Raspberry Pi 400 3,5mm jack

[corsairetc]

Hello all,
It is possible to get 3,5mm jack on Raspberry Pi 400 via GPIO ?
Thank you.

[jamesh]

Yes, but would require some extra electronics.

https://learn.adafruit.com/introducing- ... io-outputs

[PhilE]

From a software perspective you just need to add "dtoverlay=audremap" to config.txt, which will enable the audio outputs on GPIOs 12 & 13 (header pins 32 and 33). "dtoverlay=audremap,pins_18_19" will send the output to GPIOs 18 & 19 (header pins 12 & 35) . But as James says, you really need some external circuitry to turn it into a usable audio signal.

[cleverca22]

ahh, i see how it works
the `compatible = "brcm,bcm2835-audio"` in the device-tree handles both the ALT muxing, and presenting a vchiq service as an alsa device

but i dont see the code in `drivers/staging/vc04_services/bcm2835-audio/bcm2835.c` that is responsible for setting the pins to the chosen set from the gpio block in the dts
and how does the firmware know if it should use pwm0 or pwm1?
i notice a swap_lr in the overlay, but the driver never reads it, is the firmware peeking at the dtb after applying overlays and then acting on that?

[LTolledo]

AFAIK, extracting audio from GPIO using passive circuity does not really give a good high quality audio...

so if I were to venture into doing that...I'd rather:
a. use a USB audio dongle with 3.5mm jack, or
b. make my own I2S audio adapter to be plugged to the 40 pin header
audio will be directed to a 3.5mm jack that I will install (on the I2S adapter)

I still have a few PCM5102A modules with me...

enabling the USB audio dongle or the I2S module are much easier to do...

[cleverca22]

AFAIK, extracting audio from GPIO using passive circuity does not really give a good high quality audio...

except, that is exactly what the 3.5mm jack on every existing pi model is doing

so its more a question of if the old quality is good enough, or if you want to go for something better

[LTolledo]

AFAIK, extracting audio from GPIO using passive circuity does not really give a good high quality audio...

except, that is exactly what the 3.5mm jack on every existing pi model is doing

so its more a question of if the old quality is good enough, or if you want to go for something better

you have a very good point there...everyone has their preferences...
"tinker for good.... or tinker for even better...."

well anyway what will eventually be done will be sort of a "pluggable module" to the GPIO header...
that may be swapped according the the user's use case...

[cleverca22]

AFAIK, extracting audio from GPIO using passive circuity does not really give a good high quality audio...

except, that is exactly what the 3.5mm jack on every existing pi model is doing

so its more a question of if the old quality is good enough, or if you want to go for something better

you have a very good point there...everyone has their preferences...
"tinker for good.... or tinker for even better...."

well anyway what will eventually be done will be sort of a "pluggable module" to the GPIO header...
that may be swapped according the the user's use case...

i'm thinking that a hat (with the overlay eeprom) and a 3.5mm jack would be a pretty easy board to design, and would make upgrades trivial for the end-user

[gordon77]

except, that is exactly what the 3.5mm jack on every existing pi model is doing

so its more a question of if the old quality is good enough, or if you want to go for something better

you have a very good point there...everyone has their preferences...
"tinker for good.... or tinker for even better...."

well anyway what will eventually be done will be sort of a "pluggable module" to the GPIO header...
that may be swapped according the the user's use case...

i'm thinking that a hat (with the overlay eeprom) and a 3.5mm jack would be a pretty easy board to design, and would make upgrades trivial for the end-user

I'd prefer fitting the audio interface internally, soldered to the back of the GPIO connector. Leaving the outside gpio clear.

[cleverca22]

I'd prefer fitting the audio interface internally, soldered to the back of the GPIO connector. Leaving the outside gpio clear.

that could also be done, and id argue that there was already room for that, but the RPT says there wasnt

you just need to make sure the end-user doesnt try to use those pins for something else, and then drill a hole somewhere for the headphone jack

[LTolledo]

those new and upcoming devices may not be called HATs anymore as it wont be on "top" of the board

HATs as defined in raspberry.org.jpg (39.32 KiB) Viewed 4145 times

it might be a bit awkward to call it HAB (as in Hardware Attached to Back)
I wonder how the RPi folks will name those.....

[PiGraham]

I think the simplest, easiest, cheapest pluggable audio connection is a USBaudio card.

Or USB speakers with single USB connection

[Edited to remove info on speakers that, it turns out, require 3.5mm socket]

[thagrol]

i'm thinking that a hat (with the overlay eeprom) and a 3.5mm jack would be a pretty easy board to design, and would make upgrades trivial for the end-user

You mean like the many that already exist?

[trejan]

it might be a bit awkward to call it HAB (as in Hardware Attached to Back)

HATS - Hardware Attached to The Side

[PiGraham]

...beret?

[LTolledo]

it might be a bit awkward to call it HAB (as in Hardware Attached to Back)

HATS - Hardware Attached to The Side

"back side"... yes...
confusion may arise if thought to be as "left side" or "right side"

maybe better wait for any "official announcement" from RPF folks on this matter.....

anyways.... it seems that RPi400 may not be powered via GPIO headers..(read from another forum topic).
may dampen some of my upcoming audio projects with the RPi400..

thus in the meantime... USB audio dongles or USB attached speakers are the only option at the moment
good thing though... my Sound Blaster XiFi Go is available for use....

[W. H. Heydt]

thus in the meantime... USB audio dongles or USB attached speakers are the only option at the moment
good thing though... my Sound Blaster XiFi Go is available for use....

Not *quite* the only options...one could use Bluetooth attached speakers as an option.

[LTolledo]

thus in the meantime... USB audio dongles or USB attached speakers are the only option at the moment
good thing though... my Sound Blaster XiFi Go is available for use....

Not *quite* the only options...one could use Bluetooth attached speakers as an option.

yeah... forgot about that...sorry
being used to physically wired connections only...

[cleverca22]

i'm thinking that a hat (with the overlay eeprom) and a 3.5mm jack would be a pretty easy board to design, and would make upgrades trivial for the end-user

You mean like the many that already exist?

i mean a new shape/design, that fits the shape of the pi400, and wont stick out like an old hat that perfectly fit the pi4b

[thagrol]

i'm thinking that a hat (with the overlay eeprom) and a 3.5mm jack would be a pretty easy board to design, and would make upgrades trivial for the end-user

You mean like the many that already exist?

i mean a new shape/design, that fits the shape of the pi400, and wont stick out like an old hat that perfectly fit the pi4b

I see. Personally I think RPT missed a trick by not rotating the GPIO connector 180 degrees. Existing boards would then plug in vertically. I guess the desire to have the pins in the same orientation as all other models and all those internet diagrams won out.

[hippy]

Personally I think RPT missed a trick by not rotating the GPIO connector 180 degrees. Existing boards would then plug in vertically.

I am not sure how you mean because it seems to me they do. Admittedly the keyboard user would be looking at the underside of a HAT when plugged in rather than the top.

I have to admit the official RPF "You can connect HAT hardware directly to the 40-pin GPIO header, but it will be pointing backwards and down" makes no sense to me. The 400 looks to be using a right-angle header set against a normal Pi. HAT's which covered a Pi would therefore pivot about the header, facing backwards but up.

Maybe I'm missing something.

[thagrol]

Personally I think RPT missed a trick by not rotating the GPIO connector 180 degrees. Existing boards would then plug in vertically.

I am not sure how you mean because it seems to me they do. Admittedly the keyboard user would be looking at the underside of a HAT when plugged in rather than the top.

Is that what you are meaning; put the HAT connector on its top, adjust the 400 header to suit ?

That would work but they'd have to find a connector with some means of preventing existing HAT's being plugged in and I can foresee problems when ribbon cable gets used.

Sorry for the lack of clarity. What I meant was leave it where it is but instead of having physical pin one in the top right have it in the bottom left. HATs would then plug in directly with the board extending upwards not downwards as they would now.

Edit:
Scratch that. I think I've visualised it incorrectly.

[hippy]

Sorry for the lack of clarity. What I meant was leave it where it is but instead of having physical pin one in the top right have it in the bottom left. HATs would then plug in directly with the board extending upwards not downwards as they would now.

I edited my post, but will repeat here as it better facilitates flow ...

I have to admit the official RPF "You can connect HAT hardware directly to the 40-pin GPIO header, but it will be pointing backwards and down" makes no sense to me. The 400 looks to be using a right-angle header set against a normal Pi. HAT's which covered a Pi would therefore pivot about the header, facing backwards but up.

Maybe I'm missing something.

[gordon77]

Sorry for the lack of clarity. What I meant was leave it where it is but instead of having physical pin one in the top right have it in the bottom left. HATs would then plug in directly with the board extending upwards not downwards as they would now.

I edited my post, but will repeat here as it better facilitates flow ...

I have to admit the official RPF "You can connect HAT hardware directly to the 40-pin GPIO header, but it will be pointing backwards and down" makes no sense to me. The 400 looks to be using a right-angle header set against a normal Pi. HAT's which covered a Pi would therefore pivot about the header, facing backwards but up.

Maybe I'm missing something.

They should have removed one pin (an earth) to make a key to prevent the connector being connected the wrong way round.

[cleverca22]

They should have removed one pin (an earth) to make a key to prevent the connector being connected the wrong way round.

its already using the standard 0.1" keyed shroud, where the ribbon cables will only fit in one way
but no past model had that key, so nobody has been making keyed hats