Running STM32U083 directly from Li-SOCL2 battery

[Magvel]

I am using STM32U083 MCU in a project that runs on a 3.6v Li-SOCL2 lithium thionyl chloride battery (OC voltage 3.65V / Nominal voltage at 1mA 3.6V).

Currently I have used a LDO of 2.5v to power the MCU, but I was thinking if we can eliminate the LDO and directly power the MCU from the battery.

Going through the datasheet I have found that the recommended input supply voltage is 1.55v-3.6v, and max is 4.0v, I have seen a mass produced device using the same battery directly on a STM32L053 MCU. I am expecting the same might be possible with U083 as well.

Please suggest if this is okay or is there can be any potential issues that can happen.

 

[Ozone]

> Going through the datasheet I have found that the recommended input supply voltage is 1.55v-3.6v, and max is 4.0v, I have seen a mass produced device using the same battery directly on a STM32L053 MCU. I am expecting the same might be possible with U083 as well.

You might be right, and the short period above the 3.6V rating does not damage the MCU.

However, I think you need to consider the other end of the discharge curve.
When the battery voltage drops below the brown-out limit, your device will be dead - until the battery is replaced.
With some battery management system, you could avoid that, and schedule some actions before charge runs out.

Obviously this is not a point many mass products deem worthwile.

[Andrew Neil]

---

@Magvel wrote:

max is 4.0v

---

As your screenshot shows, that is the Absolute Maximum - that's not a guaranteed operating voltage. 

 

---

@Magvel wrote:

I have seen a mass produced device using the same battery directly on a STM32L053 MCU.

---

As you know, many mass produced devices prioritise low cost against reliability, etc.

They're taking a gamble - do you want to take same gamble?

Are you aiming at the same market - where minimum cost trumps all else?

Sure, you'll probably get away with it, most of the time - is that sufficient for you?

[Magvel]

Hi @Andrew Neil, the product is competing in the same low cost market, but I didn't wanted to gamble so I asked here in the community, as I have mentioned even though the absolute max is 4.0V our system absolute max is 3.6v-3.7v in normal operation it will never exceed 3.6v,  which is just at the corner of operating voltage(3.6v) given in datasheet. So I was doubtful if we can use the MCU in such a corner condition safely. I believe it is perfectly safe as the absolute maximum is around 4.0v as confirmed by @Ozone.

I will anyway test and validate the product working thoroughly before taking a final call.

[Ozone]

> I believe it is perfectly safe as the absolute maximum is around 4.0v as confirmed by @Ozone.

I cannot "confirm" anything in this regard, just interpret the datasheet.
My statement "you might be right ..." is not quite a confirmation as such.

Do your due diligence and make some tests, with different MCUs and fresh batteries.
If none dies, you are likely fine.

Similiar with the behavior at battery EOL. 
Perhaps it doesn't make a difference to the customer, and your fine as well.

I had been involved in projects for commercial devices that had to (permanently) store parameters at power-down, because a device standard required it.
Nothing of this kind is possible with your device, it will just stop working suddenly at one point.