Are we sure its not booting up? Maybe the display is just garbled? I’m not sure that helps but I’m just cuprous as to whether it actually fails to boot or not. You could try hooking up an HDMI display.
It does fail to boot. I left it go for an hour on mine and could not SSH into it. Something is different between the Pi 3b and Pi 4. My Pi 3b has been running the fan shim and 3.5 display together for about a month now with no issues. The same parts I gave up trying to get to work on my Pi 4.
Ok, it is in fact not booting up. I have no idea why though?
@Silarous I am going to need your help to get the current state of those pins on the PI3
What I am trying to achieve with this is to know the difference between the Pi4 and Pi3 as far as GPIO configuration and current state.
Please install only the Fan Shim on your PI3 and once is fully online open a terminal and run the following commands:
-
sudo raspi-gpio get 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40" > fan_full_gpio.log
-
watch -d "sudo raspi-gpio get 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40"
The previous command will highlight if there is any change on any field. for instance if one of the levels change. Most likely you will see changes in some levels. Please let it run for 2-3 min and Please write down the gpios that change.
-
Please repeat steps 1 and 2 without the fan and Only with the LCD and rename the output file to lcd_full_gpio.log instead of fan_full_gpio.log
-
Please repeat steps 1 and 2 with the Fan and LCD, and rename the output file to lcdfan_full_gpio.log instead of fan_full_gpio.log
I am doing the same with the PI 4 and I will compare my finding and try to findout if something can be done at the software level.
Cheers
Not a problem. I will try and get that done later tonight!
With Fan Shim, no Display - GPIO 34
With Display, no Fan Shim - GPIO 34
Both situations only have pin 34 that highlights.
could you share your fan_full_gpio.log and lcdfan_full_gpio.log
Sure if you tell me how to find and display a log.
-
fan only
sudo raspi-gpio get 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40" > fan_full_gpio.log -
lcd only
sudo raspi-gpio get 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40" > lcd_full_gpio.log -
fan and lcd
sudo raspi-gpio get 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40" > lcdfan_full_gpio.log
when you run these commands you will endup with 3 files : fan_full_gpio.log, lcd_full_gpio.log, lcdfan_full_gpio.log.
They will be in whatever directory you run this commamd
How do you display the log? I know nothing when it comes to Linux. I can follow a guide though. Have to be detailed for me unfortunately.
you can just run the following commands:
- cat fan_full_gpio.log
- cat lcd_full_gpio.log
- cat lcdfan_full_gpio.log
cat fan_full_gpio.log comes back “Too many arguments”
I am having the exact same issue currently using a Pi 4. Trying to work it out but not having too much luck. I have reinstalled the fan shim software as per in this post in GitHub https://github.com/rootzoll/raspiblitz/issues/806, but to no avail!
I am just setting up the raspiblitz tonight hopefully as I’ve managed to ssh in with the LCD removed. Let me know if I can be of any help debugging this for future users, I’m not the best technically but willing!
Here’s a great solution.
- Drill 2 tiny holes into the fan connector, on its side, at the very bottom on the connector. (The holes are the width of the pins and drill them all the way thru the connector)
- Insert the fan connector “sideways” onto pins 2 and 3, it will stick out instead of up, but it will make the electrical connection and it only takes up a few mm of space when mounted sideways.
- Add the LCD on top, lots of pin real-estate is left.
Works! :-)
So having the fan always on is hardly a solution as we can get fan only with a 5V connector for 10x cheaper.
wondering if anyone having solved their shim fan issue with sudo apt-get purge plymouth*
has had any further problems? Things like the sudo reboot
or sudo shutdown now
not working or else?
It might just affect or latest build: #885
Also did anyone have luck with changing the shimfan to a never one? I wonder if this is issue is just a manufacturing error or a wrong update to their PCB?
Removing plymouth might have got through the booting problem but unfortunately not catching the problem early can lead to weird issues later on.
I have tested with a GPIO Expansion board. Physically separating the connection of the shim fan from the LCD did not help my faulty one. Also unlikely to be an EM interference as distance form the board did not make a difference either.
I was wondering if anyone is still having issues with this now that the software includes disabling the button at the software level?
I looked through the Python code for the fanshim and it looks like pin 11 is set to a pill up/on state, so the button nacts as a sink/ground when pressed. Since the 3 5" LCD touchscreens all seem to do the same, it shouldn’t be an issue. No other pins are shared besides for power and grounding.
That leads me to ask next – did this happen only with both installed at the software level, or does this happen only if both devices are physically connected to the PI? Conversely, does it happen with only one installed? What type of 5V power supply are you using – what amperage? The touchscreen is quite power hungry, I’d be surprised if the fan is hhngry enough to put it over the edge (don’t see it in documentation). But Waveshare suggests a 5V/3A supply for their 3.5" LCD, and most of these 3.5" screens are essentially the same product being resold by different supplies, so it’s likely good advice for all of them.
If this was the case it would seem odd that you don’t see this coming up more in discussions of the LCD. Surely some people would be bound to have any fan alongside the LCD and some of them would use a low amp supply without realizing and end up complaining about the screen preventing booting. That makes power seem unlikely too.