microphones and the cirrus audio card

Hi Grant,

The circuit diagrams for the two cards (old and new) show the difference. You can see a 22k resistor on the L and R input channels connected to the WM5102 MICBIAS pins (described in the WM5102 datasheet). It is likely to be sufficient for any small electret element.

The WM5102 datasheet contains more info, but according to it, the MICBIAS pins can be software enabled or disabled.

There is definitely a microphone bias circuit in the schematics, and even a script to enable recording from a biased microphone.  But I tried just now and couldn't actually measure any bias voltage.on the line-in connector.

Hi Shabaz

Thanks very much for your reply. So this is potentially great news. To clarify: this means the bias voltage will be available via the Line-In?  [I haven't found the circuit diagrams]

thx

grant

Ok Great - thanks Tim

so maybe it has to be enabled?

bw

grant

Grant, if you take a look at the "scenario" scripts supplied in the Cirrus Logic disk image you'll see that Record_from_lineIn_Micbias.sh includes a line to "Enable MICBIAS".  Though for me it doesn't seem to work.

Thanks, Tim

I'll download and have a look.

I was waiting to buy the card until finding out if this functionality exists - it sounds from this as though it does and presumably just needs some tweaking. It sounds like the hardware setup visibly supports PiP via the Line In. So I'll get a card and test.

If anybody is still working with audio streaming with the Wolfson, we have a disk image available from  here, currently being used for a nice live audio stream from Seoul here. This image was very much dependent on support from Ragnar Jensen.

I can't say I particularly look forward to trying to make a new image with Darkice and the Cirrus drivers.. well but with this new capability it should be well worth it

bw

grant

Great thread, all information is very useful.

Some additional points

• The Cirrus card has been updated to include stereo MICBIAS on the Line Input.
• The script setup file in the image will configure the ability for microphone recording
• After executing the setup script, the MICBIAS voltage will only be enabled when an active record is configured, i.e. arecord. This is part of the software power management.
• The MICBIAS voltage is typically 2.7V.

Thanks, Scott

That's very useful and clarifies the query about enabling the bias voltage.

2.7V is not as high as we'd like for the electret microphones we're using, but we will see how it goes.

I ordered a couple of soundcards yesterday and we'll run some tests and post updates, and no doubt questions, here

bw

grant

Electret condenser microphones via the Cirrus Logic Line In - updated

Hi Scott and all

We successfully installed Darkice over the new Cirrus image and all works well. Much easier than previously with the Wolfson.

We ran the use script to record from the Line in with Mic Bias as suggested above, and plugged in a pair of omni stereo microphones of decent quality.

We are now streaming what appears to be very good, detailed audio with a good tone and nice spatial sense - very promising.

However, we have a loud hum which sounds like an earthing type issue.

It looks like there is a hi-pass filter in the script:

# Configure the input path for 0dB Gain,  HPF with a low cut off for DC removal

amixer $1 -Dhw:sndrpiwsp cset name='LHPF1 Mode' High-pass

amixer $1 -Dhw:sndrpiwsp cset name='LHPF2 Mode' High-pass

But the loud hum remains and filtering it would severely impact on the audio.

Any thoughts on how to address this would be great

Thanks

Best wishes

Grant

Hi all

Further to my post on hum using ECM microphones with the Cirrus Line In, we've found that the hum is not present on all microphones.

So this probably  shouldn't  be considered a serious problem.

We'll test the Primo EM-172 and report on how those work.

bw

grant