diydrones / pxf Goto Github PK

While I was reviewing the design, I have noticed that the Serial 3 (sheet nº2) has not GND signal connected. This could be the reason why the port can't power up connected modules.

In this image, it can be appreciated:

@proficnc OUTPUT_EN (signal used to enable/disable the voltage translators) is connected to two different MIC5332:

One of them is used to generate the delay but the other one? Could you explain why?

For SPI function:

• Pin 24: SDI
• Pin 23: SDO_XM
• Pin 22: SDO_G
• Pin 21: SPC

Thereby the connections should be:

• SDI (LSM) -> SPI0_MOSI (P9_21)
• SDO_XM (LSM) -> SPI0_MISO (P9_18)
• SDO_G (LSM) -> SPI0_MISO (P9_18)
• SPC (LSM) -> SPI0_SCLK (P9_22)

Please refer to this thread for a discussion.

Resistor 101 causes a brownout, ensure this is removed for version 1.1

Change resistor values to suit the 1.8v max limit

PWM 5-12 share pins with the ROM Bootloader....

Comment removed, marking as closed.

We're having issues with the MPU6000 gyros and accels values. The issue might be related to the fact that C409 and C414 don't follow the datasheet's specifications. Instead of 2200 pF they should have a value of 0.1 uF.

This is a guess, it doesn't necessarily mean that the samples will improve. We might have a software bug.

According to the Package Specification the REGOUT capacitor should be 0.1 uF but a 2200 pF has been placed in the board:

@proficnc why such change?

Jumper on the 3.3v enable (J1) needs to be removed for revision 1.1. as a fix, it should be set to position 2-3, rather than the default 1-2

(left side corresponds with rev. 1.2, right side rev. 1.4):

In rev. 1.2 the VDD_ADC is not connected properly (that signal provides 1.8 V, the voltage divider is not needed).

OE from the voltage translators is connected to one of the pins in the P8 header and activated by one of the MIC (LDO) ICs.

Test points removed. Sensors need adjustments (rev. 1.4 sensors work fine).

Capacitor changed, signal added to POR2

Resistor changed from 10K to 100K.

Straight connection of the Power Module DF13-6 pins. Add CAP to the LDO. Add more resistors to the CURRENT and VOLTAGE sense lines.

i2cdetect -r 1 command gives
0 1 2 3 4 5 6 7 8 9 a b c d e f
00: -- -- -- -- -- -- -- -- -- -- -- -- --
10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
50: -- -- -- -- UU UU UU UU -- -- -- -- 5c -- -- --
60: -- -- -- -- 64 -- -- -- -- -- -- -- -- -- -- --
70: -- -- -- -- -- -- -- --

This is clearly a wrong response the expected pings were to be from addresses 0x55(Toshiba LED) and 0x77 (Barometer) instead we are getting responses from 0x5C and 0x64. Until this post the result is proved to be this way on two boards on Siddharth's and Victor's board.
i2cset and i2cget also fails to work for all the mentioned addresses.

I went through the schematics and the changes in the PXF for the ADC side. We expect 5 V at the DF13 pins, right? If so, that'll put the voltage at BATT_CURRENT_SENSE and BATT_VOLTAGE_SENSE at 5/3 = 1.6 V which gets into the 0-1.8V range of the BBB ADCs.

Can you confirm this reasoning?

Remove option of servo rail power backup, connect main power directly.