LMV321 as line level audio buffer?

[dobsonr741]

I received a ton LMV321 opamps, planning to use it as input buffer and differential output driver. I wonder if you’ve experienced any issues using it as line level buffer in audio?

[strawberry]

isnt it something to do with 600
inputs seem to be well protected

[coppercone2]

I don't think there will be a problem, especially if its unity gain, it should be similar to a LM741 for noise.

I did not see you mention high-fi requirements (do not use for microphone amplifier)

The input protection circuit is valuable for a buffer. Circuit that does not break easily while performing its function is a great asset

[dobsonr741]

Signals at 1Veff coming off of 3.3V powered codecs. Not a Hi-Fi environment, it's a synthesizer.

[coppercone2]

its probobly good enough

[macboy]

The Equivalent input noise voltage of LMV321 is listed as 39 nV/sqrt(Hz) at 1 kHz, which is on the high side. Very cheap "audio" opamps typically have less than 15 nV/sqrt(Hz), and better ones have less than 5 uV/sqrt(Hz). E.g. NE5532 or RC4580 or LM833 - inexpensive, decades old designs commonly used for audio - all come in around 4.5 nV/sqrt(Hz). Some of the best designed by National Semi (and mostly cancelled by TI) were ~2 nV/sqrt(Hz).

The open-loop gain is also a little on the low side. This is in line with the distortion figures in the datasheet, which are not great: no better than 0.01% THD+N at unity gain, and worse with gain. This might not sound like much, but most opamps for audio are at least an order of magnitude better, and the best of them are three orders of magnitude better (0.00003%)

The above doesn't mean that they won't work for audio, they will of course. You will need to determine for yourself whether the performance is adequate. "Audio" is a broad category; are we talking about Hi-Fi or an intercom system or ...?

[donlisms]

A little more theoretical analysis...

With 39uV * sqrt(20kHz) giving 5.5uV noise, if this was a mic signal at, say, 5.5mV, that would be 1/1000 or -60dB SNR, which is not too good.  For line level at sqrt(600/1000) volts (0.775 volts), its -103dB, which is a pretty reasonable number.  So... the amps would be quiet enough if you used them appropriately.

And since you have them... try it!    They will speak to you.

[schmitt trigger]

As others have mentioned, you already have got them, might as well use them.

The important thing is that these opamps have industry-standard pinouts, such that if you build the circuit and the noise performance is not acceptable for your application, you can always replace them with an improved device.

[dobsonr741]

The important thing is that these opamps have industry-standard pinouts

This is my pain point. The LMV321IDBV SOT23 layout is non-standard. Anyhow, I listened to it on the breadboard, had no audible noise at line level. The PCBs are already churning at OSH Park, fingers crossed.