MAX9814 - Too much noise on output.

The datasheet specifies 5.5uV RMS as the typical output noise, so 10mVp-p is not expected.  Could you provide details about your setup?

What pins are you shorting?

\What are you using to measure the output noise?

Based on the block diagram in the DS, nothing on the input should be shorted.

You could remove the microphone, but the MIC bias should always be connected to the coupling capacitor feeding into the input.

Thank you for your interest in this matter.

I used a shorted phono plug on the input connector on the EVM whilst looking at the output using a scope with a close coupled Gnd lead. The 10mV pp is an eyeball estimate of the differrence seen on the scope between power on and power off with power being from 3x1.5V (AA) type batteries.

A couple of things to note:

-The node you are shorting out is not meant to be shorted.  If anything, just leave it open to measure the output noise.

-Measuring the output with an O'Scope does not provide an accurate measure of output noise.  Measured to specification, you would want the integrated noise from 20Hz to 20kHz.  Test gear like an Audio precision will provide the required filtering and integrated output noise.

-I want to correct something from my previous post.  The 5uVRMS I previously mentioned is the noise on the MICBIAS.  The output noise of the MAX9814 is 430uVRMS.

If I leave it open the noise level is higher. Using DSP on the scope to filter out anything above 20kHz, the noise level is approximately double between shorted and open.

My ears tell me there is too much noise. If I connect the preamp EVM to a power-amp EVM (MAX 9768) - which has no audible noise until the preamp is powered there is then  back gound hiss. 

I will get ahold of one of these and test it on the bench myself to see what I measure.

Can you please provide complete details of your setup?  Anything that will help me understand your setup like photos, wiring diagram, testing procedure.

Test set up shown, 5V bench PSU for the supply.

I do have data from Picoscope measurements if you can read them (the free Picoscope app will open them from picoscope.com) but they won't upload directly to this forum.

On a speaker the background hiss with the pre-amp powered is possibly acceptable but through a headset amp and headset it is not. 

Thank you for your support in this matter it is much appreciated.

I should have an EVK in house today for a look at the output noise.  I was thinking about this a bit and even the quoted 430uVRMS in the DS is likely audible depending on the speaker used and how close it is to the ear.  For reference, our newer in-ear speaker drivers aim for <10uVRMS so that the noise is not audible.

I'm after a microphone pre-amp not a speaker driver per se. The speaker driver seems Ok and is quiet enough until the pre-amp is attached.

Alright we've looked at this in the lab and the noise is in-line with the specified 430uV in the datasheet.  Keep in mind the noise from the MAX9814 will be amplified by the next amplifier in the chain, so the noise will only get worse.

Perhaps we could back up a step figure out if there is a better solution for your design.  If you can describe your design, perhaps we can find a better set of parts to meet your needs.  

Hi Mike

Thank you for checking that out for me.

Whereas I would greatly appreciate further help, I can't discuss the application in detail on a public forum I'm afraid as my client would not appreciate that. Can we take it off line?

Neil