Hi everybody,
I am running a network of 12 sensors for our project #viadukaduk. Now and then sensors fail simultaneously. I was wondering if you ever experienced this too and was it linked to in your opinion:
Greets,
Bron
What patterns (such as geographical location) can you observe in the failure s ?
Check #viadukaduk on the map or lawaai.viadukaduk.be: It’s a monitoring project we run on sck 2.1 since april 2020.
The sensors are all located within about a kilometer circle and sometimes fail simultaneously - within minutes of each other.
That sounds bizarre. It’s hard to rule out a common cause in such circumstances.
Possibilities include spikes on the mains supply (assuming they are connected).
Is lightning out of the question ?
I’ll tell you a story from my past. I was a technician working for the main Telecom Company in Australia; in Darwin. Darwin is a tropical area; with high rainfall. The soil has been leached of salts making it low conductivity.
Whilst lightning often saw failures of equipment by way of ground strikes; sometimes failures would occur when there appears to be no clouds around. What happened was that high voltage power transmission lines sometimes fall to the ground; causing a spike in ground potential in that area. And equipment failed because of it. HVAC travelled up the telecom earth wire and into our equipment; looking for a low impedance path to earth potential!
We also saw failures of equipment several days after a storm. A spike during the storm caused electrical stress; but actual device failure occurred later.
Another possibility is batch failures of electronic components. If the boards are made from a batch of components sometimes they can all have the same weakness; and fail at a similar age. But not normally in the same short time interval. It’s unheard of; making me more certain it’s likely to be an external agent.
Thirdly is if the units are located in the line of a beam of strong radio signal such as airport or military radar.
Another possibility; you may have heard of recent solar activity in the form of a coronal mass ejection in the last few weeks. Apparently this causes problems with satellites; but can also affect sensitive ground based equipment.
Due to the multiplicity of possible causes it may be difficult to nail the exact cause; instead find a way to handle the fallout and face up to the need to replace boards.
May I suggest that you take a close look at all of the affected units and determine if the issues can be dealt with by way of CLI resets or other commands. Also check for obvious issues such as urban sensor 18 pin connector not being fully inserted etc.
Also positively confirm a given board is faulty by replacing it with a known good board from another kit.
It’s entirely possible that you might actually have one or two faulty data boards too, you see.
Come up with a count of boards that can only be fixed by replacements. I know you are looking at this already from your other posts.
Then reach out to the SCS team direct (@oscgonfer @victor ) via this forum to see what can be done about a special order for the necessary items. I agree it’s tough to have to order every component when you only need a specific board (and there are several involved).
Cheers and all the best
@bryn.parrott
Hey @bryn.parrott,
Thank you for the good piece of advice. Darwin, that is a cool story as such! I’d love to see these areas myself, and then the outback. You still live down under?
Greetings from Belgium,
Bron
No, I am living in Taiwan now. It’s a better place.
Yes, Australia is nice; but not so great when you are getting old.
But feeling the need for an expansive environment monitor (see my system under Projects)