I'd like to generate a feed for some custom packages I'm using on the omega2+, like Onion has done for http://repo.onioniot.com/omega2/packages/onion.
How is this custom feed generated?
The usb option could work but I'm interested in pcb integrated options for mechanical and cost savings. Something that could live on other busses like SPI, I2C etc but that would have integrated support in the kernel through bluez.
Onion publishes https://hub.docker.com/r/onion/omega2-source but this container hasn't been updated for several months, it looks like prior to the upgrade to LEDE 18.
Any plans to publish a newer container? I'd rebuild the container here but I can't find the originating Dockerfile so maybe it was built manually??
@zedf yep, building on windows works but it sounds like Linux and macOS it doesn’t. I thought docker was supposed to be the same across all OSes? There must be some differences. Would be great if it worked on mac and Linux.
I was able to build the image, I think with the docker container on OS X after disabling mosquitto or something like that. At least I’m almost certain I could build a full image...
Same SD card works fine on another Omega2+ so I'm less likely to think its the card. Supply voltage looks good as well, 3.45v (need to adjust regulator resistors to bring that down to 3.3v) That's a good path to check and more data sheet driven then my approach of swapping the cards and performing a functional check.
I saw that diagram with those pinouts as well but isn't EPHY_APGIO_AIO_EN set in software and not affected by the boot pins?
After a few hours of experimenting here the issue turned out to be a relatively strong pulldown on GPIO45. This was affecting a boot pin. Looking at http://plan44.ch/downloads/Onion-Omega2-Pinout-annotated.png and the MT7688 data sheet though it isn't clear exactly what the issue is. Measuring the module itself there is a 4.7k pullup to 3.3V on that pin, so no need for external strapping. Pulling that pin down would appear to be enabling JTAG mode, page 33 of the MT7688 data sheet fyi.
JTAG pins look entirely separate from SD interface pins though so maybe I'm close on the bootstrap pin issue but off on the effect? I still can't figure out why everything would boot fine if that GPIO45 was strapped low (overriding the 4.7k pullup) but just the SD card wouldn't be working......
Looking at the Omega2S-reference schematic it looks like the SD card pins are entirely separate pins but its tough to know for certain without the Omega2 module schematic I think.
I'm using a lot of header pins (not the SPI ones related to the internal memory) but if none are related to the microSD then I'm left with an issue with this particular Omega2+ module, a firmware difference (both are upgraded to latest, I upgraded the working module once it worked thinking maybe it would break after the upgrade), or some pin configuration that isn't documented that I'm missing.... any ideas?
The symptom I'm seeing is the kernel driver is reporting that there is no support for the card's volts. This same card works fine in my other Omega2 (not mounted on my custom PCB).
4.901374] MTK MSDC device init. [ 4.944622] mtk-sd: MediaTek MT6575 MSDC Driver [ 4.954890] sdhci: Secure Digital Host Controller Interface driver [ 4.961182] sdhci: Copyright(c) Pierre Ossman [ 4.966206] mtk-sd 10130000.sdhci: no support for card's volts [ 4.972137] mmc0: error -22 whilst initialising SDIO card [ 4.978735] sdhci-pltfm: SDHCI platform and OF driver helper [ 4.989205] usbcore: registered new interface driver usb-storage [ 4.998530] mtk-sd 10130000.sdhci: no support for card's volts [ 5.004520] mmc0: error -22 whilst initialising MMC card [ 5.010722] kmodloader: done loading kernel modules from /etc/modules-boot.d/* [ 5.020594] init: - preinit - [ 5.254946] mtk-sd 10130000.sdhci: no support for card's volts [ 5.260882] mmc0: error -22 whilst initialising SDIO card [ 5.356121] mtk-sd 10130000.sdhci: no support for card's volts [ 5.362058] mmc0: error -22 whilst initialising MMC card [ 5.600540] mtk-sd 10130000.sdhci: no support for card's volts [ 5.606528] mmc0: error -22 whilst initialising SDIO card [ 5.626395] mtk-sd 10130000.sdhci: no support for card's volts [ 5.632331] mmc0: error -22 whilst initialising MMC card
That's the same url I posted above :-P
The Omega2S schematic is on there but the Omega2S doesn't have the same pin headers as the Omega2. I think you are correct that I could correlate the two manually but that's a lot more effort.
I asked separately about the schematics but if those aren't available would someone know if any, and if so which, pins on the Omega2+ pin headers are also used for the MicroSD card? I'd like to make use of a number of the header pins for various things but also ensure that I'm not breaking MicroSD support.
I've looked in the GitHub repository https://github.com/OnionIoT/Onion-Hardware/tree/master/Schematics, but don't see them there.
Trying to debug an SD card not showing up and I suspect that I'm using pins that conflict with the SD card pins but have no way to easily check at the moment.
@luz I'm still seeing it here with a pretty new Omega2+ and the latest firmware, although I'm not sure if the oupgrade process upgrades the boot loader or not.
@luz what did you end up doing to resolve this issue? I'm not too excited about rebuilding the boot loader with a change but I'm also using this pin to control an external relay and the relay is activating during boot time when GPIO18 is in analog mode and goes to ~1.65V.
I'm thinking about moving to another available GPIO that doesn't exhibit this behavior but wanted to hear what you had to say first.
@Erik-Hougaard isn’t the container saved unless you delete it? I built the toolchain and images here under docker on macOS and just start the container when I need access to it.
Are you running out of space on your local drive? The build is huge here, like 10gb or something.
Is it possible to use smartconfig to configure the omega2 wifi? Like http://processors.wiki.ti.com/index.php/CC3200_SmartConfig_Provisioning ?
We are looking at an optimal approach for provisioning an IoT device built with an Omega2. BLE is another option we considered but it looks like there isn't integrated BLE Support with the Omega2 at the moment.
Our next thought was to use the self hosted AP but from our initial investigation on Android and iOS it looks like we may need user interaction to connect to the self hosted AP and enter the wifi password.
Thoughts? Comments on how you've done it to achieve the smoothest user experience?
I'd like to measure fan frequency via the tach/sense pin. I'm also converting this signal level to be within the input range of the Omega2.
My question is, is there hardware support for a GPIO counter? Measuring input frequencies is a pretty common activity with embedded controllers, where the Omega2 may be used. I'm surprised nothing turned up with a google search so wanted to ask here about how this might be done and any pointers you might have! (I'd also prefer not to have to add an additional ic to measure frequency.)
This makes sense. You'd like to be a ubus client. A google search turned up this example:
I'm trying to use Poco with a piece of software I'm writing for the Omega2. I can't find any other package, not in the OnionIoT OpenWRT packages or in the full OpenWRT packages that is doing so. I keep thinking someone else must be using Poco in an Omega2 package....
I'm trying to set up a build environment so I can generate opgks of a few applications I'd like to load onto the omega2.
I retrieved the omega2-source docker image and followed the instructions in https://onion.io/2bt-cross-compiling-c-programs-part-1/ but the build is failing on mosquito.
Tried a 'git pull' and that got a bunch of new changes, also did a
scripts/feeds update -a
and the build is still failing. Disabling libwebsocket support in mosquito appears to be working but it isn't clear how this can be the official build repository if it doesn't build :-)
Looks like your connection to Community was lost, please wait while we try to reconnect.