doppelgrau / esp32-f9p-io-board Goto Github PK

Tests with the oscilloscope show that the inductor in front of the ADS1115 is not much more effective than the RC in front of the LSM9DS1. So to make the BOM shorter changed the inductor to a resistor.

If the xbee-uart is currently connected to the esp32 change to uart2

Look again into the datasheet, I think a bit clearance could help

The front mounting hole to secure the F9P does not fit, movement required

Change the Xbee uart PINs with the pwm pins so there won't be any problems if pulled to gnd during startup
(Test before it not pulled down from the VNH and if PWM s okay)

Some Redesign do allow larger screws, the M2 work but not enough reserves to be comfortable.

adding inductor and supressor diode?

Some removed Parts are still referenced

current position of the switch leaves not enough space for good cable management around the power connector.
So improve free space (top/bottom, wider?) or move power switch

make the edges a bit less sharp
Round with a small diameter or a 45° cutoff (0.5mm-1mm?)

Check if reasonable to isolate the USB from the rest.

Evaluate if the parts over the LEDs can be made/printed thinner.
If printed with transparent filament it should be more clear and with two color prints it would look nicer.

Make sure everything can be tested without the need of external cables (using pogo pins)

The uart <-> RS232 converter supports a power-safe mode that might cause trouble.
Pull the FORCEON-pin to high instead of gnd

Thank you for answering my question!
I noticed a small point.

In the circuit diagram "C35" is "100nF/16V"
But in "Parts.ods" C35's Mouser Partnumber is "603-CC805KRX7R9BB103"
it's 10nF.

I think Row"C35,C47,D2,D5,D9" Colum"Mouser Partnumber-" is wrong.

Connection of MDIO is broken, fix it to enable ethernet

Provide better examples (maybe libraries?) for:

• IMU iron correction
• IMU gyro correction
• using the port expander
• using the VNH7070

• More robust for USB?
• Keep the PTSM? If so Change Variant? SMD/THT

Check if there is an option to put different IMUs on the board and only populate the requested (MMA, LSM9DS1 ...)

Hello doppelgrau.
It is a wonderful work!
I want to use this board for NTRIP Client to F9P, and AutoSteer to AgOpenGPS .
Do you think it is possible in ONE esp32?

To use two ardusimple F9P (simpleRTK2B and simpleRTK2Blite) for RTK + heading with NTRIP a second UART is needed with an interface so it can be connected to the top module.

• If you use an LR kit with corrections coming over XBee, you only need to connect to 1 UART. This UART will give RTK position + heading, but only in UBX protocol
• If you use NTRIP server, you will need 2 connections, one to send corrections to simpleRTK2Blite, and another one to read RTK position+heading in UBX protocol.

Change the part names/values in such a way, all relevant parameters are visible (e.g. for capacitors not only the capacity but also the allowed voltage)

Evaluate if a ESP32 with more flash is an option, so larger Software can use OTA-Updates

After everything is done, update gerber, bom, ibom, seedstudio files

some ldo or a reference voltage

• Same behavior for all connectors
• All pads on USB for better mechanical connection
• maybe less paste for the port multiplier

Related a bit to #37, check if protection against to high/reverse voltage should be improved

Evaluate if there could not be some sort of support on two or three sides, so the pcb is a bit secured during installation.

Idea: plane with axes, in software rotate pitch/yaw/roll until plane fly in direction of travel

It's very hard to press the Button if a F9P Board is mounted.

Evaluate if there are any drawbacks (for usual motors should be more than enough).
If no drawbacks, eliminate or use for additional (digital) outputs

Divide the power supply of the PCB from the Power suppl to the motor driver.
So the motordriver can be easily driven with an other voltage (e.g. 24V)

including drill files

Get two output pins for the slect on the VNH and three analog inputs (or check if two would be enough)

Currently placed near the RS232 connector. but connectors and cables introduce metal in the area that can disturb the magnetic field after calibration.
Maybe better placed under the F9P, magnetic error should be more constant there => after calibration not a huge problem

Include a good documentation what the board has / what parts, internal and external inout

README.md

one uart (IO0/2) next to the arduino headers on two pins (J2) for the new simpleRTK2Blight. Since the two pins have a special meaning during boot, pulling RX to GND during startup results in the ESP going into programming mode.

pulling TX to GND ,I think

The following labels are inverted relative to others:

• ESP-F9P-IO-Board
• Version 2.2
• GND / 12V / M1A / M1B
• Uart / RX / TX

Two external Ethernet-connections would make daisy-chainning easier

Add some fidducials, so a pick & place machine might be used

• Improve Termination 2x60 Ohms to Capacitor (one end GND) instead of 120Ohm. Reduces Common Mode Problems
• Check ESD, improve with something like a PESD2CAN

Make sure all SMT-connectors have the Stencil/Paste layer activated for the footprints