px4 pixhawk 2.4.5 hardware related sensors

[Alex Angel]

Hy all,

I have recently noticed something strange about the pixhawk 2.4.5 hardware.

One of the sensor on the pixhawk is called a LSM303D ( a 3D magnetometer and accelerometer)

I was looking at the pixhawk schematics that are to be found on github posted by Lorenz Meier or 3DR and saw that the sensor surrounding parts are different then what the manufacturer suggest. ST shows a completely different arrangement for this.
Wondering why did the people/team who build pixhawk chose to reduce or replace this surrounding parts with other values. I wanted to hear a VALID explanation from an expert on this matter and how it affects or improve the flight controller.

Please check the file in attachment. 

Greets

[Tom Pittenger]

I don't have the answer to this, but I do know that demanding help will not get you answers faster. No one is obligated, they help because they like to help.

--
You received this message because you are subscribed to the Google Groups "drones-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to drones-discus...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Joonas Melin]

Hi Alex,

Since I didn't design this, I cant give you a definitive answer, but from the looks of it, they have opted for more capacitance on the capacitors.

Now if you read through the datasheet of the sensor, I am quite sure that it will tell you that those capacitors are some form of smoothing capacitors that filter out noise/make the operation of the sensor smoother. Generally speaking, opting for larger capacitors results in larger reserves for smoothing out noise.

Because the pixhawk operates in noisier environment than normal, I believe that opting for larger capacitors is a reasonable design decision.

What leads you to believe that this would not be a reasonable choice?

[Jeffrey Wurzbach]

It's been a long time since I've touched the Pixhawk 1 drawings.  I don't recall why we set the values that way off hand.

 

I know a couple of the caps on the MEMS elements are there to make internal boost regulators for bias voltages that are used internally on the devices, but without reading the datasheets again, I can't say for sure which cap did what.  

Also, what's the motivation for keeping the LSM303D and L3GD20 on the board?  

--

[Alex Angel]

Hello,

Thank you all for the information,

I have added the datasheet in attachment from ST for the LSM303D and also again the picture showing the differences. Not just the red circled cap value but allot more differences.

If you like to have a look check attachment 

Greets

Op dinsdag 2 augustus 2016 03:44:52 UTC+2 schreef Jeffrey Wurzbach:

[Alex Angel]

Hello Tom,

I just posted this question several days ago I don't know why you jump to this conclusion of demanding answers just like that without even introducing yourself first. I just post my question like all other people nothing more.

There is a difference between demanding answers and not to give an answer. I don't have any clue why is it that you choose to pick on me for this Im sure you have good reasons.

greets

Op maandag 1 augustus 2016 12:04:01 UTC+2 schreef Joonas Melin:

[James Brown]

The capacitors (on VDD / VDD_IO pins) values are not important, the smaller 100n (0u1) C507 is for decoupling, and 10u C506 provides bulk capacitance for power smoothing.

The value of the set/reset capacitor is not critical, as long as it is 'big enough'. It is used to provide a current burst across the sensor, as described here link

Similarly, the 'reservoir' capacitor is simply larger (10u instead of 4.7u), so it will work fine.

You should be aware that value of a ceramic capacitors is not especially accurate, in particular with DC bias applied (even when using X5R or X7R dielectric). There are very few instances where changing the (rated) value by a factor of 2 will make a large difference to circuit operation, such a circuit would be prone to reliability issues.

[Alex Angel]

Hello,

Yes but some values here were changed by a factor of 10 ! and some parts /caps  were not even placed as suggested by the manufacturer. What about the fact to don't place a cap were it is suggested ... ?

greets

Op dinsdag 9 augustus 2016 03:06:40 UTC+2 schreef James Brown:

[James Brown]

What components are you talking about? All the caps shown (4) in the application example are included, and approximately the same in value, C1 = 10u vs 4.7u, C2 = 470n vs 220n. C3, 10u and C4, 100n are the same as indicated.