I have set up 4 Photon2’s and they have all exhibited the same strange behaviour.
Two of the processors were mounted on a custom PCB which is normally powered from a LiPo through on-board regulators and a power switch. However, to set them up I connected the processor directly through USB without the LiPo connected. Both boards appeared to set up correctly but when they rebooted they reported an SOS with 4 blinks (Bus Error). The condition persisted when I powered them from a USB power supply. I did not try powering them from the LiPo.
To my surprise, the next morning they booted normally when powered either from my MacBook or from the USB power supply.
I wondered whether it had something to do with the custom PCB so set up a further 2 Photon2’s on an Adafruit quad backplane but with no other boards mounted on the backplane. Same result - SOS with 4 blinks regardless of whether they were powered from the MacBook or the USB supply. Presto - next morning they both booted normally.
I am concerned that this may be indicative of an intermittent issue and that the successful reboots may be a false sense of confidence. I am about to test them repeatedly over a longer time period.
An almost identical set of 10 custom PCB’s loaded with B542 processors behaves normally.
Has anyone observed similar behaviour with Photon2’s and, if so, have you found a cause/solution.
That's not a known failure mode. If it happens again, could you try putting the Photon 2 in both safe mode (blinking magenta) and also DFU mode (blinking yellow) with the buttons to see if it's able to stay in those modes without SOS+4?
Thanks @rickkas7. I will certainly do so. When I started up one of the processors this morning - surprise surprise it booted normally (connected to an iPad as a power source). I will continue to pursue to see if I can further isolate the triggers, including your request.
Thanks for your prompt assistance.
The Photon2 worked ok for 15 minutes then crashed and on reboot it came up with the SOS+4 error code.
The following are the answers to your question re modes:
Does not enter safe mode - it resets and goes back to SOS+4
Does enter DFU mode
Does enter Listening mode
I have had one Photon2 working ok for some months and it runs reliably, The last 4 that I have setup all display this SOS+4 behaviour.
I have also tried one of them on a Mikroe board with the same results. Since they exhibit the same behaviour on Adafruit and Mikroe boards I do not believe it is an external component problem. I will try this evening running the custom PCB powered from the LiPo battery.
Whatever the cause it is intermittent - after a period of SOS+4 behaviour one of the processors is once again connected to the network. However, an attempt to put it safe mode causes it to go back to SOS+4 mode.
Are those photon 2's running a firmware other than tinker when they SOS?
If they were running a different firmware, do they SOS when you flash tinker in them and let them run on your boards?
@gusgonnet Thanks. I was in the process of registering two sets of 10 processors - B542 and Photon2. I did not get to load any user software before I became aware of the problem with the Photon2’s.
I have since used the CLI to update the processors. I have updated to both 5.9.1 and 6.2.0 followed in both cases by loading tinker. These activities are all reported as successful but the processor blinks green followed by SOS+4
I stopped registering the Photon2’s when I became aware of this issue. In the near term I will have to change focus to code development on the B542’s pending finding a solution for Photon2’s. Whatever the cause, this problem has affected 5 Photon2 processors and I am not going to proceed further until I have a solution.
@rickkas7 I have had two Photon2’s and 1 Photon running for several days. One Photon 2 and a Photon are on breadboards and the other Photon2 is on an Adafruit Quad backplane with no other modules loaded. All three are within 5-10 cms of each other.
The Photon has been solid for the entire time. The Photon2’s have been fluctuating between online and SOS+4. There is no discernible pattern - sometimes one will be online while the other is in SOS+4, other times both will be SOS+4 or online. Sometimes they recover from SOS+4 and go online, other times it requires multiple resets.
I am a little suspicious that it could be a comms problem and will develop some code to log behaviour (they are currently loaded with Tinker). I am using a mesh system with a pair of ASUS XT12’s. I tried using my cell phone as a hotspot but for some reason the Photon2’s do not list the phone even though it is identified and connectable by my iPad, MacBook and an ASUS notebook.
Any suggestions as to what tests I can apply would be appreciated. I have 7 Photon2’s and all are displaying exactly the same behaviour.
@rickkas7 The plot thickens. I have been busy on other matters so have left a set of processors running to see if I can detect any pattern in the SOS+4 behaviour.
I was suspicious there could be an issue coupling in to my home mesh WiFi network. However, I have registered an M542 on a Muon board. I was surprised to find that it too flashes SOS+4 although with a slightly different behaviour. It periodically drops connection and tries to reconnect. When it does so there is a chance that it gets to flashing cyan, looks as though it will blink cyan but quickly drops into SOS+4 mode. It shows 2 cycles of SOS+4 then reboots. Sometimes it will drop into SOS+4 mode again but usually manages to reconnect and breath cyan.
This is similar to the Photon2’s. They drop into SOS+4 mode when trying to connect. The two Photon2’s seem to last breathunbg cyan longer than they were a week or so ago but when they go into SOS+4 mode they persist for extensive times before sometimes reconnecting.
Since I understand that the M524 will prefer to use the 4G network this suggests it is a failure to login once connected to the Particle server rather than a WiFi problem at my end,
A B425 just sits there breathing cyan for 24 hours without missing a beat.
The processors are on a desk 1-2 metres away from the router.
I need to find the time to put some logging code on to all 4 processors but it seems that the issue is related to the 4th Gen processors (Photon2 and M524) and not the 3rd Gen (B524) or for that matter two Photons, although the latter are running an earlier version of the OS.
A small number of users are experiencing issues using the P2/Photon 2/M-SoM (RTL872x) on certain Wi-Fi networks, typically those with repeaters. This is being investigated.