Ok, so the heatermeter rPi setup is fully constructed, inoperative, and my troubleshooting has not illuminated any solutions so I am throwing myself on the mercy of the forums. Here's the deal:
1) I constructed the board late one night. For whatever reason, I left out the "R7-0" strap for the LCD. I know, first step bungled, not a good sign. Moving on. I flashed "openwrt-brcm2708-sdcard-vfat-ext4_224.img" to the SD card via Win32 DiskImager, inserted into the rPi, and connected the rPi to the HM board. I did not have a microUSB cable so the initial power-up was performed with the 12V barrel connection.
2) When I powered up for the first time the top two LED's (red and yellow on my HM board) lit up solid and stayed that way for more than 2 minutes. The LCD backlight did not come on, so I powered down. I rechecked the board, realized that I was missing the R7-0 strap, soldered it in, and powered it back up. Same result.
3) Searching the forums and finding this post on a nearly identical issue, I rechecked each solder connection, reheating and adding solder when necessary. I checked the ATmega328P and made sure that it was properly seated, no pins were bent or not making contact. I also procured a microUSB cable before powering back up. Same result.
4) Concerned that I had fried the rPi before it could flash the ATmega328P, I disconnected it from the heatermeter board, connected it to a monitor & keyboard, and powered up via microUSB. OpenWrt appeared to load, but would not recognize keyboard input. I then downloaded Raspbian wheezy, flashed to SD, and powered that up on the rPi. It loaded normally and I was able to type commands, which seem to put the rPi in the clear.
5) More forum searching resulting in finding the rPi bootloader issue, so I went through the process described to replace the bootcode.bin and the start.elf with the one from wheezy before powering up the rPi again. Same result: it loads to a point, then stops and does not accept input from the keyboard. The forums seem to indicate that this is the normal stopping point, but here is a screenshot of the boot process where it freezes:
Needless to say, pressing "enter" does not have any effect.
6) Unconvinced that the rPi was really ok, I installed rasplex media center and gave it a network cable. It powered up and worked perfectly, showing the plex media library on my pc server and playing 1080p video files without fault. I think it's safe to say the rPi is fine.
My suspicion is that perhaps the OpenWrt software doesn't recognize my wireless USB keyboard. Wheezy and Rasplex being larger projects may have a more extensive peripheral driver library? I don't know. As far as the HM board, I'm stumped. By chance I discovered that with the power on, when my multimeter probes were measuring across the collector and base or base and emitter of the LCD driver (Q2), the LCD backlight turns on. Whisky. Tango. Foxtrot. I'm guessing I've got a bad Q2, but that only addresses display issue.
Is there some sort of bat-signal that I can shine in the night sky that will summon Bryan Mayland to my house? Because that would be helpful.
1) I constructed the board late one night. For whatever reason, I left out the "R7-0" strap for the LCD. I know, first step bungled, not a good sign. Moving on. I flashed "openwrt-brcm2708-sdcard-vfat-ext4_224.img" to the SD card via Win32 DiskImager, inserted into the rPi, and connected the rPi to the HM board. I did not have a microUSB cable so the initial power-up was performed with the 12V barrel connection.
2) When I powered up for the first time the top two LED's (red and yellow on my HM board) lit up solid and stayed that way for more than 2 minutes. The LCD backlight did not come on, so I powered down. I rechecked the board, realized that I was missing the R7-0 strap, soldered it in, and powered it back up. Same result.
3) Searching the forums and finding this post on a nearly identical issue, I rechecked each solder connection, reheating and adding solder when necessary. I checked the ATmega328P and made sure that it was properly seated, no pins were bent or not making contact. I also procured a microUSB cable before powering back up. Same result.
4) Concerned that I had fried the rPi before it could flash the ATmega328P, I disconnected it from the heatermeter board, connected it to a monitor & keyboard, and powered up via microUSB. OpenWrt appeared to load, but would not recognize keyboard input. I then downloaded Raspbian wheezy, flashed to SD, and powered that up on the rPi. It loaded normally and I was able to type commands, which seem to put the rPi in the clear.
5) More forum searching resulting in finding the rPi bootloader issue, so I went through the process described to replace the bootcode.bin and the start.elf with the one from wheezy before powering up the rPi again. Same result: it loads to a point, then stops and does not accept input from the keyboard. The forums seem to indicate that this is the normal stopping point, but here is a screenshot of the boot process where it freezes:
Needless to say, pressing "enter" does not have any effect.
6) Unconvinced that the rPi was really ok, I installed rasplex media center and gave it a network cable. It powered up and worked perfectly, showing the plex media library on my pc server and playing 1080p video files without fault. I think it's safe to say the rPi is fine.
My suspicion is that perhaps the OpenWrt software doesn't recognize my wireless USB keyboard. Wheezy and Rasplex being larger projects may have a more extensive peripheral driver library? I don't know. As far as the HM board, I'm stumped. By chance I discovered that with the power on, when my multimeter probes were measuring across the collector and base or base and emitter of the LCD driver (Q2), the LCD backlight turns on. Whisky. Tango. Foxtrot. I'm guessing I've got a bad Q2, but that only addresses display issue.
Is there some sort of bat-signal that I can shine in the night sky that will summon Bryan Mayland to my house? Because that would be helpful.
Last edited:
