I am currently using some ESP32s and BLE beacons with PIR motion sensors to make some sort of home security wireless motion sensor system. I have my first beacon completed, but it draws a whopping 19mA!
Sooo.... the beacon module I'm using requires 1.1 to ~3.6V DC to run. The motion sensor requires 3.3V to 5V to run. I settled on just doing a 3.3V supply, but I'm using a single cell lithium battery, which means it has a voltage range of ~3.2 to 4.2V. See the pickle!
I decided to do a very bad setup, just to get it working (and it does work!) but it is very power hungry. I have the battery connected to a boost converter that outputs a steady 5V, which then goes through a AMS1117-3.3 3.3V LDO regulator to the parts.
I measured 19mA of current consumption overall, with only 0.5mA being used by the motion sensor and beacon combined!
Any suggestions on how to cut this awful power bill? I feel like I am missing a very simple solution, but am too worn out to see it.
Home security... don't have wall power? A phone charger and buck converter to 3.3V makes sense if you do. 19mA, call it 20mA is a lot? Check what 3 AA batteries will supply,
quote: A typical AA battery has a 2 ampere-hour capacity, and can supply peak currents of over 2A. The battery can typically supply 1.5V when fully charged, down to 0.9V when discharged.
3 AA's in series should give 2000 mA-hours, mostly at 4.5V amd you need 20mA. Buck that to 3.3V and get like 100 hours before needing to change or recharge the AA's. With high-end AA's get even longer.Use more cells, get more time. With wall power you'd need the batteries when power goes out and then so would the base system!
Just what do you need? 100 hours is over 4 days. 2 x 3-in-series-AA's in parallel should get you 8 days, a week with a day late.
How much will you spend? How about a solar panel with charge circuit and li-ion battery pack?
If my memory is working properly there were several X10 PIR sensors that ran for months on two AAA batteries. I probably still have some. The schematics were published which would give you some ideas. You could remove all the LEDs or change the resistors so they do not pull as much current. LEDs will visibly light with less then 1 mA. The sensor could trigger a zero circuit circuit that would turn on the ESP32, and it could turn itself off when finished. This will give you a major power reduction.
Which one. Idle current could be 10mA. Why boost, and then reduce. There are regulators with a dropout voltage of 0.1volt, and a very low idle current.
Another poor choice. Idle current of this regulator is 5-10mA.
So connect that one directly to the LiPo battery.
Post a links to the beacon module and motion sensor.
Which ESP32. There are dozens of different ones.
Leo..
I actually am just trying to get my battery to last as long as possible . I am currently using a li-ion cell with a capacity of 1000mAh.
A solar panel with a charge circuit is a good idea, but I am trying to make them as small as possible, while also being able to be easily repositioned.
The ESP32 is only the base station... I am using these BLE beacons:
I just used that setup on my last project though! Awesome easy way to save power.
I mentioned it was a bad setup . Im trying to use what I have on hand. The booster I am using is this one:
Do you know of a regulator that has a low idle current and a lower dropout voltage? I actually boosted then reduced because the AMS1117-3.3 has a 1.1V drop, so it cant be connected directly to the battery and have it work.
Beacon module linked above, and I am using the classic HC-SR501 PIR motion sensors as I already have plenty. The ESP32 is a FireBeetle ESP32-S C6, but is irrelevant because there is only one, and it is the central hub.
Thanks so much everyone! I hope I provided enough feedback?
Also, I was just thinking... would it be possible to hook the motion sensor directly to the battery, and power the beacon module through a 1n4007 Diode? If I remember correctly, this diode has a voltage drop of ~0.7V, which means it will be under the 3.6V max of the beacon.
Your choice of parts and inability to wire power in may leave you needing more batteries in parallel despite a desire for smallness.
One option may be changing the battery every 2 days. Or running wires from a bigger battery pack in one place to a smaller sense and transmit unit at some distance. If batteries can be spread out, they won't be in a lump, they only have to be together electrically, through wire. Otherwise, as small as possible may just not be as small as you'd like. If it's about visibility, consider spreading the totality out and/or embedding as much as possible into existing framework and drill some holes.
" I am trying to make them as small as possible, while also being able to be easily repositioned."
Well oops on the embedding idea! How about need vs want?
Thank you again @Wawa ! I was having a lot of false positives, and removed the diode and it is a little better. There is still a lot of false positives though. Do you have any idea why this might be?
Also, I measured the current consumption again of everything, and its drawing about 500µA in total! It will last for 83 days now off of this battery (except that the battery can't drop below 3.4V because of the 3.3V regulator and its 100mV drop, so more like ~65 days).
I've also found that the SR505 PIR devices are prone to false triggering and would recommend the AM312 instead. These appear much more stable provided that the supply voltage is kept above 3v3. See: AS312 (AM312) Mini PIR module review – Unusual Electronics . I don't think that you should share the regulator of the PIR with any other device because these [PIRs] are very susceptible to noise.
Thanks again @Wawa! I was going to come on here and ask for more advice on how to reduce the false positives when I saw your last post. The capacitor fixed it!
@kgray9
Yes it has a low dropout but you don't want to operate your regulator in dropout mode since the output is no longer regulated and I'm sure the dropout is much higher when you draw more current.
Ohhh I didn't realize that @jim-p ! Also, I am still getting way too many false positives, even after adding a large capacitor right after the regulator on the 3.3V line and removing the protection diode .
The motion sensor can be directly connected to battery, but I need to figure out a different way to power the beacon. I did notice it can take from ~1.1V to 3.6V, which seems like an opportunity, but I'm still not sure how I would do that with what I have. The lowest LDO I have is 3.3V and I have no buck converters. I might just have to order something if I can't get it to work with what I have, although I would really rather not to.
Thanks again everyone! I hope someone might see a different idea that would work!
I think this will be perfect for your application.
Efficiency is >90% for input voltages between 4.5V an3.6V and currents between 2mA and 50mA
With a 50mA load it will operate with battery voltages at low as 3.4V
If you think false positives are from external voltage regulator, I would just power beacon with 3.2V lifepo4 cell. Ok, two batteries need little more space, mut you get clean supply and zero waisted power from external regulators.
So... when I had my previous setup with the battery going through the booster to 5V, and then the 3.3V LDO, the motion sensor was very accurate, and I didn't have a single false positive. With this new setup I'm running right now, there's at least a dozen false positives a minute .
Give LFP a try. Even if you find that problem is something else, it's always useful to have a 3.2V battery in hands. It's perfect match to modern electronics.
.
)
. I am currently using a li-ion cell with a capacity of 1000mAh.
. Im trying to use what I have on hand. The booster I am using is this one:
.