ESP8266 Community Forum

- Tue Jan 31, 2017 9:32 am #61817 Hi,

I'm working on building some IoT buttons (similar to https://www.youtube.com/watch?v=6JD2RMDM88Y&t=46s ). I have a few different boards, some ESP-12E (plain board) and Wemos D1 mini. I'm trying to understand when/if I should be adding a capacitor across the power / gnd. It appears based on the spec that the wemos D1 has some small capacitors built into it but very small ones. Right now I'm powering it via 5v USB but my eventual button will use a 3.2V (Lifepo4) battery. Would that increase the value of me having a larger capacitor? Is there really any risk of having a capacitor across Vcc/Gnd?

On a related note for my ESP-12E boards, I'm trying to determine if using an LDO is necessary or perhaps problematic is already below 3.6v. I'm assuming using the LDO when its not required isn't a good idea, but want to verify before I waste or screw up any components.

Thanks

- Tue Jan 31, 2017 6:59 pm #61844 Adding a capacitor across the supply can only do good, provides stability and smooths out any short power perturbations, risk is tending to zero. Q=c.v therefore i^2.t=c.v and if v=3.3v and i=75mA and assume you want to smooth out a 100mS power lost then ~ c= 170uF so use 220uF, or adjust accordingly by varying time.

AFAIK there is only one LDO regulator as used on the Wemos and Node mcu boards that can work down to 3.6 volts. Don't forget that battery voltages are quoted as nominal a Lipo for example is quoted as 3.8v but it's usable voltage is 3.66 discharged to 4.2 fully charged, are you sure the Life type won't exceed the ESP max voltage?

- Tue Jan 31, 2017 7:13 pm #61846 Thanks, your info on the capacitors was assuring.

As for the battery, from my reading lifepo4 batteries dont require the LDO. Measuring a fully charge battery I get 3.3v and its listed as a nominal voltage of 3.2. This video, https://www.youtube.com/watch?v=heD1zw3bMhw (at around the 4:50 second mark) shows the relatively flat discharge curve.

- Sat Sep 15, 2018 5:33 am #78239

Adding a capacitor across the supply can only do good, provides stability and smooths out any short power perturbations, risk is tending to zero. Q=c.v therefore i^2.t=c.v and if v=3.3v and i=75mA and assume you want to smooth out a 100mS power lost then ~ c= 170uF so use 220uF, or adjust accordingly by varying time.

AFAIK there is only one LDO regulator as used on the Wemos and Node mcu boards that can work down to 3.6 volts. Don't forget that battery voltages are quoted as nominal a Lipo for example is quoted as 3.8v but it's usable voltage is 3.66 discharged to 4.2 fully charged, are you sure the Life type won't exceed the ESP max voltage?

Are you sure about this? I just put a 1000uF capacitor between 5v and GND on my Wemos D1 Mini and now it's completely dead. Does not even respond to USB-connection to the computer any more, so I assume it's now trash.

The reason I was trying this, is because there seem to be some weird power leakage in my house. I have my Wemos D1 Mini connected to a USB-adapter powered from cables in my light switch socket. The problem is that if I switch on the light in ANOTHER room, the Wemos detects an interrupt on all the pins! Very strange. Any ideas?