New reworked driver for Wolfson/Cirrus Logic audio card

Hias,

A huge thank-you for all the work you've put into supporting the Cirrus Audio card.  Until now I've been using your kernel packages, and have just got round to trying the latest official 4.9 kernel which is also working very nicely.  It's a real shame that the Cirrus card has (or appears to have) gone out of production just when the software support became so good.  The older Wolfson cards are still available, so perhaps I'll get a couple and try modifying them to fit a 40-pin header.  Anyway, thanks again for everything you've done.  I really do appreciate it.

Tim

The Cirrus Card does seem to be out of stock everywhere - what a shame.  I just started working on an alternative ultrasound sensistive microphone, since the electret type I was using has been out of production for years and is now virtually impossible to get hold of.

What other options are there for recording audio at 192kHz with the Raspberry Pi?

The Cirrus Card does seem to be out of stock everywhere - what a shame.  I just started working on an alternative ultrasound sensistive microphone, since the electret type I was using has been out of production for years and is now virtually impossible to get hold of.

What other options are there for recording audio at 192kHz with the Raspberry Pi?

Tim and hysterix,

Do not dispair.

If you look earlier in this thread (2 weeks ago)  Matthias and I were discussing how to adapt the Wolfson variant of this card to run on Pi2/3.  He should take all the credit for verifying that this is a simple matter  ..... indeed trivial from a software perspective.  The hardware modifications are not to everybodys taste and I am still looking into less obtrusive ways of making the adaptation (using a piggy back board, eg)

Trouble is it needs to be double sided, plated through and gold plated on the 8 "thick" pads!  But such a board (with connectors installed) would allow the Wolfson card to plug into a Pi 2/3 with no hardware modifications to either (Hias, if you are reading this, can you confirm that I have not done something stupid like reverse clks or tx/rx?).  For the more intrepid/impatient, the solution which Hias has already posted works perfectly.  For those who do not want to unsolder the 8 pin pogo connector, you can cut-off pin 2 of the connector (which snags the I2C pull up resistor on the Pi) and place a small piece of insulating material beneath the pins for extra safety.  That would make the only unsoldering the removal of one resistor on the Wolfson card.

Hi John. I have a spare Wolfson card, so I might try adapting it for Pi2/3.

It's a shame that the Wolfson card doesn't have Line In micbias, but I could live with that.

Many thanks for the info.

Hi,

Interesting,  I really had not investigated that part of the schematics.  You are right of course  ..... can you not tie the inputs to the bias as the signal leaves the jack socket? Much easier there as you are not working at a microscopic scale.

Let us all know what route you take for the modification  ..... as you rightly note the only way of using the Wolfson device is through this card, so we all need to work on finding the best way forward. Best of luck

Hi John - Is removing R39 on the Wolfson board absolutely essential?

Hystrix

As always Hias is the guru, but (for a direct plug in connection between pi and Wolfson) yes removal of R39 is necessary, otherwise you will be pumping the I2S bit clock directly into the 8804 SPDIF chip CSB pin.

This (and the pogo connector) are the reasons I am looking at a piggy back board .... I do not trust myself on components of that size. The down side of this is board construction and the extra height that is involved in the sandwich as a whole

If you do not want to unsolder R39 and do not want to use a piggy back board, the only alternative is to cut pin 12 out of the 40 way connector on the pi and run a flying lead from the stump of that pin up to pin 3 of the pogo connector on the W (working on cutting one of the tracks on the Wolfson board is in my opinion more difficult than removing R39). If you have no intention of reusing the pi with its expansion header, this is perhaps the easiest modification

Hi John!

The hardware modifications are not to everybodys taste and I am still looking into less obtrusive ways of making the adaptation (using a piggy back board, eg)

Trouble is it needs to be double sided, plated through and gold plated on the 8 "thick" pads! But such a board (with connectors installed) would allow the Wolfson card to plug into a Pi 2/3 with no hardware modifications to either (Hias, if you are reading this, can you confirm that I have not done something stupid like reverse clks or tx/rx?).

Looks like you have pins 3/4 and 5/6 of P5 swapped - the P5 layout (as eg here http://elinux.org/RPi_Low-level_peripherals) is from the bottom side of the board (as opposed to the P1 layout which is from the top side). Pin 1 of P5 is next to the P1 GPIO connector.

Instead of using "hard" gold plated P5 pads you could try ENIG finish - it's cheaper and should be OK (though not as durable as hard gold).

You'd probably need to make the PCB full-size so you can add mounting screws for the Wolfson card and the RPi2/3. Then mechanical stability should be OK.

Other than these issues your approach look fine!

so long,

Hias

Ha! I was just looking at those pogo pin connections myself.  From the photo, it looks like the connections are:

P5 Pin 3 to GPIO Pin 12

P5 Pin 4 to GPIO Pin 35

P5 Pin 5 to GPIO Pin 38

P5 Pin 6 to GPIO Pin 40

And remove R39.

Are those the only chnages to make?  On the piggy-back board, what is that connection on GPIO Pin 33 on the 40 pin header and GPIO Pin 12 on the 26 pin header?

Many thanks :-)

It's a shame that the Wolfson card doesn't have Line In micbias, but I could live with that.

Looking at the schematic it seems you can add it quite easily: remove C8 (it's next to the WM5102), then connect that to 2 2k2 resistors and solder the other end to line in (junction between R55/L6/C33 and R56/L7/C32). Then it'd be like the Cirrus card.

Taking Micbias from the headset in would also be a possibility, just remember that'd you'd need to enable Headset input as well (you could just do a "cirrus-listen headset headset" then micbias on headset in will always be on).

Is removing R39 on the Wolfson board absolutely essential?

Yes. During reset pin 5 of the WM8804 is used as in input to select the I2C address - so you need to have it connected to the pullup. Afterwards it's an output so if you leave it connected to P1 header pin 12 it'd clash with the I2S clock. I haven't found an easier way to cut the connection from GPIO pin 12 to the WM8804 so removing the in-line 0R resistor R39 seems to be the easiest method ATM.

so long,

Hias

Hystrix

See Hias comment, it looks as though I have mirrored the J5 pins, so wait for the revised diagram! Sorry!  But the link you are referring to provides the correct connection back to the 8804.  That is done for completeness, because as Hias says, the pin on the 8804 can be left open circuit without a problem.

The minimum set of connections are just as you say.

As per Hias suggestion, I had thought of making the piggy back card "full size" so that it can offer mechanical support.  At the moment I am thinking of what else to use the card space for.  I need to drive a relay(!) based digital attenuator for the output of the Wolfson, so I will probably put a couple of I2C ports on the card.

Hias,

Once again, thank you for your help

And herewith the revised pcb