开云体育

ctrl + shift + ? for shortcuts
© 2025 开云体育
  1. MicBuilders
  2. Topics

FET OPAs for mikes

 

I'm a bit out of touch with modern OPAs but note there are a number of new DIY mikes using FET OPAs like Homero's Charis and the OPA Alice.
?
I'm after a FET OPA with less than 3.5nV/rtHz noise @ 1kHz and good 1/f, popcorn noise bla bla.
?
What do you guys use and recommend ??
 

I am using the OPA1642. It is 5.1nV√Hz. The OPA1652 is 4.5?nV√Hz? Both are very good. What is your noise floor based on out of curiosity?

Jules

On Mon, Apr 7, 2025 at 4:39?PM Richard Lee via <ricardo_lee=[email protected]> wrote:
I'm a bit out of touch with modern OPAs but note there are a number of new DIY mikes using FET OPAs like Homero's Charis and the OPA Alice.
?
I'm after a FET OPA with less than 3.5nV/rtHz noise @ 1kHz and good 1/f, popcorn noise bla bla.
?
What do you guys use and recommend ??



--
Best Regards,

Jules Ryckebusch

214 399 0931
 

It's actually for a Jurassic application with of no modern importance;? RIAA MM preamps in DIYaudio.? I'm surprised that NE5534 is still one of the best for this; slightly beaten only by $$$ OP627 or Unobtainium NJM2068.
?
In theory, a FET OPA with <3.5nV/rtHz & low 1/f, popcorn noise bla bla should trounce 5534
 

开云体育

Look at this one:?? One of my reasons for choosing the OPA1642 was idle current. The OPA1633 ?has >10ma which for a phono preamp with external supply is not an issue?
Best Regards,

Jules Ryckebusch?

On Apr 7, 2025, at 21:32, Richard Lee via groups.io <ricardo_lee@...> wrote:

?
It's actually for a Jurassic application with of no modern importance;? RIAA MM preamps in DIYaudio.? I'm surprised that NE5534 is still one of the best for this; slightly beaten only by $$$ OP627 or Unobtainium NJM2068.
?
In theory, a FET OPA with <3.5nV/rtHz & low 1/f, popcorn noise bla bla should trounce 5534
 

ADA4625-2 bifet from Analog Devices is 3.3nV/rootHz at 1kHz, and 5.5nV/rootHz at 10Hz. Current noise 4.5fA/rootHz at 1kHz

The OPA828 JFET has 4nV/rootHz at 1kHz, 7.5nV/rootHz at 10Hz, and 1.2fA/rootHz current noise. Available as a dual OPA2828. Surface mount of course.

And the only bipolar opamp to beat the NE5534 is the OPA210. 2.2nV/rootHz at 1kHz, an amazing 2.5nV/rootHz at 10Hz, and crucially for MM cartridges current noise of 400fA/rootHz (same as for the NE5534). It is just better in all respects than the Neanderthal NE5534.? And available as a dual OPA2210.

If you put it into my RIAA amp spreadsheet, for typical cartridge R and L the OPA210 only finds you 0.6dB improvement in SNR as compared to the NE5534. For a factor of 5 times greater cost. But distortion is vanishingly low, it only draws 2.5mA quiescent, and inevitably is only surface mount. And cheerfully drives 600 ohms with no degradation in distortion.

Craig



------ Original Message ------
From "Jules Ryckebusch via groups.io" <ryckebusch@...>
Date 08/04/2025 03:55:22
Subject Re: [MicBuilders] FET OPAs for mikes

Look at this one:?? One of my reasons for choosing the OPA1642 was idle current. The OPA1633 ?has >10ma which for a phono preamp with external supply is not an issue?
Best Regards,

Jules Ryckebusch?

On Apr 7, 2025, at 21:32, Richard Lee via groups.io <ricardo_lee@...> wrote:

?
It's actually for a Jurassic application with of no modern importance;? RIAA MM preamps in DIYaudio.? I'm surprised that NE5534 is still one of the best for this; slightly beaten only by $$$ OP627 or Unobtainium NJM2068.
?
In theory, a FET OPA with <3.5nV/rtHz & low 1/f, popcorn noise bla bla should trounce 5534
 

On Tue, Apr 8, 2025 at 04:55 AM, Jules Ryckebusch wrote:
Look at this one:?? One of my reasons for choosing the OPA1642 was idle current. The OPA1633 ?has >10ma which for a phono preamp with external supply is not an issue?
Best Regards,
?
Jules Ryckebusch
?
OPA1633 is a bipolar differential amp with an Optimum Source Impedance of about 800 ohms. hardly suitable for a MM preamp.
You must have confused it with another.
OPA1655/56 are JFET with 2.9nV/sqrtHz IN density.

?
 

On Tue, Apr 8, 2025 at 01:21 AM, Craig Sawyers wrote:
And the only bipolar opamp to beat the NE5534 is the OPA210. 2.2nV/rootHz at 1kHz, an amazing 2.5nV/rootHz at 10Hz, and crucially for MM cartridges current noise of 400fA/rootHz (same as for the NE5534). It is just better in all respects than the Neanderthal NE5534
Thanks for these suggestions Craig.
?
OPA210 has EVIL input bias current cancelling.? The damning evidence is its +/- input bias current spec.? These invariably have much higher noise for stuff with different impedances on each input pin like single OPA RIAA preamps.? You may remember us looking at the ???? preamps with Steve Sells and finding significant improvement replacing OP27 or OP37 with 5534 on their RIAA
?
It's more than a decade since I last looked at this seriously
?
Wayne (and NWavGuy IIRC) measured NJM 2068, sadly Unobtainium today, quieter than LM 4562 so env < 2.5nV/rtHz @ 1kHz and inv good enough in to beat 5534.? Of course da Jurassic NJM topology (dating back to Raytheon 4136 ie much older than 5534) can't give 1 ppzillion THD like more modern stuff but good enough for an old beach bum like me.
?
BTW, I hope you are not suffering various health issues like me which may mean I need to give up the beach
 

Indeed bias cancellation is a hidden evil in recent (and not so recent) audio opamps. It is simply omitted from the specs, and in the neat and ideal noise calculations with resistive generator impedances.

Thanks for spotting that for the OPA210. And like I said, even ignoring bias compensation it only gains you a mere 0.6dB SNR as compared with the 5534 in MM SNR.

And that is something in favour for the 5534 and 5532 still - no bias compensation. And still available in 8-pin DIL packages. I just wish that they had the comprehensive set of performance graphs of more recent devices.

Sorry to hear of what ever has caused the beach bum from quitting the beach. Following a five month chest infection that needed kick ass antibiotics to kill, I'm luckily back to firing on all cylinders.

Craig

------ Original Message ------
From "Richard Lee via groups.io" <ricardo_lee@...>
Date 09/04/2025 05:00:30
Subject Re: [MicBuilders] FET OPAs for mikes

On Tue, Apr 8, 2025 at 01:21 AM, Craig Sawyers wrote:
And the only bipolar opamp to beat the NE5534 is the OPA210. 2.2nV/rootHz at 1kHz, an amazing 2.5nV/rootHz at 10Hz, and crucially for MM cartridges current noise of 400fA/rootHz (same as for the NE5534). It is just better in all respects than the Neanderthal NE5534
Thanks for these suggestions Craig.
?
OPA210 has EVIL input bias current cancelling.? The damning evidence is its +/- input bias current spec.? These invariably have much higher noise for stuff with different impedances on each input pin like single OPA RIAA preamps.? You may remember us looking at the ???? preamps with Steve Sells and finding significant improvement replacing OP27 or OP37 with 5534 on their RIAA
?
It's more than a decade since I last looked at this seriously
?
Wayne (and NWavGuy IIRC) measured NJM 2068, sadly Unobtainium today, quieter than LM 4562 so env < 2.5nV/rtHz @ 1kHz and inv good enough in to beat 5534.? Of course da Jurassic NJM topology (dating back to Raytheon 4136 ie much older than 5534) can't give 1 ppzillion THD like more modern stuff but good enough for an old beach bum like me.
?
BTW, I hope you are not suffering various health issues like me which may mean I need to give up the beach

1 - 8 of 8