Potting of the LIS2DW12 related to increased current consumption issue.

[Stefano_82]

Hi All,

My name is Stefano, and I developed firmware for an embedded sensor that integrates the LIS2DW12 MEMS accelerometer. We encountered the following issue:

1. At the end of the firmware development, we tested the current consumption across many of our new sensors, and the results were as expected;
2. The new firmware was then put into production;
3. After a few months, we noticed that some sensors had a lower-than-expected battery level, despite their typical lifespan;
4. We retrieved those sensors from the field for analysis;
5. Upon analysis, we discovered that the issue was due to abnormal current consumption by the sensor, with an average current (Iavg) of approximately 300 mA instead of the expected 7 mA;
6. The sensor's current consumption returned to the normal range when the firmware powered off the accelerometer;
7. After extensive investigation, we determined that his abnormal consumption is due to the resin potting phase,

 

Main resin infos are follows: 

shore A hardenss of 74-80,

Dielectric Costant 7-8,

Dielectric Strength of 22 to 23 kV/mm,

water absorption rate of 0.2%-0.3% after 24h at 25°C

 

• In order to isolate resin hardening, we proceed with pre-conformal coating above the LIS using a special silicone specifically made for electronic coatings

Main resin infos:

Name MOMENTIVE TSE397 seires

Shore A hardness of 20

Dielectric Costant 2.9

Dielectric Strength of 22 to 23 kV/mm

• We tested the sensor's current absorption without conformal coating and everyting went well;
• We then applied conformal coating only above the LIS accelerometer;
• We let the silicon dry for 2 days;
• We tested the sensor's current absorption with the conformal coatin eand 100% of the sensors where affected from abnomal current absorption.

 

Can you tell us what we can do in order to resinate our sensor and avoid axtra current absorption?

P.S. attached you can see one of the conformal coated sensors.

Thanks in advance.

 

 

[Federica Bossi]

Hi @Stefano_82 ,

300mA is a very high value that we do not expect can be handled by the device. By retesting do you always get this value?

[Stefano_82]

Hi Federica,

Thanks for answering my question. As I mentioned in my first post, I have a lot of sensors with very strange behaviors in my office.

Normally, our sensors work at 7-10mA, but the faulty sensors (only with the integrated accelerometer) have different absorption behaviors.

Most of them work at 80mA, and if I touch the resin, the current absorption rises to 200mA and slowly fall down to 80 mA. Some have an absorption of 200mA, and others at 300mA.

Unfortunately, I'm currently out of the office for work and can't provide you with the log of the sensors with 300mA of absorption. However, I can give you a log of one of them that I stored locally.

This is the current absorption with the sensor fully resinated and the accelerometer powered up in sleep state:

Normally, its current absorption is under 10mA, and if I power off the accelerometer its absorption returns to normal ranges.

 

Do You have te possibility to open the .ppk2 extension files?

Is there a possibility to have a private ticket discussion?

 

Thanks in advance.

[IIvan.22]

Hi Stefano,

Do you still have this issue? We have a similar one LSM6dsm instability and high current consumption. You can try our solution (adding 300pF capacitors), or, if you have pull-pus or pull-downs, to replace them with direct driving by push-pulls in the software. Do you have a schematic that you can share by any chance?