Very low noise op amp recommendations?

[bikerhifinut]

Hi guys, if this is the wrong section mods please move to DIY but I thought this bit may get more exposure.
I am mulling over a DIY RIAA preamp stage, I do have very good disc stages in the solid state commercial preamp but both for personal satisfaction (How many percent better again for DIY RW?) and for the occasion when I use my preferred integrated KT88 Valve amp, basically a power amp with selector and Volume pot etc, I'd like a good quality disc stage to do justice to a Good Moving magnet cartridge (Ortofon 2M Bronze) and deck. I do have an ECC83 based RIAA pre amp which is pretty nice but my main criticism of it is that for me it has excessive levels of noise (Hiss and various valvey rustles.) I could have a go at that and see if theres any underlying reason for the noise but frankly I think I'll end up going round in circles.
So I have dug out some circuits and have pretty much settled on a fairly simple 2 stage circuit based on op amps. It's entirely conventional, being a fairly low gain first stage of around 20 db into a second stage with higher gain and the EQ network nested in its feedback loop, again I like that idea over passive EQ having weighed up the pros and cons I feel it gives a better overall compromise. I can post the circuit if forum rules allow, as its cribbed from the R A penfold op amp book by babani and I imagine as its a dIY bok there's no issues around building and modding as far as copyright etc goes. And I don't feel like reinventing the wheel when messrs baxandall and Lipschitz etc have done the hard work already!
The circuit uses NE5534 as first stage into an LF351 second stage, I suspect the TL072 wasn't mainstream when the book was written.
So to the nub of the matter, the first stage is obviously to me the crucial one as far as noise is concerned. I can reduce it with the use of metal film resistors and careful wiring/layout. The second stage as far as I can figure will be well served with the trusty TL071/2 series as the signal levels will outweigh any underlying noise issues in any practical sense.
Given that the NE5532/4 types are marketed as low noise audio devices anyway, is there any perceivable benefit to the hiss levels by choosing a different op amp? Previous experiments with a single stage RIAA preamp kit from velleman showed that it aint always a given that a "better" device provides a better result. Thanks to David RW for the tutorial on that a couple of years back.
I appreciate that discrete transistors can produce a better result but that's not the way I want to go.
I'm also thinking to use batteries to power it as a pair of 9V PP3 in series should last a long time, or maybe use one of the AA holders I have that gives 15V on a single rail.
A long question, I hope the answers aren't so rambling as my efforts!

PS I do realise that the surface noise off many records swamps the hiss from my valve stage but I find the hiss when between sides to be somehow very annoying especially when I have worked so hard to get background noise levels down on the rest of my kit. and some of my better pressed LPs do let the background hiss of the disc stage through.

Many thanks in advance.

Andy

[Ambientnoise]

Suggest you have a look at ESP website (Elliott sound products), project 187 for example. Some of the lowest noise opamps are SMD only, have particular supply or source impedance requirements, or are merely very expensive !

Ken

[bikerhifinut]

I am familiar with Rod Elliots site, thanks for the reminder, however I don't agree with some of his philosophy especially around the use (or non use) of feedback on Disc stages and Tone controls. And law faking for linear pots, if its that important you have perfect channel balance, then these days its cheap and easy to buy or make a switched attenuator. OT there sorry.
Others may disagree with me, I wont get upset.

I do have a feeling that it's simply not worth trying to improve on the NE5532/4 types at Moving magnet voltages, perhaps moving coil levels would warrant it but I decided to walk away from MC this year when the cost of replacing an accidentally damaged ortofon quintet made my eyes water.

A.

[paulsherwin]

The op-amp types that you mention are mass market general purpose audio devices, and while perfectly adequate for most jobs will not give you the best possible performance. There is plenty of discussion of more esoteric op-amps if you google, though there is a lot of the usual hifi woo to wade through.

If you build the preamp with IC sockets you can start with cheapies like the TL072 and then try different chips as inclination and budget permits.

[ajgriff]

Originally Posted by bikerhifinut

I do have a feeling that it's simply not worth trying to improve on the NE5532/4 types at moving magnet voltages .....

Agreed.

Alan

[Radio Wrangler]

With careful amplifier choice you might almost do 1 dB better noise floor, but at a cost of money and getting less-available parts.

The NE5534 has a number of good parameters for the job. Low input noise, yup. Low input noise with source impedances similar to MM cartridges is more subtle, but it's a YUP too. Good driving capability, Yup. This allows you to use low-Z feedback networks which also save on noise. Remember the noise-matching impedance thing on the non-inverting input for the cartridge applies just as much to the feedback network into the inverting input. Enough gain and plenty of bandwidth.

Most of these parameters can now be beaten by one device or another, but the combination of them all together makes finding something that is overall better somewhat more difficult.

You'll always get some noise from magnetic cartridge, tape head and microphone input stages and you'll be able to hear it when there is no material being transduced. Once you know it's there, or suspect it's there, you will concentrate on listening for it. I don't think there is anything good enough to avoid this perception. So you might just as well do a reasonably good job and get on with listening to some music.

TL072 are fairly low noise, not as good as NE5534 in basic terms, but their optimum Z source for noise figure is much less suitable in this application.

There is a slight disadvantage to using a flat gain amp ahead of an RIAA stage. You don't get the same headroom, and the common mode input of the second stage amplifier is larger, leading to more distortion. To an audiophile once such things are said, they can never be unsaid and they assume great importance. Non audiophiles are, in contrast, allowed to do sums and even measure things to find the extent of such effects thereby to see them in proportion.

A single NE5534 with the missing pole added after works fine for me. Maybe having a buffer after that added pole would be good if a cable comes next.

101.832% of all statistics are made up on the spot.

David

If I can hear the noise on a recording (not the medium) that's a system good enough for me.

[Electronpusher0]

Originally Posted by Radio Wrangler

A single NE5534 with the missing pole added after works fine for me. Maybe having a buffer after that added pole would be good if a cable comes next.

This is the system used in a circuit available as either a bare board or kit on e-bay.
Search for
MMCF10 HIFI LP phonograph MM amplifier

I traced the schematic (attached)
It is intended to use OPA134 but I used the NE5543 instead.
I use this to feed my Mullard 5-10.

Peter

[mhennessy]

About 20 years ago I built a version of the Douglas Self pre-amp, and I think to improve on it would take you seriously into the realm of vastly diminishing returns.

It uses a 5534 at the input, and a unity gain stage for the IEC amendment. It could do with perhaps another 10dB of gain, which would be easy to achieve with a 5532 or similar.

I'm not a big vinyl user, so haven't ever felt the need to revisit this over the years - been a bit busy with bringing up the kids, etc. But if I did get the itch, I'd pick up a copy of his dedicated book: https://www.amazon.co.uk/Electronics.../dp/1138705454 - though based on some of the reviews, it appears that I might already have a lot of the content in my copy of Small Signal Audio Design. If you have neither, I'd definitely pick up the latter first of all. I refer to my copy all the time - like all of Doug's books, it manages to be both a great read and a great reference. If only all text books were that well written...

[stevehertz]

In my Quad 44 preamp I replaced the TL071 op amps with OPA 134 ones that are designed specifically for high quality audio and were not available when the 44 was designed. These offer lower noise, higher slew rate and much lower distortion than the TL071. The improvement was noticeable. There is a further upgrade to OPA627s but they are a lot more expensive and you're getting into the realms of diminishing returns.

[Ted Kendall]

For moving magnet, 5534 every time. The combination of virtues for this application is still practically unbeatable. The LM4562, which betters the 5534 around most of the course, albeit at a price, falls down on current noise.

[McMurdo]

I'd be tempted to obtain a diagram for a commercially available pre-amp circuit fitted to some decent amplifier, and then just copy that. Saves relying on some hobbyists sketchings that may or may not work as intended. At least a production model will have been proven beforehand!

[Ted Kendall]

Self's design has been commercially available for several years, so can be said to be proved in the field.

[Electronpusher0]

Originally Posted by McMurdo

I'd be tempted to obtain a diagram for a commercially available pre-amp circuit fitted to some decent amplifier, and then just copy that. Saves relying on some hobbyists sketchings that may or may not work as intended. At least a production model will have been proven beforehand!

I hope you are not refering to the hand drawn sketch in post 9, that is, as you suggest, the tracing of a commercial model, and it does work, I measured it on the bench as well as listening tests.

Peter

[bikerhifinut]

Thanks for the input.
I think my question has been answered, i.e the 553x types, given a genuine article, are quite good enough for a fairly simple op amp RIAA step up for moving magnet with around 5mV output.
Paul Sherwins observations about audio foo gave me an inward smile, I remember from a previous dabble with a single stage velleman kit that "better" op amps can bite you on the rear end if you aren't careful. And his advice about sockets is my default position with circuits using op amps for reasons other than being able to swap different ones in and out.
Given that I have stocks of 5532 and 5534 (the difference only being that the 5532 is a dual op amp, and I have stocks of TL071 and TL072 again its the single vs the dual then I won't waste time trawling the suppliers for other chips.
I reckon its going to be quieter (less hissy) than a valve stage especially as RW states, after the first stage the circuit works at lower impedances. Thanks David I think I get it and it gives me food for thought about experiments around the feedback based EQ. The whole thing has to be a compromise anyway to get a decent approximation of the EQ.

The explanation of the circuit I have been perusing, and its a dead simple one really, is that a modest gain stage into a second stage giving the lions share and the EQ helps in as much as the treble cut from the EQ helps reduce any noise from the first stage. The TL07x isn't a bad chip in my book and as its not in the low signal section of the circuit any self generated noise will be immaterial to the audible result. I'll take a photo of the page if anyones interested. it's from the Babani preamplifier and filter circuits book all based around op amps. there are a couple of tweaks it needs to improve any stability issues and decoupling but these are minor and easy to incorporate. I'm also intending to power it from a battery rather than a dedicated mains PSU, again its a theory that it helps reduce noise induced via the power supply. If I simply put a DC socket on it then there's the option of a stabilised wall wart or purpose built PSU, that's for the future.
Its as much a bit of fun as a serious attempt, I have a commercial preamp that is probably as good as it gets without getting silly and that's what I very happily use with what I consider to be better than average turntable/arm/cartridge. This is to see how close something rather modest costing beer money gets to a competent modern commercial product, and for when I am experimenting with different amplifiers. Which I do for the hell of it and it keeps me off the streets.

Again thanks for the input people, I do appreciate everything that is contributed and knowledge is never wasted.
Andy

[bikerhifinut]

Originally Posted by McMurdo

I'd be tempted to obtain a diagram for a commercially available pre-amp circuit fitted to some decent amplifier, and then just copy that. Saves relying on some hobbyists sketchings that may or may not work as intended. At least a production model will have been proven beforehand!

In my original post I did say I wasn't going to reinvent this particular wheel, It's pretty much all been done before so I concur.

A.

[Ted Kendall]

The 5534 is a couple of dB quieter than the '32, so would be worth choosing here. Note also that the drive capability of the 5534/2 is considerably better than that of the 072, which matters here with a low impedance feedback network.

[bikerhifinut]

And there's the added bonus Ted of being able to keep both channels well physically separated which may or may not assist with crosstalk? I was already thinking about going down this route, or if using a dual op amp it would be one op amp per side. The issue with the 5532/4 as far as I am aware is that if used for very high gain it can get stability issues, I seem to remember this from a previous experiment with a single stage op amp RIAA preamp.
I must get a copy of Selfs book, it will probably go over my head but I'm not too old to learn!

A.

[GrimJosef]

Originally Posted by Radio Wrangler

... You'll always get some noise from magnetic cartridge, tape head and microphone input stages and you'll be able to hear it when there is no material being transduced. Once you know it's there, or suspect it's there, you will concentrate on listening for it. I don't think there is anything good enough to avoid this perception ...

Optical sensor detecting tone-arm lift and operating a mute circuit between the pre and power amp ?

Cheers,

GJ

[mhennessy]

Another difference between the 5534 and 5532 is that only the 5532 is unit-gain stable. The 5534 needs an additional compensation capacitor if you plan to use it at gains of x3 or less.

I honestly wouldn't worry about crosstalk, given the source But the 5534 is better-suited to the task.