Good to give mains respect. It can deliver kilowatts of power without breaking a sweat, and breakers are slow enough you could see multiple joules of energy into your device. It's an expert magic smoke emancipator.
Disagree on the surprise that the setup worked, though. Mains is only regulated to a few % in frequency from what I've read. But you can see 0.05 Hz deviations (or 1%aka 1000 ppm). Even a junky crystal at ~100ppm is an order of magnitude better. A 10 ppm oscillator isn't hard to find, so the computer is likely somewhere in the middle. The math all checks out.
> Mains is only regulated to a few % in frequency from what I've read.
You've read wrong. While it's a different network, there were articles talking about how if the Texas grid stayed under 59.4Hz for a few minutes longer, some generators would have started cutting out to prevent damage, and the whole thing might have collapsed. So that's a 1% deviation being defcon 1.
And I found a page saying this about the European grid: "The allowed mains frequency range in normal operation is thus obtained at 49.8 Hz to 50.2 Hz." "short term deviations until 800 mHz are allowed (49.200 Hz to 50.800 Hz)."
> But you can see 0.05 Hz deviations (or 1%aka 1000 ppm).
You're right. Within a few 0.1%. shouldn't do mental math late at night. That said, my PPM math was correct, so I'm sticking by my point, which was based on PPMs.
One watt is one joule per second. Most things designed to consume power, especially related to wall devices and consumer electronics you'd charge, are capable of ingesting multiple watts. Most things that aren't, aren't. The difference that I'm not sure I'm seeing from the other comments is that a phone takes that energy and stores it as a chemical reaction, with some losses as heat. For everything else, it's all heat. Also, it's uncontrolled.
So imagine it like this: phone chargers are capping out at 120W maybe on the highest end ones. Assume that somehow they're 95% efficient in all their regulators and stuff. That's 6W of heat. Toasty. Now think about unlimited (1500W is a low end) power all as heat. That's 250x. That'd be like charging your phone at 30 killowatts of power. That's the difference here. Nobody's charging at 30 kilowatts of power. So really it's the same issue. There's a lot of fudge factor here because I can't be bothered to work out all the losses but hopefully the point comes across.
It's a question of energy density. Multiple joules into your big phone battery is nothing, multiple joules into a small SMD component means it evaporates immediately in a bright flash!
"slow" is still measured in milliseconds. That said, the energy is probably in the low hundred Joules. which is a lot when theres no time to dissipate it.
So... The setup was already running and they happened to catch the Feb 8 in incident? What a luck!
There is open distributed monitoring for all kinds of signals, like seismometer networks, weather, ads-b...
Is there anything like this for the power network? Like a reference design or an esp32-shield?
How would it look like if we were serious?
I would make it three phase, with direct coupling to also see the exact voltage changing over the day. Sometimes we have issues with local voltage rising too high and PV inverters shutting off. I'd like to see and log this.
An audio ADC should be good, but needs three channels.
For distributed sensing and logging, you would need a reasonable accurate time synchronization. Raw ntp over internet might not be good enough, at least not for localizing fault propagation issues over the whole continent. Better stick a 5€ GPS module on there.
I've been thinking about this for a while and did a bunch of research and planning, though had no time for building yet.
Stepdown transformer or 12V AC power bricks with barrel jack output are easy to get and cheap. Lower voltage, easier to measure and you outsource the safety/galvanic isolation.
Zero crossing ICs tied to GPS PPS signal. Add on ADC to measure voltage swings. Esp32 can record and compress data locally, only sending summary frequency, phase/time alignment.
This is obviously nice for outage detection too. For some extra cost add SD card for longer storage buffer and ability to backfill data once connection is restored. Small battery to not lose 1-2 minutes of time sync from GPS on powerup.
My last rough estimate was ~$40 for minimal single phase version and ~$100 for kitted out three phase battery backup, in volume of tens of units. That's pretty decent, I could then mail a couple dozen to people for free to get global coverage.
My main unsolved concerns were: liability (I don't want anyone to sue me if their house burns down), trust in data (it's easy to send spoofed garbage, so initially I would want only trusted parties to send, all data would be public of course), and the most important one: finding time and motivation next to my day job to do this.
I've been using a transformer & resistor voltage divider direct into audio in to decode Decabit signals (a form of ripple control aka audio frequency load control): https://web.archive.org/web/20140127003936/http://www.anime-...
I'll test if this antenna methods works as an alternative. I'd feel more comfortable sharing with others if mains voltages are eliminated entirely.
Good to give mains respect. It can deliver kilowatts of power without breaking a sweat, and breakers are slow enough you could see multiple joules of energy into your device. It's an expert magic smoke emancipator.
Disagree on the surprise that the setup worked, though. Mains is only regulated to a few % in frequency from what I've read. But you can see 0.05 Hz deviations (or 1%aka 1000 ppm). Even a junky crystal at ~100ppm is an order of magnitude better. A 10 ppm oscillator isn't hard to find, so the computer is likely somewhere in the middle. The math all checks out.
> Mains is only regulated to a few % in frequency from what I've read.
You've read wrong. While it's a different network, there were articles talking about how if the Texas grid stayed under 59.4Hz for a few minutes longer, some generators would have started cutting out to prevent damage, and the whole thing might have collapsed. So that's a 1% deviation being defcon 1.
And I found a page saying this about the European grid: "The allowed mains frequency range in normal operation is thus obtained at 49.8 Hz to 50.2 Hz." "short term deviations until 800 mHz are allowed (49.200 Hz to 50.800 Hz)."
> But you can see 0.05 Hz deviations (or 1%aka 1000 ppm).
That's 0.1%
You're right. Within a few 0.1%. shouldn't do mental math late at night. That said, my PPM math was correct, so I'm sticking by my point, which was based on PPMs.
> you could see multiple joules of energy into your device
Is that supposed to be a lot? Your phone receives multiple joules every second when charging, even with a slow charger.
One watt is one joule per second. Most things designed to consume power, especially related to wall devices and consumer electronics you'd charge, are capable of ingesting multiple watts. Most things that aren't, aren't. The difference that I'm not sure I'm seeing from the other comments is that a phone takes that energy and stores it as a chemical reaction, with some losses as heat. For everything else, it's all heat. Also, it's uncontrolled.
So imagine it like this: phone chargers are capping out at 120W maybe on the highest end ones. Assume that somehow they're 95% efficient in all their regulators and stuff. That's 6W of heat. Toasty. Now think about unlimited (1500W is a low end) power all as heat. That's 250x. That'd be like charging your phone at 30 killowatts of power. That's the difference here. Nobody's charging at 30 kilowatts of power. So really it's the same issue. There's a lot of fudge factor here because I can't be bothered to work out all the losses but hopefully the point comes across.
It's a question of energy density. Multiple joules into your big phone battery is nothing, multiple joules into a small SMD component means it evaporates immediately in a bright flash!
Yeah, and batteries are deeply weird in that they're a component you put power into and they mostly don't get warm.
"slow" is still measured in milliseconds. That said, the energy is probably in the low hundred Joules. which is a lot when theres no time to dissipate it.
I bet if you had something like a box fan running you could hear when the desync happened.
So... The setup was already running and they happened to catch the Feb 8 in incident? What a luck!
There is open distributed monitoring for all kinds of signals, like seismometer networks, weather, ads-b... Is there anything like this for the power network? Like a reference design or an esp32-shield?
How would it look like if we were serious?
I would make it three phase, with direct coupling to also see the exact voltage changing over the day. Sometimes we have issues with local voltage rising too high and PV inverters shutting off. I'd like to see and log this. An audio ADC should be good, but needs three channels.
For distributed sensing and logging, you would need a reasonable accurate time synchronization. Raw ntp over internet might not be good enough, at least not for localizing fault propagation issues over the whole continent. Better stick a 5€ GPS module on there.
Anybody seriously working on this..?
I've been thinking about this for a while and did a bunch of research and planning, though had no time for building yet.
Stepdown transformer or 12V AC power bricks with barrel jack output are easy to get and cheap. Lower voltage, easier to measure and you outsource the safety/galvanic isolation.
Zero crossing ICs tied to GPS PPS signal. Add on ADC to measure voltage swings. Esp32 can record and compress data locally, only sending summary frequency, phase/time alignment.
This is obviously nice for outage detection too. For some extra cost add SD card for longer storage buffer and ability to backfill data once connection is restored. Small battery to not lose 1-2 minutes of time sync from GPS on powerup.
My last rough estimate was ~$40 for minimal single phase version and ~$100 for kitted out three phase battery backup, in volume of tens of units. That's pretty decent, I could then mail a couple dozen to people for free to get global coverage.
My main unsolved concerns were: liability (I don't want anyone to sue me if their house burns down), trust in data (it's easy to send spoofed garbage, so initially I would want only trusted parties to send, all data would be public of course), and the most important one: finding time and motivation next to my day job to do this.