Alex waxed lyrical about the audio performance of the TDA1034 which had originated in their own Philips labs. Alex may well have advised on the performance spec. So we adopted it, firstly in this our first bespoke portable mixer, then in the highly successful 5422 ‘suitcase’ mixer which also subsequently gave birth to the entire ‘54’ series of compact studio mixers.

It wasn’t long before I wanted to introduce the TDA1034 into our mainstream desks, leading to Neve circuit designer Jon Brown cleverly modifying the BA438 and BA440 modules to become the BA638 and BA640 where the voltage amplification is handled by the TDA1034 instead of the previous discrete components. That’s why you’ll find the TDA1034 in the rather traditional but highly regarded 8078 music recording desk. It wasn’t long though before the IC migrated onto the actual channel motherboard in the 81 series music recording desks and 51 series broadcast desks within 2-3 years or so.

These more recent desk systems also used a split rail power supply, which does give a superior overload recovery performance compared with the original single 24V B+ rail (which by the way you’ll still find on the highly valued 8078). By the way, any perceived difference in sound quality or reliability is likely to be due to factors other than the IC.

Martin

P.S. The BBC acquired a great many desks with the discrete amplifiers and the older single rail supply, so they can’t have been too troubled by the indifferent overload recovery.

[Radio Wrangler]

Thanks, Martin, that's an interesting insight into where a few legends came from.

In our lab in the late 70's the NE5534 came to notice. I'd been looking into opamps worth using in low phase-noise PLLs and the NE5535 tripped the 'Blimey, that looks useful' alarm. Useful enough to re-engineer some circuitry to bring the source impedance of a phase detector down low enough to play into the opamp's strong suit with low-Z sources. The low-Z capable output was also exploitable ro keep resistor values low in feedback networks and to use choke insertion of tuning voltage into VCOs, all avoiding high value resistors and their inevitable noise. High sensitivity, low noise audio is a critical area for low noise amplification, but so is low noise PLL design.

David

[m0cemdave]

I remember somebody showing me the latest Midas input module, I think it was for the original PF mixer. They pointed out the great innovation - a TDA1034, the original 5534 in a TO5 can. They also told me what it cost, something like £13 (and that was to an OEM). Which was getting on for 2 days' wages for most people at the time!

[Craig Sawyers]

Indeed the end of an era; what are TI thinking? When there is a continuing resurgence of record buying, currently handsomely exceeding the "Pure Perfect Sound Forever" CD sales, TI obsoletes for the lowest noise magnetic cartridge opamp going.

It is similar to the corporate blindfold that caused Shure to exit the cartridge market during the record revolution.

Aagh

Craig

[Craig Sawyers]

The only version that OnSemi has left as a current device is the SOIC version. So there seems to be some sort of agreement between the two main manufacturers of the device.

Craig

[Valvepower]

Tell me about it, obsolescence of linear semiconductors both IC and discrete is a constant battle… there is a void between a lot of Semiconductor companies and the continuing resurgence of record buying and associated equipment. A few semiconductor manufacturers are on board - so to speak - however a lot of the the big boys aren't

Oh well, as Bod Dylan sung…, The Times They Are A-Changin'...

Terry.

[Valvepower]

Forgot to say…,

As mentioned earlier in the thread Nisshinbo/JRC are still making suitable parts. These parts perform OK for me in both technical and listening tests.

Terry.

[knobtwiddler]

Originally Posted by Craig Sawyers

The only version that OnSemi has left as a current device is the SOIC version. So there seems to be some sort of agreement between the two main manufacturers of the device.

Craig

The Onsemi version I can find at Mouser is listed as EoL. I hope that I am missing something here and there is another package of 5534 available

[Craig Sawyers]

Although it is only SOIC8 (and tinier packages), the OPA1611/12 (single and dual) give the same input referred voltage noise in 20kHz bandwidth with 10k source resistance (2uV).

I haven't done the comparison in my RIAA calculator yet, but I will do and report back.

Craig

[knobtwiddler]

Thank you very much for putting me straight on the Neve story, as well as taking the time to write a concise history, Martin. I have read it at a few places over the years and I can only conclude that people have conflated the Neve company's early adoption of the IC with Mr Neve having been involved in the development.