To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/LearnEmbeddedSystems/. The first 200 of you will get 20% off Brilliant’s annual premium subscription.
@@johnchapman1903 another youtuber, @clough42, recently did a vid replacing the mercury batteries in a piece of old lab equipment with long life lithium cells and an ultra-low power regulator that would be *perfect* for this setup. It measured in the microamps for draw on standby. Texas Instruments TPS7A2501
Would be very good to see a practical example of using ESP32 C6 utilising the thread programming aspect in conjunction with existing thread network ie Apple HomeKit ecosystem… I expect others would be interested in this also!
There are two _completely different_ versions of the C6, one has 4MB of Flash _on the die itself_ and a much lower pin-count, the other has *zero* Flash on-die and requires and external Flash device and has the exposed pins for that purpose (higher pin-count). The C3 had a similar execution, albeit with the same pin-counts, where one version has 4MB Flash on-die and others have none.
4:40 that is not a typo on the datasheet. The 22pin has 4mb flash and the 30pin had no flash and needs an external flash chip which is 8mb for the module
The module itself draws 5 uA in deep sleep, but the board has components such as the regulator, which draws 5 mA. If low power is needed, the board must be modified.
I just checked the schematic. The 3.3V LDO Output is connected to CP2102's Regulator Input Pin. This has also been the case with the Pico Kit D4 Board as well. Removing CP2102 from the board and testing the sleep current can get you the right result. I did measure 5uA range in deep sleep by removing the LDO and CP2102 in the Pico Kit D4 Board.
I would be very interested in a thread/matter example! It would be amazing if there was a straight forward matter on thread or wifi example that could be used as a starting point for custom iot devices. If the deep sleep works with the wake on thread I would be ecstatic. I've been using the nRF52840, but it's community seems much smaller then the esp32 community and getting things running on it has taken weeks with a lot of pain points.
18:02 Spec is for the module, you measured the dev board. You can measure the module consumption by removing the jumper (and putting your ammeter between the pins)
i'd like to use one of those to make my little Electric appliance intelligent and contrôolable via zigbee. first on my list : my automated coffee machine (check the water, binger and coffee levels before sleep, distance wake up (the wake-up takes 3 annoying minutes) display the number of cup before running up of water, coffee and binge ect..) and of course display on home assistant the state, the number of coffee per day etc... for this i will have to "reverse enginer the coffed maker mother board, and that's my goal for now...
Hi, just curious about the power measurement. Can you sometimes make a video about how to measure microcontrollers power? Or maybe using other tools? Thanks, appreciate your great explanation video!
Great video! BTW, could please share what kind of breadboard you use at 11:25? I am struggling to find a board with 6 holes in a row, like yours has. I checked the links in the description, but those are for standard 5 holes per row boards.
zigbee battery powered devices are controllable (i.e. temperature/motion sensor or so) but when you put your ESP32 to deep sleep, there is no radio. I am wondering how they are doing this to control zigbee devices while still consuming low power and yet to be controllable - do they wake up device every few seconds to check if there is any message from the controller? that is what I do for my devices while using ESPnow: go to deep sleep, wake up, measure whatever needed, send data to the controller, check if there is any command waiting for you and if yes - then do it, if no then go to sleep again - but that is my ESPnow not zigbee - I would like to know how it is done in zigbee
Scheduled Wake up in longer intervals. You poll sensors each 5 minutes or half an hour or so and then go to deep sleep, you don't need temp sensor readings every couple of seconds usually... I think that presence of advanced sleep modes is the main advantage of ESP, when compared e.g. to Raspberry Pi Zero. -- OR -- There is a possibility to query Zigbee when in sleep, because there's still running internal microcode and the power needed to communicate is very low (in contrary to wifi or usb). It depends on code possibilities to communicate. Maybe run simple sentence of signals from sensor might be possible ? Need to investigate it...
@@vencdee in your first case the sending device needs to be modified to accommodate this. it needs to keep retrying until there is an 'ack'. so we need a TCP like guarantee system that the message has been delivered. maybe the Thread provides this guarantee?
If you have 10yrs in cash to ride out any storm, why shouldn’t you have at least 90% in sp500? You can catch the upsides without worrying about any downsides
The chip itself consumes a bit too much to transmit anything, be it WiFi or Bluetooth, just like in previous iterations (C3, etc). For power-conscious applications, it's better to use a single-stack chip (only BT LE, or only WiFi, etc.), like nrf52840 instead. Too bad we don't have a generally-available, cheap (for small makers like me) and royalty-free solution for most of this stuff, especially WiFi -- and by that, I mean outside Espressif offerings. All in all, impressive chip, but power consumption (especially on TX/RX) really needs to improve.
Sadly, the esp32-c6 does not preserver its WiFi-6 connection in deep sleep, it only does so when in automatic light sleep - which uses much more power than deep sleep. So if you are going to use deep sleep you have to code to have the esp32-c6 reconnect to wifi when it wakes up
