 Hi guys, first of all thank you for your support in the last video. I will get right to the point, in my last video I developed a smart window and door sensor, and the last few weeks I have been researching methods to improve the energy efficiency of the SP32 and in general of the device I am designed. Also many of you gave me a lot of devices about power consumption, energy efficiency and so on. And I thank you for that. This video will be more a video update of the project, so I will show you some tests and how I plan to improve this prototype. But first this project is possible with the help of PCBWay. And yes, this video is sponsored by PCBWay. You can buy your PCBs for only $5, check the link below to place your first order now. Thank you again PCBWay for the support. So using a deep sleep of the SP32 to wake up the my controller is an efficient method because it does not allow us to go below 3 or 5 microampere. The problem can be solved in two ways by using a more efficient deep sleep my controller such as a Tini 10 or a STM32 or by using a power latch to control the SP32 power supply. Design a power latch can be more complex for some application like digital sensor without interrupt pin. A few weeks ago I designed this new version of the door sensor. The main changes from the previous version are first I removed the digital RGB led with the free leds that are powered only when needed. I replaced the read sensor with a magnetic sensor. It's a test to compare the two different solutions and then I added a Tini 10 to control the sensor interrupt and enable of the LDO that provides power to the SP32. And then the final change is try to replace the T9080 with SPX. Spoiler was not a good idea. And this guys was my first idea to improve the design. While the PCBs were in production I got the components and I test the SPX hoping it would be more efficient when the enable was disabled than the RT9080. But as you can see the LDO have the same efficiency and in general RT9080 it's better. Also one of you suggest me to try using a low power mic controller to handle the interrupt in the power block of the SP32. So I figured it was a good idea because I found several mic controllers that in deep sleep consume very low power in order of non on pair. This is the reason of the Tini 10. After place the order of the new PCBs I realized that the best solution is actually to develop a custom power latch. It's the only way to reduce the power consumption as much as possible. So the best idea is to use a power latch or design a power latch or something like that. Doing some research I found this component, the TPL by Texas Instruments. For this type of application it will be perfect maybe. You will say why isn't perfect for a window and door sensor. Because this design to have a transient height signal like a button. So zero I push button, one and go again to zero the signal. Actually this circuit is perfect for IOT button project, the power consumption is perfect only few non on pair. But for this project smart door sensor I only need to power the SP32 when the signal rises from zero to one and when it falls from one to zero. So when the door from closed opens and vice versa. So the door may stay open but the SP32 should not stay on. So when the door is open maybe the sensor give to my controller one or zero. If the signal go to zero to one or one to zero the power latch should power up the my controller. In addition using TPL means that the design depends on these components so I would like to develop something with MOSFET and transistor to avoid the problem for the shortage. This prototype I develop is more efficient than the previous version. I have decreased the power consumption of some components. I am also thinking of developing a single PCB design and not a modular design. As you know I like to develop modular solution with the idea of allowing easy use of the hardware. The problem is that for these kinds of application you need to optimize the hardware a lot. So it's difficult to ensure the modularity of the device. For sure modularity is useful to benchmark with different my controllers for the same final application. But on final device designed to be sold maybe it doesn't make sense. Guys what do you think about modular design it's useful for this type of project. I have already almost finished developing the new version of the project with a custom power latch that should ensure low power consumption. In the next video I will show you the new version of the dev kit for energy harvesting is an evolution to the previous solution I showed you a few months ago. But this time is compatible with Objects Link. Thank you again to PCBWay for supporting the project and guys thank you for watching until the end and see you next time in the next video and next project, bye.