Battery - Githubissues

danpowell88 commented 5 years ago

Any chance you could do an update or show an example of the type of battery that would be compatible with this instead of requiring AC power

recklessnl commented 5 years ago

Seconding this, some locations just have no way of wiring it to AC power without it looking awful.

Additionally, I saw someone using the solar power strips strapped on the glass, which you couldn't see from inside, but they would charge the battery during the day. Is this something that's feasible with this build?

thehookup commented 4 years ago

I have a working prototype of this, trying to ensure its safety and longevity before I release it into the wild.

tung256 commented 4 years ago

that would be awesome! i volunteer to be the first to try if the battery can last at least 2 weeks.

incognit-joe commented 4 years ago

Hi, I have built the kit and have a dual motor set up. Everything running perfectly thanks! I would also be interested in the battery option. Can you advise if you've had any progress with this as i would be pretty gutted if i proceed to drill holes in my walls then a week later you put up a tutorial to add a battery..

Either way i want to extend my thanks, this is the first electronic project i have ever tried and im over the moon to have succeeded!

Rgds, Joe

thehookup commented 4 years ago

In my initial test the battery only lasted 2 months, even with the solar panels on it. I need to revamp the code a bit more but currently have no time to devote to it. Probably in mid-december I'll revisit.

joelwetzel commented 4 years ago

Any updates? I've built 2 of these awesome things now. But would love to get rid of more cables.

1ntroduc3 commented 3 years ago

In my initial test the battery only lasted 2 months, even with the solar panels on it. I need to revamp the code a bit more but currently have no time to devote to it. Probably in mid-december I'll revisit.

Do you have any update?

JustMaier commented 3 years ago

@thehookup Any chance you can at least share a few of the details on what components you used and maybe what adjustments you made? 2 months on battery is actually pretty good.

I'm going to attempt the battery-powered route here's my plan.

Additional Components:

5000mAh USB battery pack
7 Solar Cells (60mA) + Schottky Diodes (to bypass dark cells) (anticipating 70mA - 210mA total)
Micro USB cable to recharge battery pack from 5V from solar cells
Step Up Converter (5v to 12v) between the battery pack and driver (since my base power will be 5V, I plan on dropping the buck converter)

Code Adjustments:

Modem sleep the device and wake it every 5 seconds to check for MQTT messages. Optimally reduces daily power consumption by ~60% (80% of every minute it's running at 25mA instead of 100mA)
Auto close and deep sleep from 8pm - 8am. Hopefully reduces daily power consumption by an additional ~37% (I believe it still consumes 10mA in this mode)
Assign static IP (reduces the power needed on wake)

Hopeful Outcome:

-600mAh/day for nodemcu
-5mAh/day for motor (assuming 1.25mAh per open or close with 4 operations per day)
+560mAh on bad days or +1,680mAh on good days from solar cells

Potential Problems:

Solar is likely to underperform what I've hoped for. I tried to measure output with a multimeter on my desired windows and a single cell was producing 3mA at noon on a sunny day which is considerably less than even my "bad day" estimate. Maybe different panels would help, or maybe my window just doesn't get enough light... Either way, if I'm only able to get 168mAh/day from the cells, I can only expect to go 10 days or so without having to recharge
I haven't accurately measured the consumption of the actual motors to open and close the blinds. I figure they'll only be used 2-4 times a day and I'm hoping that I can essentially just not worry about their overall power draw with such limited use. That assumption could be very wrong.
Sleep cycling the nodemcu like that might not actually save as much energy as I hope
I'll probably have to tweak the USB battery pack so that it doesn't turn off when the nodemcu deep sleeps
I've never done something like this before and I could be horribly horribly wrong with all my calculations and components needed 😂

Mehick57 commented 3 years ago

@thehookup Any chance you can at least share a few of the details on what components you used and maybe what adjustments you made? 2 months on battery is actually pretty good.

I'm going to attempt the battery-powered route here's my plan.

Additional Components:

5000mAh USB battery pack

7 Solar Cells (60mA) + Schottky Diodes (to bypass dark cells) (anticipating 70mA - 210mA total)

Micro USB cable to recharge battery pack from 5V from solar cells

Step Up Converter (5v to 12v) between the battery pack and driver (since my base power will be 5V, I plan on dropping the buck converter)

Code Adjustments:

Modem sleep the device and wake it every 5 seconds to check for MQTT messages. Optimally reduces daily power consumption by ~60% (80% of every minute it's running at 25mA instead of 100mA)

Auto close and deep sleep from 8pm - 8am. Hopefully reduces daily power consumption by an additional ~37% (I believe it still consumes 10mA in this mode)

Assign static IP (reduces the power needed on wake)

Hopeful Outcome:

-600mAh/day for nodemcu

-5mAh/day for motor (assuming 1.25mAh per open or close with 4 operations per day)

+560mAh on bad days or +1,680mAh on good days from solar cells

Potential Problems:

Solar is likely to underperform what I've hoped for. I tried to measure output with a multimeter on my desired windows and a single cell was producing 3mA at noon on a sunny day which is considerably less than even my "bad day" estimate. Maybe different panels would help, or maybe my window just doesn't get enough light... Either way, if I'm only able to get 168mAh/day from the cells, I can only expect to go 10 days or so without having to recharge

I haven't accurately measured the consumption of the actual motors to open and close the blinds. I figure they'll only be used 2-4 times a day and I'm hoping that I can essentially just not worry about their overall power draw with such limited use. That assumption could be very wrong.

Sleep cycling the nodemcu like that might not actually save as much energy as I hope

I'll probably have to tweak the USB battery pack so that it doesn't turn off when the nodemcu deep sleeps

I've never done something like this before and I could be horribly horribly wrong with all my calculations and components needed 😂

Sounds like a great idea and plan! I am just wondering if you have gotten this to work? This is exactly what I am looking for since I can't run wires to ten sets of blinds I would like to convert! Thanks for your time!

Mehick57 commented 3 years ago

In my initial test the battery only lasted 2 months, even with the solar panels on it. I need to revamp the code a bit more but currently have no time to devote to it. Probably in mid-december I'll revisit.

It would be great if one of you geniuses could figure this one out!

JustMaier commented 3 years ago

In my initial test the battery only lasted 2 months, even with the solar panels on it. I need to revamp the code a bit more but currently have no time to devote to it. Probably in mid-december I'll revisit.

It would be great if one of you geniuses could figure this one out!

I actually abandoned this route. In my findings, it seemed that WiFi takes too much energy to be able to run on a battery for an extended period of time.

The best way to do this in my research was actually using RF on a micro-controller designed for low-power usage, like a moteino. This was a helpful guide project.

Mehick57 commented 3 years ago

Thanks so much for your
      quick reply and being pointed towards this project! I will be
      trying to see if I can put this together and get it working on
      HA!

On 5/4/2021 4:54 PM, Justin Maier
  wrote:

      In my initial test the battery only lasted 2 months, even
        with the solar panels on it. I need to revamp the code a bit
        more but currently have no time to devote to it. Probably in
        mid-december I'll revisit.

    It would be great if one of you geniuses could figure this
      one out!

  I actually abandoned this route. In my findings, it seemed that
    WiFi takes too much energy to be able to run on a battery for an
    extended period of time.
  The best way to do this in my research was actually using RF on
    a micro-controller designed for low-power usage, like a moteino.
    This was a helpful guide project.
  —
    You are receiving this because you commented.
    Reply to this email directly, view it on GitHub, or unsubscribe.
  [

{ @.": "http://schema.org", @.": "EmailMessage", "potentialAction": { @.": "ViewAction", "target": "https://github.com/thehookup/Motorized_MQTT_Blinds/issues/3#issuecomment-832275991", "url": "https://github.com/thehookup/Motorized_MQTT_Blinds/issues/3#issuecomment-832275991", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { @.": "Organization", "name": "GitHub", "url": "https://github.com" } } ]

JustMaier commented 3 years ago

@Mehick57 let me know how it goes and what changes you end up making. I tried to do this project but after running into issues with dependencies that had changed I struggled to get the code working with my limited knowledge of developing for micro-controllers.

thehookup / Motorized_MQTT_Blinds

Battery #3