Forum Rule: Always post complete source code & details to reproduce any issue!
Results 1 to 17 of 17

Thread: Teensy 4.x RTC battery life

  1. #1
    Junior Member
    Join Date
    Jul 2021
    Posts
    3

    Teensy 4.x RTC battery life

    Hello

    I have several home-made boards using Teensy 4.0 and 4.1, and on every one I noticed that the external RTC battery has a much shorter life than expected.
    On these boards a CR1220 coin cell battery is directly connected to the VBAT and GND pins, nothing else. But after slightly more than one month I find the battery voltage quickly going to zero.
    At the beginning of August I put a fresh battery into a board and measured around 3.1V. The board was never powered up or otherwise connected. Last week I measured about 2.6V. Yesterday it was zero.
    I'm used to having these type of batteries last for several years, e.g. on an iMX6 with external RTC, and I'm quite surprised of the Teensy behaviour.
    Has someone else noticed this? Can it depend on some leak in the diodes used on the VBAT input?

    Alberto

  2. #2
    Junior Member
    Join Date
    Jul 2021
    Posts
    3
    Sorry, tester problems. For 'Yesterday it was zero' read 'Today is 1.6V'. Still out of specs.

  3. #3
    Senior Member
    Join Date
    Oct 2019
    Location
    Calgary
    Posts
    127
    I've had the same CR2032 hooked up to a Teensy 4.0 for 23 months now. I just popped the top off the case it's in to measure the voltage and it's only down to 2.73V. Not bad for a dollar store battery.

  4. #4
    Senior Member+ defragster's Avatar
    Join Date
    Feb 2015
    Posts
    15,271
    There was one other post some many weeks back reporting the same thing. I didn't have one set aside with a battery in to verify.

    And of course with those CR Batteries - you never really know what you are getting. Having a $_$tore battery survive 23 months is impressive!

    I do have a T_4 in an early TallDog/@loglow breakout where the SD wasn't right so not in active use - and some just ordered 'Energizer ECR1220' cells ... I could put on a time sketch and set aside ... but would take months to offer feedback.
    Test begun: TimeTeensy3.ino
    Code:
    RTC has set the system time
    12:15:14 9 9 2021
    Though I didn't test NOOB voltage ...

    Voltage good at 3.20V ... but bad gnd needs fixed to the holder - Doh!

    Flux and solder the GND wire to fit the socket in use and re-Programmed:
    Code:
    RTC has set the system time
    12:30:57 9 9 2021
    12:30:58 9 9 2021
    // UNPLUGGED - idle - POWERED
    RTC has set the system time
    12:33:50 9 9 2021
     ... // matches minute change on PC
    12:35:00 9 9 2021
    Last edited by defragster; 09-09-2021 at 08:35 PM.

  5. #5
    Senior Member
    Join Date
    Oct 2019
    Location
    Calgary
    Posts
    127
    Quote Originally Posted by defragster View Post
    And of course with those CR Batteries - you never really know what you are getting. Having a $_$tore battery survive 23 months is impressive!
    I was impressed. I decided that it had done its duty and replaced it with a "Basics" battery from a certain well known Internet retailer. We'll see how it does. 3.5V to start.

  6. #6
    Senior Member+ defragster's Avatar
    Join Date
    Feb 2015
    Posts
    15,271
    Quote Originally Posted by Silverlock View Post
    I was impressed. I decided that it had done its duty and replaced it with a "Basics" battery from a certain well known Internet retailer. We'll see how it does. 3.5V to start.
    Wow 3.5V! I just ordered those 'name brand' hoping for good life ... as noted started at 3.2V ... and worked 15 mins so far when unplugged.

  7. #7
    Senior Member
    Join Date
    Oct 2019
    Location
    Calgary
    Posts
    127
    Quote Originally Posted by defragster View Post
    Wow 3.5V!
    Yeah, I wondered about that... then I noticed that the voltage measurement was slowing drifting upwards. I turned on a regulated 5V bench supply and measured that and got 5.7V, drifting upwards. Hmm! Turns out it was long past time for the battery in the multimeter to be replaced! Changed that and got a nice solid 5.01V out of the bench supply, and 3.14V out of the Basics battery.

  8. #8
    Senior Member+ defragster's Avatar
    Join Date
    Feb 2015
    Posts
    15,271
    Quote Originally Posted by Silverlock View Post
    Yeah, I wondered about that... then I noticed that the voltage measurement was slowing drifting upwards. I turned on a regulated 5V bench supply and measured that and got 5.7V, drifting upwards. Hmm! Turns out it was long past time for the battery in the multimeter to be replaced! Changed that and got a nice solid 5.01V out of the bench supply, and 3.14V out of the Basics battery.
    Arrrgggh! Batteries

  9. #9
    Junior Member
    Join Date
    Jul 2021
    Posts
    3
    Quote Originally Posted by Silverlock View Post
    I've had the same CR2032 hooked up to a Teensy 4.0 for 23 months now. I just popped the top off the case it's in to measure the voltage and it's only down to 2.73V. Not bad for a dollar store battery.
    Yes, but a CR2032 is 6x more powerful than a CR1220 (235mA vs 37).
    I suspect that the CR1220 is not powerful enough to keep up a Teensy 4 for more than a couple of months; if you look at the specs at https://data.energizer.com/pdfs/cr1220.pdf this battery lasts only about one month with a 46uA drain.
    Now, the RTC in the iMXRT106x should consume no more than 6.5uA, but they become 20 if the ring oscillator is active; factor in some other leak and the battery will need replacement quite often.
    I will now try to use a CR2032 and see what happens.
    Alberto

  10. #10
    Senior Member+ defragster's Avatar
    Join Date
    Feb 2015
    Posts
    15,271
    Still good on 1220 after 5 whole days - will check for voltage change when I get back:
    Code:
    RTC has set the system time
    15:12:53 14 9 2021
    Yes, 1220 is tiny

    Battery 'in circuit' when not powered and showing 2.91V today - still good but not 3.2V of 9/9:
    Code:
    RTC has set the system time
    2:09:06 16 9 2021
    Last edited by defragster; 09-16-2021 at 10:10 AM.

  11. #11
    Senior Member+ defragster's Avatar
    Join Date
    Feb 2015
    Posts
    15,271
    Plugged this in again - still good with the 1220 cell: seeing 2.81 volts
    Code:
    RTC has set the system time
    21:28:34 26 10 2021
    Down from 3.20 V on Sept 9th - may not make it much longer ... 47 days

  12. #12
    Senior Member
    Join Date
    Oct 2019
    Location
    Calgary
    Posts
    127
    2.95V on the Basics CR2032 after 48 days.

  13. #13

  14. #14
    Senior Member+ defragster's Avatar
    Join Date
    Feb 2015
    Posts
    15,271
    Quote Originally Posted by bicycleguy View Post
    Thanks ... Saw that then ... though didn't look it up ...

    Quote Originally Posted by bicycleguy View Post
    Thought I was getting defective CR1220 button cell battery's. However, measuring RTC currents:
    T3.2 1.24uA
    T4.0 30uA

    Doesn't seem like much but with a 35mAH, 20mm dia. coin cell battery that's only 48 days vs 3 years.
    Still great for power failures but not so good for keeping time on intermittent use projects.
    ...
    Seems about right - depends on the cell - but doesn't seem much more should be expected from 1220 than 47 days ...

    Though the CR2032 of @Silverlock may not make it 3 years given the voltage drop already ....

  15. #15
    Senior Member+ defragster's Avatar
    Join Date
    Feb 2015
    Posts
    15,271
    Plugged in again - time lost and doesn't even hold a second across repowering ... not measuring a half volt ...

  16. #16
    Senior Member
    Join Date
    Oct 2019
    Location
    Calgary
    Posts
    127
    That earlier 2.95V reading may have been a poor connection on my part. Checked a couple of times over the past week and still reading a steady 3.00V here on the CR2032. Double checked the DVM against a known steady regulated supply (and swapped the DVM battery out just in case). After having a dollar store CR2032 still keep time after 23 months, I wouldn't bet against ~3 years.

    Looking at a typical discharge curve of a coin cell, there's a short initial steep decline from above 3V before stabilising at 3V; maybe I just got nice fresh CR2032s out of Am$$on.

  17. #17
    Senior Member+ defragster's Avatar
    Join Date
    Feb 2015
    Posts
    15,271
    DVM here first suggested 20 mV ... and was moving around ... Pushed the probes harder and it showed just over half a volt - but not enough even to keep it alive on a quick switch off to on ... couting form zero again ...

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •