220nF capacitor ESR LSM303AGR

rvanvreden · ‎2020-04-27

It is stated that the 220nF capacitor should have a ESR smaller than 200mOHm. At what frequency is this rated? not sure what frequency the LSM is switching this capacitor.

I had a look at the EVAL MKI72V1, but that one has an error in the BOM. Any suggested part number? regards, rob

Uwe Bonnes · ‎2020-04-27

ESR is at low frequency. Impedance rise at high frequence for X and COG ceramic is caused by ESL. ESL gets bigger with size and bad layout practice as long traces between pin and capacitor.

rvanvreden · ‎2020-04-27

hello Uwe, Thanks for the reply. The LSM303 is used in a commercial product. we have great difficulty to get it soldered properly in the engineering stage. so soldering with a hot air gun. for some reason, the magnetic axes totally mixup values. looks as if the sensor is broken. totaly unusable. I gues that 1 out of 10 works in this phase. So I was trying to find oud if maybe something is wrong with the layout of the board. I also used the EVAL board. that one seems fine. Speaking of layout; ST mentiones in the datasheet that copper traces must me short as posible, but in their pcb eval design, they are quite long. I have tried both pcb layouts. no success. So that is why I was wondering if maybe the Capacitor is the culprit. regards, rob

Eleon BORLINI · ‎2020-04-28

Hi @rvanvreden , you should take care in using the hot air gun for soldering purposes (instead of a standard reflow oven) on a device whose absolute maximum rate are -40 to +125°C, especially for the magnetomer, which could suffer from magnetization issues. You can btw try to perform the Magnetometer self-test as described in the LSM303AGR AN4825 application note, p.21, to check if the damage is permanent or not. Regards

rvanvreden · ‎2020-04-28

Hello Eleon,

I have also tried ( several times) to solder in the a small reflow oven. the results were not better. Yesterday, I have mailed to ST for support. The temperature range you are referring to is the storage range.

We do have once in a while a sensor that delivers the right data. we will try the selftest and see what that tells us.

regards,

rob

Eleon BORLINI · ‎2020-05-05

Hi rob, the problem with the hot air gun is that it's hard to control the directionality and the exposure of the device to the hot air flux, differently from the reflow oven which is more uniform. Maybe the magnetic sensor is more susceptible of a long exposure to temperature higher than the maximum storage range: the magnetometer technology is related to AMR and not to moving silicon masses like the standard motion/pressure/microphone sensors. Regards

rvanvreden · ‎2020-05-05

Hello Eleon,

You are probably right. Thing is though, that I work with electronics over 35 years, and solder a lot of those tiny footprint IC chips. Never have any problem with it. After the engineering and test phase, the boards are produced in a factory.

But now, it looks like there is no other option then to have them produced, even we are still in the engineering phase. I have hooked up the ST eval board to our board( SCl/SDA) and that works fine. ( in stead of the own soldered LSM)

would be nice to sit down with a ST engieer to discuss this.

regards,

rob