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.
> and breakers are slow enough you could see multiple joules of energy into your device.
The magnetic part of a miniature circuit breaker will trip in nanoseconds with enough fault current or over voltage, but the thermal elements can take longer to trip for a lower amount of fault current or voltage. Instantaneous trip ratings are generally max out at 16.67ms to clear the fault in one cycle.
Large frame circuit breakers have protection relays that detect fault current and over voltage and trip the breaker.
I am not sure if you're agreeing with me or not. Assuming you are, thank you. Nanoseconds could save you but milliseconds will likely not. It takes very little to explode a chip that isn't fortified and designed for robustness.
Instantaneous trip will kick in with 3-5x over current which happens in nanoseconds. Thermal overloads take longer so they don’t nuisance trip on inrush motor current (not applicable to single-phase motors due to start/run capacitors)
Also, what chip is taking in line voltage single-phase AC at 120V or 208V or 277V or 480V?
All of the semiconductors I am familiar with run on low voltage DC with an inverter between AC and DC.
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. The phone charger circuit can be upwards of 95% efficient, so very little power is getting turned into heat. The ADC input to your scope, on the other hand, would turn 100% of that into heat, which is why it'd blow up.
Practical example is the 50 ohm term. Most scopes I've seen max that at 5 Vrms. P = V^2/R, so 0.5W being dissipated. Now assume you hooked your scope to mains and accidentally turned on 50 ohm term. A low mains voltage is 100Vrms. That's 200W. 400x the maximum. Could a device take 200W? Sure. Could that device? No
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.
Ting by Whisker Labs [1] in the US has this data aggregated by their home electrical safety monitoring devices, and there’s GridRadar [2] in Europe. As you mention, its GNSS based time sync with something a little less featured than PMUs [3] (phasor measurement units) traditionally used for grid health monitoring.
I remember reading an article about this being used in some forensic capacity to determine the date/time a video was taken by comparing the frequency noise.
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.
> and breakers are slow enough you could see multiple joules of energy into your device.
The magnetic part of a miniature circuit breaker will trip in nanoseconds with enough fault current or over voltage, but the thermal elements can take longer to trip for a lower amount of fault current or voltage. Instantaneous trip ratings are generally max out at 16.67ms to clear the fault in one cycle.
Large frame circuit breakers have protection relays that detect fault current and over voltage and trip the breaker.
Breaker trip curves for Cutler Hammer BR breakers: https://www.eaton.com/content/dam/eaton/products/low-voltage...
I am not sure if you're agreeing with me or not. Assuming you are, thank you. Nanoseconds could save you but milliseconds will likely not. It takes very little to explode a chip that isn't fortified and designed for robustness.
Instantaneous trip will kick in with 3-5x over current which happens in nanoseconds. Thermal overloads take longer so they don’t nuisance trip on inrush motor current (not applicable to single-phase motors due to start/run capacitors) Also, what chip is taking in line voltage single-phase AC at 120V or 208V or 277V or 480V?
All of the semiconductors I am familiar with run on low voltage DC with an inverter between AC and DC.
> 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. The phone charger circuit can be upwards of 95% efficient, so very little power is getting turned into heat. The ADC input to your scope, on the other hand, would turn 100% of that into heat, which is why it'd blow up.
Practical example is the 50 ohm term. Most scopes I've seen max that at 5 Vrms. P = V^2/R, so 0.5W being dissipated. Now assume you hooked your scope to mains and accidentally turned on 50 ohm term. A low mains voltage is 100Vrms. That's 200W. 400x the maximum. Could a device take 200W? Sure. Could that device? No
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.
Ting by Whisker Labs [1] in the US has this data aggregated by their home electrical safety monitoring devices, and there’s GridRadar [2] in Europe. As you mention, its GNSS based time sync with something a little less featured than PMUs [3] (phasor measurement units) traditionally used for grid health monitoring.
[1] https://power-quality.tingfire.com/
[2] https://gridradar.net/en/wide-area-monitoring-system
[3] https://en.wikipedia.org/wiki/Phasor_measurement_unit
Is there any API for the US grids?
I remember reading an article about this being used in some forensic capacity to determine the date/time a video was taken by comparing the frequency noise.
https://fnetpublic.utk.edu/frequencygauge.html