LPS33W Reflow profile unclear

[SGano.1]

Hi.

According to the DS of the LPS33W - the Reflow profile should be according to JEDEC J-STD-020 , with no actual determination of the profile.

we had several boards assembled , and with very little force applied, the sensor disintegrated - the metal part from the plastic socket.

I suspect it might be the reflow profile that weaken the glue and caused this disintegration.

Is there a recommended profile for the LPS33W?

We have used before that the STEVAL-MKI205V1 - which is the LPS33W EVB ,

And had no such issue.

The pic below is of an actual reflow performed on the boards.

Thanks.

[Eleon BORLINI]

Hi @SGano.1​ ,

Which soldering technique are you using (convection / vapor phase)?

The profile looks overall good, especially in the first part.

However, comparing it with the one in AN4450 application note (Hardware and software guidelines for use of LPS25H pressure sensor... well it's another pressure sensor but the soldering profiles are all very similar):

I would suggest you to increase the speed of cooling ramp from 1°C / sec as in your case to at least 2°C/4°C / sec as in the above picture.

Moreover, in AN4450 application note you can find other considerations to be taken into account:

In using non self-cleaning solder paste, proper board washing after soldering must be carried out to remove any possible sources of leakage between pads due to flux residues. However, take care not to perform the cleaning process on top of the pressure sensor.

Let me know if you can try the modified solder profile.

-Eleon

[SGano.1]

Thanks Eleon.

I will check wit my SMT vendor for profile modification.

Is there another explanation Why would the LPS33W Disintegrate that way?

[Eleon BORLINI]

Hi @SGano.1​ ,

I checked internally and it could be that -in rare cases- due to tight tolerances at mounting on the specific application platform, the cap could receive lateral stresses and worse its adherence.

Which is the failure rate of this issue? Do you have other components near the sensor on the board? And could you estimate the force you apply while you are soldering the device?

If you could better tune the soldering / mounting process, you should solve the issue since it only occurs at the high soldering temperatures in specific "surrounding" conditions.

-Eleon

[SGano.1]

Hi @Eleon BORLINI​ ,

All the LPS33W that were assembled, have Disintegrated.

Just a little touch disintegrates the Sensors.

Moreover - after a rework process for replacing the faulty component - two components gave false data (80mBar ambient pressure instead of 1010mBar, and false data of temperature too.

The I2C comm is OK and the sensor communicates with the uC- but gives the false data.

Attached another pic from another Board.

[Eleon BORLINI]

Hi @SGano.1​ ,

but does the chip decap during the soldering process or after?

-Eleon

[SGano.1]

Hi @Eleon BORLINI​ ,

This happens after the soldering.

After it comes out of the SMT,

It looks intact, even though it is held only by the Gel inside.

[Eleon BORLINI]

So it is something happening during the soldering process that is weakening the capping for some reason...

Do you perform an IQC to check the integrity of the devices before soldering?

Could you try modifying the cooling step of the reflow profile as suggested above?

-Eleon

[Eleon BORLINI]

Hi @SGano.1​ ,

did you eventually manage to try to slow down the cooling ramp of the soldering profile?

-Eleon

[SGano.1]

Hi @Eleon BORLINI​ 

We have not performed another SMT assembly yet.

We had all our boards fitted with the LPS33W  reworked,

and so far so good.

Didn't you mean a faster cooling ramp by

"I would suggest you to increase the speed of cooling ramp from 1°C / sec as in your case to at least 2°C/4°C / sec as in the above picture" ?

[Eleon BORLINI]

Hi @SGano.1​ ,

sorry, a faster cooling ramp, you are right.

Although our internal qualification didn't show such behavior of LPS33W  cap for standard SMT mounting, we will go deep in this topic.

Thank you for having reported your issue.

-Eleon

[SGano.1]

Hi @Eleon BORLINI​  ,

We are experiencing a lot of failure rate by the LPS33W  -some times the device is communicating via I2C - but the values of pressure and temperature are not Physically possible (garbage data).

after replacing the device- the data is valid.

The LPS33W  seems very sensitive mechanically and we had to replace at least 8 pcs during our system's debug.

Is there a recommended method of strengthening the LPS33W  so it is not damaged by any minor lateral force?

are there particular batches that are more vulnerable ?

We had no issues with the EVBs.

Please advise.

[Eleon BORLINI]

Hi @SGano.1​ ,

>> Is there a recommended method of strengthening the LPS33W so it is not damaged by any minor lateral force?

You should try not to apply any side force during (or just after) the soldering process...

I have done what in my power contacting our internal pressure sensor quality team and the feedback from them is the one reported in my comments above.

The only additional thing I can ask you is the lot ID of the failing parts, so we can correlate any eventual weakness in production among the assembly lots with equal or similar ID code.

-Eleon

[SGano.1]

Hi @Eleon BORLINI​ 

Please refer to the attached photo of the reel.

[Eleon BORLINI]

Hi @SGano.1​ ,

unfortunately this is not the bulk / lot ID, but the distributor commercial references (Farnell)...

You should check with Farnell, I believe...

-Eleon