Hexaphonic Pickup Project

[Mark Hammer]

The diversity of op-amps out there is a reflection of the difficulty of designing a chip to deliver particular sorts of features or specs under ANY set of circumstances. It's night impossible, so what you find is that chip X will deliver such-and-such a spec under THESE conditions, and chip Y manages to deliver the same spec under another set of conditions. And THAT'S why we keep seeing newer op-amp designs that aim for a particular niche, where a particular set of specs are delivered under a particular set of conditions.

The 5534/32 is a great chip for doing certain things, but its best characteristics (like low noise and high current-delivery capability) are exhibited under certain conditions. Look at page 10, here, and you'll see what I mean: http://hammer.ampage.org/files/Device1-8.PDF This information is old but still relevant.

[cycfi]

There's no cure for raising the noise floor with tons of gain, even more cowbell won't do. Some things just are what they are so it's like a moot point talking low noise and excessive gain within the same notion IMO.

The 5532 (and 5534) isn't really a "current" benchmark as it's an old chip from the late 70's which has long been upstaged by newer low noise chips but for musical intrument applications it's a very good general purpose low-noise chip , I've updated/upgraded several studio consoles by replacing 5532/4 chips with lower noise modern equivalents.

It is old, but it is still the benchmark to compare against (see these 2011 EE times articles, for example). I never said they are the latest/greatest.

There are some great white-papers available on noise sources (National Semi + AD) which tend to show that currents through input resistors other than the on-chip resistors are often the "usual suspects" although we do tend to quote the spec-sheet's nV/rtHz statements as the magic number. IMHO, a badly designed circuit can always make a loud-hissy-noised slice of silicon out of a low noise chip.

Nod. I know exactly what you mean (about input resistors).

We went through all this low-noise suff in the 80's/90's until we made everything as sterile sounding as possible, now 20 years later we find ourselves back-pedaling on the notion that noise is a "problem". It's more like getting a great dynamic signal down the cable that serves the end purpose the most and we can use many more devices than the ultra-low-noise devices to get us there.

Noise is always a problem especially in this day and age of 24 bits. I personally want to use all the bits hence 0.88 nV/rtHz vs. 40 nV/rtHz matters.

[cycfi]

The diversity of op-amps out there is a reflection of the difficulty of designing a chip to deliver particular sorts of features or specs under ANY set of circumstances. It's night impossible, so what you find is that chip X will deliver such-and-such a spec under THESE conditions, and chip Y manages to deliver the same spec under another set of conditions. And THAT'S why we keep seeing newer op-amp designs that aim for a particular niche, where a particular set of specs are delivered under a particular set of conditions.

The 5534/32 is a great chip for doing certain things, but its best characteristics (like low noise and high current-delivery capability) are exhibited under certain conditions. Look at page 10, here, and you'll see what I mean: http://hammer.ampage.org/files/Device1-8.PDF This information is old but still relevant.

Here are more recent (2011) articles but still talks about the 5532/34: Op amps in small-signal audio design (part1, part2, part3) by Douglas Self (EE Times)

Nice read, BTW. Thanks for sharing.

[David Schwab]

Hi David,

The hiss would be audible if you crank it up a lot more. I don't think it is relevant for bass, but it will be for, say, heavy metal guitar with tons of gain through distortion pedals. I haven't used LM4250 yet, but I have used Op Amps with the same noise figures (i.e. 40 nV/rtHz measured at 1kHz). An example in that vicinity is the LM324 with 46 nV/rtHz.

I can't hear it even with the bass running though a fuzz pedal with lots of gain. And since it was the op amp EMG used, lots of heavy metal guys use those through high gain amps. Where's the noise?

All electronic devices have thermal noise. This isn't even an op amp meant for audio. Thats what they said when Leo Fender picked it for the Musicman preamp back in the 70s. He did it because of the low current draw. originally that bass didn't even shut the battery off when you removed the plug! Since it's a programable op amp, I increased the current consumption on my version of the preamp for better performance.

But it's surprisingly quiet. It's certainly not a LM741! lol

[rjb]

I can't hear it even with the bass running though a fuzz pedal with lots of gain.

The question is, can a person who hasn't been playing in rock bands since the 70s (60s?) hear it?

[David Schwab]

The question is, can a person who hasn't been playing in rock bands since the 70s (60s?) hear it?

Maybe they can hear it better better, but I hear lots of noisy stuff all the time. When I'm sitting at my computer with headphones of I realize at some point that I can hear some noise from my Johnson J-Station (when I'm not using it), which is plugged into the mixer which is plugged into the computer (via S/PDIF), and I have to reach over and turn down it's volume control. Hsssssssssss. And it's a fairly quiet unit, except when the gain is cranked.

[Mike Sulzer]

But it's surprisingly quiet. It's certainly not a LM741! lol

Indeed it is not. You have to look down in the detailed graphs on the full spec sheet for the 741 to find that it has a full 60 nV/rtHz! Imagine that, an astounding 50% greater than the LM4250, while the 4250 is only about a mere 350% greater than the amp cycfi favors!*

*This ad is brought to you by the Greater Op Amp Professional Users Guild, (GOAPUG), an industry group dedicated to encouraging the use of classic solid state devices and other interesting oddities from the past.

[Steve Conner]

I recently compared the NE5532 against its young upstart cousins the OPA2134 and OPA2604. The old guy actually came out on top with lower noise and distortion, provided the source impedance was low enough, as it was in my application.

With high source impedances, the newer devices stomped all over it as they are FET input, and the 5532 has a hefty bipolar input stage with lots of current noise.

The LM4250 isn't exactly what I would call an audio-grade op amp. Musical instrument grade, maybe.

[Mike Sulzer]

It is a great relief that by 1996 a FET input op amp was readily available that has less than a factor of two more noise voltage than a 12AX7 invented in 1946, I believe. But of course the 12AX7 is not a quiet triode, it is just a good compromise for not drawing much current, and getting an overall relatively inexpensive product. A really good triode is about 15 times quieter still, but that is more a matter for those spendthrift audiophiles (damn that high current, need a fan on my preamp, and it has more noise than my cranked amp!) than practical guitar amp guys.

I recently compared the NE5532 against its young upstart cousins the OPA2134 and OPA2604. The old guy actually came out on top with lower noise and distortion, provided the source impedance was low enough, as it was in my application.

With high source impedances, the newer devices stomped all over it as they are FET input, and the 5532 has a hefty bipolar input stage with lots of current noise.

The LM4250 isn't exactly what I would call an audio-grade op amp. Musical instrument grade, maybe.

[RedHouse]

...The old guy actually came out on top with lower noise and distortion, provided the source impedance was low enough, as it was in my application....

And it's why we see them in consoles a lot, but rarely in musical instrument circuits.

Steve, was the test for a lab customer or for your own interest/work? if it was your own did you happen to write it up?
(PM me if you have a PDF or something)

[Steve Conner]

It was a combination of two projects, a low-distortion oscillator for my own use and an ultrasonic piezo preamp for work. I don't have anything I can publish.

[RedHouse]

Ok Steve.

[cycfi]

I can't hear it even with the bass running though a fuzz pedal with lots of gain. And since it was the op amp EMG used, lots of heavy metal guys use those through high gain amps. Where's the noise?

It's gated out by the noise gate :-)

All electronic devices have thermal noise. This isn't even an op amp meant for audio. Thats what they said when Leo Fender picked it for the Musicman preamp back in the 70s. He did it because of the low current draw. originally that bass didn't even shut the battery off when you removed the plug! Since it's a programable op amp, I increased the current consumption on my version of the preamp for better performance.

But it's surprisingly quiet. It's certainly not a LM741! lol

Nods. I guess it shouldn't matter that much. There are lots of things in the signal chain with more noise anyway. Not to mention that 50s style Hi-Z pickups are very noisy to begin with.

[rjb]

... 741 to find that it has a full 60 nV/rtHz! Imagine that, an astounding 50% greater than the LM4250,

Question:
When doing these kind of comparisons, does one typically state the difference in ratios, or do you consider that a 10dB increase is perceived as about twice as loud? In other words, do you say "50% greater" or 20log(60/40)=3.52dB= "barely perceptible"?

[David Schwab]

It's gated out by the noise gate :-)

None of the guys I play with that use EMGs use noise gates.

Nods. I guess it shouldn't matter that much. There are lots of things in the signal chain with more noise anyway. Not to mention that 50s style Hi-Z pickups are very noisy to begin with.