I'd imagine a lot of users would encounter problems of some kind. Ne1 have a checklist and troubleshooting guide?
Here is what I'm lost at, aside from all the troubleshooting that was clear:
1. Wire color codes, my sensor(s) are cased in a long can and have a 1m cable. Lots of photos online have all sorts of colors, I've gone with green and red being ground and power with yellow being data. Perhaps it would work to list the connectors using terms like "lightest, darkest, and middle", thus allowing for any colors to be used. That's assuming lightest and darkest are always data and ground... and that everyone agrees green is darker than red.
2 This page menchions board rev having an influence, however for this project both revision should be wired the same... GPIO4 is P1-7? My rev is 000e, but I couldn't find a pinout for it anywhere. If there is multiple setups, they both should be listed explicitly to provide a complete reference and no mistake.
3. The use of "scripts/config --enable DYNAMIC_DEBUG" kernel build command. To turn on w1 debugging, try:
Code: Select all
#!/bin/sh
[ "`zgrep DYNAMIC_DEBUG /proc/config.gz`" ] || exit 4
f=/sys/kernel/debug/dynamic_debug/control
[ -f $f ] ||
mount -t debugfs none /sys/kernel/debug
modprobe w1_gpio
grep w1 $f|wc
grep w1.*" =p " $f|wc
echo 'file w1.c =p' > $f
echo 'file w1_netlink.c =p' > $f
grep w1.*" =p " $f|wc
modprobe w1_therm
Loading your modules this way with a DYNAMIC_DEBUG kernel will show you more about what's going on in dmesg output. For me it just indicates that I've no w1_therm supported devices on a perfectly good w1 bus. I've not seen any devices detected on the bus, but without kcore support that's difficult to say.
4. As this page suggests either 5v or 3.3v will work, does the resistor value need to change? I just read that the GPIO pins should be able to self pullup, would that be applicable here? I'm using 1.5k and get a voltage of 2.73 across the resistor and 0.586 to ground or 3.316.
The above is a list of topics that I've ruled out in my case, given that I still have the same issue the above must be wrong some how. Either I've made a mistake or my troubleshooting tips guide are incomplete.
My next step is to use an oscilloscope, to make sure that Pi and sensor are both xmitting. If they are then I'd have to assume that the Pi is not able to measure the input, 5v burnout.