The way that the Chinese manufacturers design fairly decent modules, but with absolutely no comprehension of electrical separation is weird. Things like opto-isolators, but then "isolated" tracks from either side just passing right next to each other. I don't think I've come across the live copper fill before though.
i wanted to buy the naked one, took look at the pcb photo and immediately noticed the plane under the optos, where it should be empty or even slotted... i knew there must be something else that's wrong in the mains side so indeed this mains on single smd resistor find was "nice" to know too... so that's what you get when you're trying to save the cost and don't want to spend 150 euros on actual modbus 3-phase 230/400v meter
Good video. It's nice to see you're back! I have messed with several energy monitoring chips like the CS5464 and the ADE7913, but these units look interesting. I was thinking of trying out micropython on the ESP32... maybe I'll wait on that.
Fans are waiting for more videos man!!! How can we go on if you don't produce new cool videos dude?!?! We wanna see more with the bench and your tek 7000 series scope. Come on, Connor!!!! Thanks for the vids!
My issue with the sonoff pow devices are that 1: I generally *don't* want the facility to interrupt power at all, and 2: the actual power measurement is kind of hokey. and I don't have a lot of confidence in it..
regarding the Slave ID the software is designed to be connected to a bus with only one slave BUT it can talk to any address the slave is set to , and it can change it to any address too. the software use a general address that is 0xF8 (248) , any PZEM on the bus will respond to command send with this particular address , but obviously if there is more than one device they will all send a response
The optoisolator closer to the center of the board seems to be referenced to both the high voltage "ground" and connected to the RS485 side. Looks like an easy rework though.
yeah with the cheap RS485 adapters they have a fixed hold time. This can cause problems if you go from 9600 baud to say 115k, because now your hold time is huge. So its easy to get bus clashing, where response data is lost. Having to set two stop bits sounds potentially related to bus hold times, etc. When I made my own "smart" usb converter (only Tx/Rx required), it would vary the line driver enable time depending on the set baud rate, so much shorter for high baud. You can also read the data you send back for timing. Good analysis.
nice, I got one of the bare pzem 014 boards a while ago and just thought it was an unusual serial protocol but it was modbus? well.. writing a driver from scratch is uh.. a good learning experience right?
We missed you and your quick wit! Except the stupid comments about open source software we could have done without which is absolute nonsense. Nevertheless, I love your work with hacking these electrical power sensing products.
The supplied software is linked on the first page of the manual at amazon. images-na.ssl-images-amazon.com/images/I/81GtkIOyZaL.pdf Interestingly the link is not in the paperwork they send you.
Welcome back Connor.....missed your video's. Hope to see some more and learn what you've been up to.
The way that the Chinese manufacturers design fairly decent modules, but with absolutely no comprehension of electrical separation is weird. Things like opto-isolators, but then "isolated" tracks from either side just passing right next to each other. I don't think I've come across the live copper fill before though.
bigclivedotcom You here makes me feel oldskool 😆
i wanted to buy the naked one, took look at the pcb photo and immediately noticed the plane under the optos, where it should be empty or even slotted... i knew there must be something else that's wrong in the mains side so indeed this mains on single smd resistor find was "nice" to know too... so that's what you get when you're trying to save the cost and don't want to spend 150 euros on actual modbus 3-phase 230/400v meter
Welcome back!
Really missed your videos.
Good video. It's nice to see you're back! I have messed with several energy monitoring chips like the CS5464 and the ADE7913, but these units look interesting. I was thinking of trying out micropython on the ESP32... maybe I'll wait on that.
I thought I'd watch an old Connor video then this turned up,,, spooky action at a distance,,,, welcome back
Welcome back!! 😀
Fans are waiting for more videos man!!! How can we go on if you don't produce new cool videos dude?!?! We wanna see more with the bench and your tek 7000 series scope. Come on, Connor!!!! Thanks for the vids!
My god your alive.
Looking forward to the next video with adding the ESP to it
Hi. Right now I am playing with sonoff pow r2. Works really well (16A max) but I want to use a current trnsformer for 100a.
My issue with the sonoff pow devices are that 1: I generally *don't* want the facility to interrupt power at all, and 2: the actual power measurement is kind of hokey. and I don't have a lot of confidence in it..
regarding the Slave ID
the software is designed to be connected to a bus with only one slave
BUT it can talk to any address the slave is set to , and it can change it to any address too.
the software use a general address that is 0xF8 (248) , any PZEM on the bus will respond to command send with this particular address , but obviously if there is more than one device they will all send a response
The optoisolator closer to the center of the board seems to be referenced to both the high voltage "ground" and connected to the RS485 side. Looks like an easy rework though.
Ants don't come out of the woodwork, termites come out of the woodwork!
yeah with the cheap RS485 adapters they have a fixed hold time. This can cause problems if you go from 9600 baud to say 115k, because now your hold time is huge. So its easy to get bus clashing, where response data is lost. Having to set two stop bits sounds potentially related to bus hold times, etc.
When I made my own "smart" usb converter (only Tx/Rx required), it would vary the line driver enable time depending on the set baud rate, so much shorter for high baud. You can also read the data you send back for timing.
Good analysis.
Hey Connor, how are you? Long time since you posted your last video. Hope you're doing fine!
How's the witness protection program treating you?
nice, I got one of the bare pzem 014 boards a while ago and just thought it was an unusual serial protocol but it was modbus? well.. writing a driver from scratch is uh.. a good learning experience right?
Yeah, Modbus-RTU.
Do a longer vid on how you got into electronics and your "appreticeship".
will it work correctly in my place having 2-wire , without neutral wire i.e. L1 and L2 each is hot wire having 110V (makes 220V).
The manual says ac80~260v input. Since peak voltage from a 220v plug is 311v I would say no.
Great video. Any chance you could try this with a raspberry pi?
Out of nosiness or curiosity or whatever,... what are you using to write C++ on the PC side?
Sublime text + Deviot + PlatformIO
Connor! Don't be a stranger! Oh wait, this is your channel.
how can i change slave ID thank you for your tuto
We missed you and your quick wit!
Except the stupid comments about open source software we could have done without which is absolute nonsense.
Nevertheless, I love your work with hacking these electrical power sensing products.
please send pc display software
The supplied software is linked on the first page of the manual at amazon. images-na.ssl-images-amazon.com/images/I/81GtkIOyZaL.pdf
Interestingly the link is not in the paperwork they send you.