Developer

3689432775?profile=originalHi there! Finally you can pre-order the ArduIMU V3 here.. You can find more info abut it in my last post here

The final version of the board for full production is here, I'm expecting to start shipping them this Friday or Monday. 

We are still working on documentation and giving the final cosmetic touches to the example code to have it ready before you have one in your hands and at the same time we are shipping samples to developers to do a full DCM integration.

 

From now on our new color will purple! All our new products will feature a Lilypad style look! 

 

(Note: this is a stand-alone IMU, and is not designed to be a replacement for the "Oilpan" IMU shield of ArduPilot Mega.)

E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • Ok. just to clarify the comments. The W5100 chip requires not just the CS to be manipulated, but the SPI Enable pin too at the same time. Could be a bug, could be the intention. I got this from WizNet support. When CS goes high, SPI_EN should go low and vice versa. This means that if you have a board where the SPI_EN is directly connected to 3.3V and not to an MCU pin, then you're likely out of luck.

    The good news is that the Sparkfun's W5100 doesn't have this directly connected, but does use an inverter on the CS that disables or enables SPI on the chip. Probably they've figured this out:

    http://www.sparkfun.com/products/9473

    I went ahead and ordered this instead. Haven't received it yet, but I'm convinced that'll resolve my problem.

  • Just some update. It's not the SPI wire length apparently. I cut the length short to 4cm, but the problem remains. It appears that the SPI MISO line on the other chip is interfering with the MPU6000. I tried different SPI speeds to no avail. Double checked that the SS is not low when the other chip starts talking. I can possibly try a couple of delays between the uses of the two chips to see if that helps.

    Anyone some interesting ideas I got try?

    The chip I'm using is:

    http://www.lipoly.de/index.php?main_page=product_info&products_...

  • I  got an ArduIMU v3 here and I'm happy with it so far. I'm trying to hook it up together with an external SPI device for communications, which means having 2 devices on the SPI bus. I noticed that the SPI bus is quite sensitive to noise. This noise manifests itself by corrupting the data stream from the MPU6000 back to the master (MISO), which means the MPU6000 doesn't even initialize or outputs spurious data if it does, often resulting in a processor hang after 5-10 secs.

    When the other device is fully connected, but the MISO is removed from the arduimu (not the slave device), the MPU6000 works reliably.

    I just wanted to check if others are getting similar results when anything is connected and if there are ways to significantly reduce the effects of this other than shortening the wires (extra resistors, etc.). My concern is basically that whenever it works at home, it may still start failing in the air when the ESC or other components start radiating RF noise. So any solution that's better than just shortening the wires is preferrable for me.

    At the moment, the length is around 12-15cm, which is very unlucky due to the vicinity of wireless here. A much better length would be 2-5cm, which I can only start testing after tomorrow. If anything else can be done to further reduce the noise, I'd be very happy to hear it.

  • Thanks jordi! Do you know the part number? I
    would like to know its specs.

    Thanks!
  • Developer
    @logan, the native freq is 16000000... (16Mhz)..

    Regards
  • Does anyone know what the native crystal frequency for the v3 board is? The value I'm looking for is F_CPU which is used by delay.h and others for proper timing calculations.

    Thanks,

    Logan

  • @Harry I am quite interested. I'm currently working on getting the umpl project from Invesense to work I'd be happy to share notes.

  • I found it, I changed GRAVITY to 4096 in DCM and in the scaling of accel vector, I used 1024/981.  IT WORKS. :).  Pitch and Roll are very very close to dead on.  The Compass adjusted yaw drift seems to work, doesnt wander around but if it doesnt move the heading returns to 0.  I expected the Yaw indication to be the heading.  It is the heading unless it doesnt move, kinda puzzling.  I suppose for a plane to be static in every axis, it would be on the ground.  If anyone would like to see what I cobbled, I can send you the code. 

  • Developer

    eagle123, I'm totally with you... Is the only way you learn while you get fun.

  • Developer

    eagle123, yes..

This reply was deleted.