Author Topic: Using Tadiran TL-4903 with buffer capacitor & LDO for 20+ year coin-cell replace  (Read 1494 times)

0 Members and 1 Guest are viewing this topic.

Offline kievitTopic starter

  • Newbie
  • Posts: 1
  • Country: be
Hi all,

I’m looking to replace a CR2032 in a device with a Tadiran TL-4903 (Li-SOCl₂) battery to achieve a 20+ year lifetime. My device has the following power profile:

Average current: ~15 µA

Pulse current: ~16 mA for short bursts (regular intervals)

Nominal voltage required: 3 V

I’m considering the following architecture:

Tadiran TL-2450 (3.6 V nominal) -> Ultra-low-Iq 3.0 V LDO (TPS78230) -> 220–330 µF buffer capacitor (Murata GRM31CR61C227KE15L) -> Device
       
My questions:

Is this architecture safe and reliable for the ~16 mA pulse loads, considering the TL-4903’s internal resistance and chemistry?

Are there better Tadiran or alternative Li-SOCl₂ cells for thin form-factor, long-life, low-current applications?

Any practical pitfalls to watch for (voltage sag, passivation, capacitor sizing, long-term leakage) for a 20+ year design?

I have provided a screenshot of my measured current profile . I’d like to confirm whether this approach is feasible before redesigning the battery holder.

Thanks in advance!
 

Offline nali

  • Frequent Contributor
  • **
  • Posts: 887
  • Country: gb
Can't help with this specifically, but I can say if this is a commercial project you could contact Tadiran and their tech guys will look at it (I've done something very similar with them recently for a project with a projected 10 year lifespan)
 

Offline edavid

  • Super Contributor
  • ***
  • Posts: 3785
  • Country: us
Can't you get rid of the regulator?  A fresh CR2032 could be 3.35V anyway, so 3.6V is not much higher.
 

Offline Gribo

  • Frequent Contributor
  • **
  • Posts: 677
  • Country: ca
The TL-4903 is an AA cell, and the nominal voltage is 3.6V, If you can use a low Iq buck instead of the LDO it might be a better solution, or better yet, power your board directly from the cell.

The cell should be able to provide the 16mA pulses without any issues. I used the same cell with 100mA pulses with 470uF input capacitance and the voltage dropped to ~3V in the worst case, -20C  and an almost dead cell.

If possible, use the largest buffer capacitor you can fit, Google can't find that Murata PN, but there are 220uF in 1206 and 1210.
I am available for freelance work.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf