-->
Page 7 of 11

Re: Ultra Low Power Consumption (ideal regulator)

PostPosted: Wed Jun 12, 2019 6:23 pm
by davydnorris
This is a *really* interesting discussion...

I have now started looking at the sensors I use to see if they are capable of running at 2.8V and the only one so far that is not is the INA219 voltage/current sensor, but there's an INA223 that looks to be the same measuring wise but is 2.8V supply friendly

Running everything at 2.8V (and many of my sensors are already running at 1.8V) could really extend low power life span

Re: Ultra Low Power Consumption (ideal regulator)

PostPosted: Wed Jun 12, 2019 10:57 pm
by rudy
I'm more than a little skeptical when it comes to the 2.8 volt operation rating. From my testing with ESP12E modules, I was getting unreliable operation when the voltage dipped below 3.0 volts. Dipped is not the right word. A very short duration glitch.

The 2.8 volt spec for the chip might be correct. But the implementation might not be good enough when power, measured elsewhere is at the minimum rated voltage.

As an example of how a part doesn't meet spec because of the implementation.
viewtopic.php?f=13&t=17514

I would do some serious testing before using a 2.8 volt regulator in a product with the ESP8266.

Re: Ultra Low Power Consumption (ideal regulator)

PostPosted: Thu Jun 13, 2019 5:17 am
by btidey
Yes. I was quite surprised to find my ESP-12F still running reliably from effectively 2.6V and performance at the margins will probably vary quite a bit across individual modules and regulator tolerances.

I do decouple pretty well with 220uF tantalum SMD very close to the module power and that probably helps.

Nevertheless 2.8V regulators will increase the risk of unreliability. 2.9V and 3.0V versions are also available and the latter may be a better compromise.

Re: Ultra Low Power Consumption (ideal regulator)

PostPosted: Thu Jun 13, 2019 1:39 pm
by lucasromeiro
btidey wrote:
lucasromeiro wrote:
btidey wrote:
Just to add to that experience. I was recently checking out the battery state on a door sensor which sends out internet notifications. That had been running fine for over 12 months on a 400mA LIPO using a LDO (xc6203). I was surprised to find the battery down to 2.8V, well into final exhaustion state, but it was still operational. With the drop out the ESP would have been on about 2.6V.


Amazing!!!
The regulator (xc6203) was giving way to how much? 3.3v?
With 2.6v at the entrance?
Tell us more how your application works !!
I am impressed with the weather!


I was just using a regular xc6203 3.3V regulator so when the battery got down to 2.8V the regulator was just passing the battery voltage minus about 150mV so the esp was running off about 2.65V.

One can get a 2.8V version of the xc6203 and that might be worth considering as the deep sleep current of the esp module decreases quite a bit on lower voltages.

The app hardware description is at https://github.com/roberttidey/SecuritySensor


Thank you. I'll consider this.
I'll look at your project!