StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Wed Jun 09, 2021 3:41 pm

Also - I am still having trouble getting my CM4 to boot. I'd like to finish this project as soon as I can, so I'd really appreciate if anyone would take the time to look at my schematics & board files to see where I'm going wrong.

I've made it up to Step 10 of the "A. Troubleshooting Section" = nACT_LED does nothing. I can't see anything wrong with my SD lines or traces. I've flashed the latest version of RetroPie onto my SD card using balenaEtcher and also the Official Raspberry Pi flashing tool. Does anyone notice anything in particular with the pinouts, schematics or footprints of the CM4? Maybe someone might be able to help me not throw in the towel for this project...

I've attached the schematics in PDF and board/sch files for EAGLE.
I'd really appreciate it if anyone could take a look! Really getting sick of debugging...

Thanks everyone.

PDF Schematic
https://drive.google.com/file/d/1zJb33g ... sp=sharing

Eagle BRD File
https://drive.google.com/file/d/13-FxUa ... sp=sharing

Eagle SCH FIle
https://drive.google.com/file/d/11BpjzE ... sp=sharing

Flavor
Posts: 5
Joined: Fri Jun 02, 2017 5:38 pm

Re: Retro Lite CM4: Handheld Gaming Console

Wed Jun 09, 2021 5:45 pm

We gave it a quick look. Here are some notes about stuff that could cause it to not boot.
Screen Shot 2021-06-09 at 11.39.58 AM.jpeg
Screen Shot 2021-06-09 at 11.39.58 AM.jpeg (141.56 KiB) Viewed 1503 times

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Thu Jun 10, 2021 3:56 am

Flavor wrote:
Wed Jun 09, 2021 5:45 pm
We gave it a quick look. Here are some notes about stuff that could cause it to not boot.

Screen Shot 2021-06-09 at 11.39.58 AM.jpeg
Hi, thanks for your reply back.

As mentioned above I have checked all of the voltages on all pins required for boot (RUN_PG, GLOBAL_EN etc... I measure 3.3v and 4.93v on those pins). All voltage rails rise to 3.29v and 1.78v, which shows that the internal PMIC is doing its job. I measure a solid 5V on the Pi input pins. The SoC heats up.

You make a good point regarding the ATtiny. Am I correct in assuming that the CM4 will not boot if power is applied to any pin before 5v is applied? I may have not set pins 3 and 5 on the ATtiny to inputs in my code, rather they may be currently floating. GPIO16_VOLTAGE_SHUTDOWN sends a GPIO high to the ATtiny to be read as input to shut down system when voltage of batteries reaches 3.3v and GPIO25_PI_SHUTDOWN_INIT drives a GPIO low on the Pi to shut down the operating system safely. If these pins are high, would this prevent boot/potentially killed GPIO banks if they go above 3.3v?

The other issues that I thought of:

1) SD lines are too long/wrong pinout - unlikely as the SD is a super slow bus at like 100Mhz
2) Pinout for CM4 is wrong entirely :P
3) I’m back feeding the CM4 with 3v3 regulator somewhere
4) Capacitance on 5V lines isn’t enough, causing instability issues. Increase output caps to be 15v-26v tolerant. As ceramic caps need to be rated well over the actual bias voltage since they derate so much, use higher voltage rated caps. Not enough output capacitance is making it difficult for the regulator to supply transient loads. The Pi switches on and off quickly, so it needs a big charge reservoir in the form of both bulk caps at the input and decoupling caps at power supply pins (right now I'm using small 0603 10V 1.0uF caps x 3 for the output, 0603 10V 1.0uF cap x 1 for the input).
5) GPIO pins from attiny causing issues (high before power up)
6) I screwed up the nACT_LED - pulled it low to GND instead of high to 3.3v. I have fixed this by soldering one end of a resistor now to 3v3 (still no activity on the green LED at all - no flashing or no signs of life. LED is not dead as checked with multimeter)

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Fri Jun 11, 2021 1:10 am

Update - I do get a flash on the ACT led now (only a single flash followed by solid green...)

Flavor
Posts: 5
Joined: Fri Jun 02, 2017 5:38 pm

Re: Retro Lite CM4: Handheld Gaming Console

Fri Jun 11, 2021 4:01 pm

The CM4 has a bootloader screen that should come up even if SD isn't working or whatever.

I might focus on trying to get that screen up and not worry about the SD card stuff at the moment.

You are using the LITE module, right?

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Sat Jul 03, 2021 4:12 am

Hi everyone,

I’ve taken a scope to the CM4 as I still cannot get the module booting. I’ve tried both a brand new CM4 and an older one and both are having the same issues on the n_RPIBOOT pin, where the voltage oscillates between ~1V to 3.3V. Every other rail and pin seems to rise in proper order. Is this normal behavior for this pin or is something fundamentally wrong?

I’ve attached some pictures below of the power sequencing on the rails at startup. I’d appreciate if someone (preferably a Pi engineer) would be able to help me spot the issue/advise where to go from this point for debugging. Cheers.

————————————————
Image
dark blue - VSYS
yellow - 5V
light blue - on-board 3.3V
purple - CM4 3.3V output
————————————————
Image
dark blue - GLOBAL_EN
yellow - 5V
light blue - on-board 3.3V
purple - CM4 3.3V output
————————————————
Image
Purple - 3V3 from CM4.
Light blue - RUN_PG.
Dark blue - GLOBAL_EN.
Yellow - nEXTRST (pin 100 of CM4).
————————————————
Image
Purple - 3V3 from CM4.
Light blue - RUN_PG.
Dark blue - nRPIBOOT
Yellow - nEXTRST (pin 100 of CM4).
Last edited by StonedEdge on Mon Jul 05, 2021 1:00 am, edited 1 time in total.

dp11
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 622
Joined: Thu Dec 29, 2011 5:46 pm

Re: Retro Lite CM4: Handheld Gaming Console

Sat Jul 03, 2021 6:50 am

That all looks fine to me. nRPI_BOOT is a complex signal.

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Sat Jul 03, 2021 7:44 am

dp11 wrote:
Sat Jul 03, 2021 6:50 am
That all looks fine to me. nRPI_BOOT is a complex signal.
OK, that is a relief then... Any idea were I should go from here for debugging? Hook up UART/check for backfeeding? The nACT_LED flashes one and then stays a solid green. I've tried two CM4s with a fresh image of RetroPie on the SD card. I thought the expected behaiour for the LED when no SD card is inserted is to flash 4 times. Maybe the bootloader is corrupted?

dp11
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 622
Joined: Thu Dec 29, 2011 5:46 pm

Re: Retro Lite CM4: Handheld Gaming Console

Sat Jul 03, 2021 8:23 am

I've never heard of the bootloader getting corrupted. Time to probe some of the SDCARD signals maybe ?

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Sun Jul 04, 2021 4:11 pm

dp11 wrote:
Sat Jul 03, 2021 8:23 am
I've never heard of the bootloader getting corrupted. Time to probe some of the SDCARD signals maybe ?
I’ve checked the SD card signals and there is no activity on them which suggests the Pi isn’t booting up.

I did have a thought - that being that the external 3v3 PSU on my PCB is backfeeding the CM4 somehow. The external regulator providing 3v3 to the TFP401 decoder and USB hub chip does power up first before the CM4 3v3 rail does. I have read somewhere that the CM4 won’t boot if it has any voltage applied on any of its pins before it’s power sequencing has finished.

dp11
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 622
Joined: Thu Dec 29, 2011 5:46 pm

Re: Retro Lite CM4: Handheld Gaming Console

Sun Jul 04, 2021 5:47 pm

I don't think back feeding 3.3v is causing your problem as it typically prevents RUN_PG going high. Can you probe I2C bus the EDID prom ? I assume the prom is currently blank ?

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Sun Jul 04, 2021 11:41 pm

dp11 wrote:
Sun Jul 04, 2021 5:47 pm
I don't think back feeding 3.3v is causing your problem as it typically prevents RUN_PG going high. Can you probe I2C bus the EDID prom ? I assume the prom is currently blank ?
The EEPROM is flashed currently and connected to HDMI0_SCL HDMI0_SDA lines. I can probe them to see what’s happening. I'm not using an official Raspberry Pi IPS display, just a generic TFT RGB display from Eyoyo (5").

See here:
https://learn.adafruit.com/adafruit-tfp ... g-the-edid

aBUGSworstnightmare
Posts: 3454
Joined: Tue Jun 30, 2015 1:35 pm

Re: Retro Lite CM4: Handheld Gaming Console

Mon Jul 05, 2021 6:16 am

StonedEdge wrote:
Sun Jul 04, 2021 11:41 pm
dp11 wrote:
Sun Jul 04, 2021 5:47 pm
I don't think back feeding 3.3v is causing your problem as it typically prevents RUN_PG going high. Can you probe I2C bus the EDID prom ? I assume the prom is currently blank ?
The EEPROM is flashed currently and connected to HDMI0_SCL HDMI0_SDA lines. I can probe them to see what’s happening. I'm not using an official Raspberry Pi IPS display, just a generic TFT RGB display from Eyoyo (5").

See here:
https://learn.adafruit.com/adafruit-tfp ... g-the-edid
Your TFP401 has no idea on the connected displays resolution, so you should be able to rule out an EDID EEPROM content issue by adding below to your config.txt and forcing the required resolution (with hdmi_ground and hdm_mode and, in case you're needing a custom timing hdmi_timings).

Code: Select all

hdmi_force_hotplug=1
hdmi_ignore_edid=0xa5000080
hdmi_drive=2
This should help you to figure out if it's EDID content related. Hopefully you have pull-ups on HDMI0_SCL HDMI0_SDA lines; not having them delays your boot process.

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Mon Jul 05, 2021 6:33 am

aBUGSworstnightmare wrote:
Mon Jul 05, 2021 6:16 am
StonedEdge wrote:
Sun Jul 04, 2021 11:41 pm
dp11 wrote:
Sun Jul 04, 2021 5:47 pm
I don't think back feeding 3.3v is causing your problem as it typically prevents RUN_PG going high. Can you probe I2C bus the EDID prom ? I assume the prom is currently blank ?
The EEPROM is flashed currently and connected to HDMI0_SCL HDMI0_SDA lines. I can probe them to see what’s happening. I'm not using an official Raspberry Pi IPS display, just a generic TFT RGB display from Eyoyo (5").

See here:
https://learn.adafruit.com/adafruit-tfp ... g-the-edid
Your TFP401 has no idea on the connected displays resolution, so you should be able to rule out an EDID EEPROM content issue by adding below to your config.txt and forcing the required resolution (with hdmi_ground and hdm_mode and, in case you're needing a custom timing hdmi_timings).

Code: Select all

hdmi_force_hotplug=1
hdmi_ignore_edid=0xa5000080
hdmi_drive=2
This should help you to figure out if it's EDID content related. Hopefully you have pull-ups on HDMI0_SCL HDMI0_SDA lines; not having them delays your boot process.

Thanks for your reply. I'll give that a shot later tonight when I get home. I do have some 10k pull-ups on those lines to 5V. I thought I didn't really need them though as the datasheet specifies that there is already 1.8k pull-ups on those pins.

aBUGSworstnightmare
Posts: 3454
Joined: Tue Jun 30, 2015 1:35 pm

Re: Retro Lite CM4: Handheld Gaming Console

Mon Jul 05, 2021 7:08 am

StonedEdge wrote:
Mon Jul 05, 2021 6:33 am
aBUGSworstnightmare wrote:
Mon Jul 05, 2021 6:16 am
StonedEdge wrote:
Sun Jul 04, 2021 11:41 pm


The EEPROM is flashed currently and connected to HDMI0_SCL HDMI0_SDA lines. I can probe them to see what’s happening. I'm not using an official Raspberry Pi IPS display, just a generic TFT RGB display from Eyoyo (5").

See here:
https://learn.adafruit.com/adafruit-tfp ... g-the-edid
Your TFP401 has no idea on the connected displays resolution, so you should be able to rule out an EDID EEPROM content issue by adding below to your config.txt and forcing the required resolution (with hdmi_ground and hdm_mode and, in case you're needing a custom timing hdmi_timings).

Code: Select all

hdmi_force_hotplug=1
hdmi_ignore_edid=0xa5000080
hdmi_drive=2
This should help you to figure out if it's EDID content related. Hopefully you have pull-ups on HDMI0_SCL HDMI0_SDA lines; not having them delays your boot process.

Thanks for your reply. I'll give that a shot later tonight when I get home. I do have some 10k pull-ups on those lines to 5V. I thought I didn't really need them though as the datasheet specifies that there is already 1.8k pull-ups on those pins.
o.k., not designed a baseboard for CM4 yet, only CM3.
Yes, there are 1k8 pull-ups, so I think your 10k is too much.

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Tue Jul 06, 2021 8:03 am

Posting the rise time of the external PSU's I'm using on my PCB for completeness.

Yellow is 5v, light blue is 3V3 and blue is VSYS (system power rail input for the regulators).
I also did add these lines to config.txt on an empty PCB with no external 3v3 PSU powered components, and the activity LED does appear to be flashing. I don't know how to check if the OS is succesfully booting though, as I was certain the HDMI1 output does not work if HDMI0 is not connected on the Pi 4. HDMI0 is connected to the TFP driver (internal screen) and HDMI1 is connected to the HDMI external port.

dtoverlay=dwc2,dr_mode=host
gpu_mem_256=128
gpu_mem_512=256
gpu_mem_1024=256
force_turbo=1
h264_freq=333

Video showing activity LED flashing on first bootup
https://drive.google.com/file/d/1dBH4OR ... YSM52/view

Image

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Thu Jul 08, 2021 8:48 am

I'm sorry for spamming my own thread, but I finally got the CM4 booting, exciting times! (as verified by pinging the IP address for raspberry pi and getting a response on command prompt... hooray, it finally works).

However, after moving the CM4 back to the more populated board with all the peripheral components, it went back to its usual ways without booting.

It definitely has to be something on the 5v rail, as I removed the external 3v3 PSU and it still wouldn't boot. Attaching photos of my schematics of everything connected to the CM4 below. I still do believe it is the ATtiny causing problems (despite the code having programmed these pins to High-Z logic) but maybe someone can spot something... Plan is to start with the empty board and build up the components that way, as opposed to building down from the assembled board. I think that will make the issue easier to spot.

I've attached the schematics below of what I think may be causing the issues.

ATTiny pins 5 and 3 are directly connected to the GPIOs of the CM4. Pin 5 is an input to the ATtiny to tell it to shutdown, Pin 3 is an output of the ATtiny to communicate with the Pi. I have programmed both of these to High-Z for testing purposes, but this is what I think is probably causing the issues, since it is powered before the Pi has a chance to boot up. There could be a small amount of leakage current on these pins preventing a boot.

Image

Fan for PWM control based on temperature of the SoC. Shouldn't be causing any boot issues.

Image

Amplifier for audio over i2s. I don't think this will be causing issues but have put here as it is connected to both 5V and 3v3 (from CM4). It is an i2s amplifier (I have not activated i2s audio in config.txt yet).

Image

MAX17055 fuel gauge is connected to the i2c lines of the Pi (again, which have not yet been activated in config.txt). The interrupt pin (ALRT) unfortunately I forgot to use a pull-up on my previous revision, so it is directly connected to the CM4 at the moment (I have not programmed in any pull-up on that GPIO yet).

Image

Backlight boost converter, connected to the CTRL pin via Pi GPIO. I don't think this is causing any boot problems.

Image

HDMI I2C EEPROM for setting output resolution to the internal screen. I have some pull-ups connected to the 5V rail but I do believe that the CM4 has internal pull-ups on the i2c lines, so these may not be necessary.

Image

Keen to here what you guys think before I start building a new board from the ground up!
Cheers,

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Sun Jul 25, 2021 3:56 pm

Hi guys. I’ve finally got the CM4 consistently booting up now, but having issues with HDMI1 video output to my Japanese TV. I have tried with the compute module IO board and successfully got HDMI out of HDMI1 (HDMI0 didn’t seem to work though which was weird) so I’m assuming that it’s either my routing or soldering of something that’s causing my issue for video output on HDMI1.

Is there anything specific I need to add in config.txt to get video output on HDMI1? I keep seeing rumors everywhere that HDMI1 doesn’t work unless HDMI0 is enabled etc, but that sounds quite strange and I don’t think that’s my issue.

As for the length matching, it’s not very great (I realized that I had the maximum length set between pairs at max 10mm in DRC) which I think is quite large, but most of the lengths thankfully within each TMDS pair is within 0.15mm (max 0.2mm within, max 2.3mm between) so I don’t think that is my problem. Would anyone be able to help? I’ve shot for 3mm or less of mismatch between all of the pairs. It was my first time routing HDMI so I don’t expect it to be fantastic. Just wanted to rule out any software issues before going back to look at the hardware. I’m using a HDMI mini port and receptacle (I.e type-C).

Uploading pictures below for reference. One of the length matching (I.e length of traces), the other shows the actual routing).

https://i.imgur.com/4o9QeFG.png
https://i.imgur.com/Morq1rB.png

aBUGSworstnightmare
Posts: 3454
Joined: Tue Jun 30, 2015 1:35 pm

Re: Retro Lite CM4: Handheld Gaming Console

Sun Jul 25, 2021 4:56 pm

StonedEdge wrote:
Sun Jul 25, 2021 3:56 pm
Hi guys. I’ve finally got the CM4 consistently booting up now, but having issues with HDMI1 video output to my Japanese TV. I have tried with the compute module IO board and successfully got HDMI out of HDMI1 (HDMI0 didn’t seem to work though which was weird) so I’m assuming that it’s either my routing or soldering of something that’s causing my issue for video output on HDMI1.

Is there anything specific I need to add in config.txt to get video output on HDMI1? I keep seeing rumors everywhere that HDMI1 doesn’t work unless HDMI0 is enabled etc, but that sounds quite strange and I don’t think that’s my issue.

As for the length matching, it’s not very great (I realized that I had the maximum length set between pairs at max 10mm in DRC) which I think is quite large, but most of the lengths thankfully within each TMDS pair is within 0.15mm (max 0.2mm within, max 2.3mm between) so I don’t think that is my problem. Would anyone be able to help? I’ve shot for 3mm or less of mismatch between all of the pairs. It was my first time routing HDMI so I don’t expect it to be fantastic. Just wanted to rule out any software issues before going back to look at the hardware. I’m using a HDMI mini port and receptacle (I.e type-C).

Uploading pictures below for reference. One of the length matching (I.e length of traces), the other shows the actual routing).

https://i.imgur.com/4o9QeFG.png
https://i.imgur.com/Morq1rB.png
HDMI lenght matching should be fine up ro 25mm, for DSI/CSI you should aim for as close as possible, considering 10mm as max. (Has been discussed here viewtopic.php?f=98&t=311536).

Which kernel are you testing with? There seem to be some issues with HDMi atm; lookat some posts here in the forum; viewtopic.php?f=29&p=1893068#p1893005, but that one is with 2 displays connected.

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Fri Jul 30, 2021 2:37 am

Hi all,

I’ve got my internal screen working on the custom board which is great! It’s just using HDMI with the TFP401 RGB888 converter IC, pretty basic stuff. I’m having a little bit of an issue now with my WM8960 amplifier circuit and the custom drivers. I’m on RetroPie, kernel version 5.10.52-v71+, cloning the repository below:

https://github.com/waveshare/WM8960-Audio-HAT

It installs fine after updating the latest kernel drivers and headers. I was getting the below issue, but it seems that has been resolved by updating the Raspbian packages via sudo apt-get update; sudo apt-get dist-upgrade and sudo apt-get install raspberrypi-kernel-headers.

Your kernel headers for kernel 5.4.72-v7l+ cannot be found at
/lib/modules/5.4.72-v7l+/build or /lib/modules/5.4.72-v7l+/source.

Here’s what the terminal is showing me now after executing the below commands:

Installing driver:
Image

systemctl status wm8960-soundcard:
Image

dmesg | grep wm8960:
Image

lsmod | grep wm8960
Image

As for hardware, I have the i2c lines connected to i2c0 on the CM4 (pins 80 and 82), with the below in my config.txt file to activate i2c:
dtparam=i2c_arm=on
dtparam=i2s=on
#dtparam=spi=on
dtoverlay=i2c0,pins_44_45=1

Does anyone see any problems with the above? It could be my soldering as usual, but SDA/SCL lines are pulled up internally to 3v3 and are rising. I haven't ran sudo i2cdetect -y 1 but will report back later on those results (update: unfortunately not getting any slave address pings).

aBUGSworstnightmare
Posts: 3454
Joined: Tue Jun 30, 2015 1:35 pm

Re: Retro Lite CM4: Handheld Gaming Console

Fri Jul 30, 2021 5:08 am

StonedEdge wrote:
Fri Jul 30, 2021 2:37 am
...
As for hardware, I have the i2c lines connected to i2c0 on the CM4 (pins 80 and 82), with the below in my config.txt file to activate i2c:
dtparam=i2c_arm=on
dtparam=i2s=on
#dtparam=spi=on
dtoverlay=i2c0,pins_44_45=1

Does anyone see any problems with the above? It could be my soldering as usual, but SDA/SCL lines are pulled up internally to 3v3 and are rising. I haven't ran sudo i2cdetect -y 1 but will report back later on those results (update: unfortunately not getting any slave address pings).
and what happens on I2C0 after fixing your type? Overlay should be changed to

Code: Select all

dtoverlay=i2c0,pins_44_45   
as ther is no '=1' value for it.

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Fri Jul 30, 2021 8:10 am

aBUGSworstnightmare wrote:
Fri Jul 30, 2021 5:08 am
StonedEdge wrote:
Fri Jul 30, 2021 2:37 am
...
As for hardware, I have the i2c lines connected to i2c0 on the CM4 (pins 80 and 82), with the below in my config.txt file to activate i2c:
dtparam=i2c_arm=on
dtparam=i2s=on
#dtparam=spi=on
dtoverlay=i2c0,pins_44_45=1

Does anyone see any problems with the above? It could be my soldering as usual, but SDA/SCL lines are pulled up internally to 3v3 and are rising. I haven't ran sudo i2cdetect -y 1 but will report back later on those results (update: unfortunately not getting any slave address pings).
and what happens on I2C0 after fixing your type? Overlay should be changed to

Code: Select all

dtoverlay=i2c0,pins_44_45   
as ther is no '=1' value for it.
Hey turns out that worked.
Also, the issue was that sudo i2cdetect -y 1 isnt correct for i2c1, rather you need to do -y 0 :cry:

I still can't see the device yet though when trying to do sudo aplay -l

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Fri Jul 30, 2021 10:13 am

Maybe I need to recompile the code after editing the .dts file? Looks like it’s pointing to i2c1 instead of i2c0.

https://github.com/waveshare/WM8960-Aud ... ndcard.dts

aBUGSworstnightmare
Posts: 3454
Joined: Tue Jun 30, 2015 1:35 pm

Re: Retro Lite CM4: Handheld Gaming Console

Fri Jul 30, 2021 10:16 am

StonedEdge wrote:
Fri Jul 30, 2021 10:13 am
Maybe I need to recompile the code after editing the .dts file? Looks like it’s pointing to i2c1 instead of i2c0.

https://github.com/waveshare/WM8960-Aud ... ndcard.dts
yes, target is I2C1.

Where will it be exposed to, kernel or users-ace? If grabbed by the kernel you wills 'UU' in the i2cdetect output.

StonedEdge
Posts: 81
Joined: Wed Oct 28, 2020 11:42 am

Re: Retro Lite CM4: Handheld Gaming Console

Fri Jul 30, 2021 11:43 am

aBUGSworstnightmare wrote:
Fri Jul 30, 2021 10:16 am
StonedEdge wrote:
Fri Jul 30, 2021 10:13 am
Maybe I need to recompile the code after editing the .dts file? Looks like it’s pointing to i2c1 instead of i2c0.

https://github.com/waveshare/WM8960-Aud ... ndcard.dts
yes, target is I2C1.

Where will it be exposed to, kernel or users-ace? If grabbed by the kernel you wills 'UU' in the i2cdetect output.
I'm seeing the hexadecimal value, not UU. I still can't get it to appear after changing target to i2c0 (i.e edit on WinSCP and then re-save, reboot). What's the best way to update the code to reflect the change in the .dts file?

Return to “Compute Module”