Connect all up

[kurt-k3po]

Have you thought about using the Innovate ISP2 protocol for connecting all the pieces together?
Especially for EngineDataLogger-Engine Data and the DashBoard?
On DashBoard side, someone could use existing software to analyze and display the logged data. On EngineData Side it would be more general to adapt/develop own sensor controller.

[techie66]

Long term plan is to use CAN2 protocol to connect sensor controllers. I would have to look again to see if ISP2 is suitable for the dashboard.

[kurt-k3po]

Btw. I'm doing all the RPI tests on a PI Zero W as this is the target platform I will use later on the bike. Is this o.k?
Ah, I see, the GPIO pins of RPi are not 5V tolerant, So I orderd 3.3v 5v level shifter first. In the meantime I try with a voltage divider.

[techie66]

I've ordered a board that is 3v3 compatible to try out, and I'm testing on a pi zero w as well. I want to put the mcp2515 on my custom board to simplify hooking everything up. If I'm reading the datasheet right, I only need a crystal and capacitors along with a few resistors to make it run so I will get that prototype made up soon.
I anticipate that socketCAN will make getting the packets into the DataLogger pretty easy so by next week I hopefully can start trying to implement at least logging for the oxygen sensor.

[kurt-k3po]

Can you provide a link to the board you ordered? I will then give it a try also. You design a custom PCB for the logger? Interesting.
All the CAN HW designs I know of uses the MCP2515 in combination with CAN Transceiver such as TJA1050. The TJA1050 converts the RXCAN / TXCAN of MCP2515 to the CANH CANL differential signal levels.

Ok. I see, using a MCP2562 as transceiver instead of TJA1050
would be smarter, as this chip use 3,3V for VIO (to connect direct to MCP2515) and VDD 5V to drive CANH / CANL lines.

[techie66]

The board- https://www.amazon.com/gp/aw/d/B07VMB1ZKH?ref=ppx_pt2_mob_b_prod_image

Custom board for front controls - https://github.com/techie66/EngineDataLogger-FrontControls/tree/development/hardware

[kurt-k3po]

I tested the waveshare RS485 CAN HAT board right now. This board work without problem with following configuration, as this board have a 12 MHz oscillator.

/boot/config.txt:

dtparam=spi=on
dtoverlay=mcp2515-can0,oscillator=12000000,interrupt=25
dtoverlay=sp1-1cs

[kurt-k3po]

I just wired a MCP2515 CAN Module to SPI0 on RPi using 3,3V for a test.
Edit /boot/config.txt and add the following:

dtparam=spi=on
dtoverlay=mcp2515-can0,oscillator=8000000,interrupt=25
dtoverlay=sp1-1cs

Reboot RPi and start can0 interface in loopback mode to test

sudo ip link set can0 up type can bitrate 500000 loopback on

then start 'candump can0' in one terminal and send with 'cansend can0 001#1122334455667788' from a 2nd terminal.

kurt@k3po-pi:~ $ candump can0
can0 001 [8] 11 22 33 44 55 66 77 88
can0 001 [8] 11 22 33 44 55 66 77 88
can0 001 [8] 11 22 33 44 55 66 77 88
can0 001 [8] 11 22 33 44 55 66 77 88

Even the loopback test works. See tomorrow if I can get some data over the true CAN wires. Not sure, because the tranceiver need 5V and the GPIO pins of RPi are not 5V tolerant. Therefore had to wait until other board arrive

[kurt-k3po]

Here is another log. I did this not on a running engine, but I put the LSU4.9 i a tube and burned with a lighter on the open end. So it was not free to air. This gets a lambda between approximately 0.7 and 1.2. I also add a timestamp to the candump output..

This is the Ignitech WB2 config I used:

The following logfile recorded on RPi with: candump -t z can0

lambda_can.log

Hope, this helps. If I should get the whole stuff to a real running engine, let me know. But this will take some time.

[techie66]

Thanks for the log file Kurt. I'm working on this right now. J1939 is more difficult than I anticipated. the Ignitech protocol is much simpler. J1939 does not give a lambda value, but % oxygen by volume. I am not sure how to convert from O2 % to lambda, so I will get the Ignitech format working first. If you can help me convert the O2 % to lambda, I can get J1939 support too.

[techie66]

I've pushed a new branch 'feature-can' with very basic support to get lambda from mode 1 "Ignitech" CAN frames. It isn't pretty, probably has bugs, but I'd like you to test it and see if it works. It will only work if you keep the CAN ID set to 802 (0x322)

[kurt-k3po]

Jacob, this is really great, thank you for that.
I just cloned the 'feature-can' branch and RPi is compiling at the moment. Make it sense, to install the bluetooth and flatbuffer at this time for this branch? What can be displayed at the 'dashboard' with this branch?
Also, I try to use the RPi GPIO 14+15 with RS232 converter for serial connection to TCIP4. I don't want to use the USB port for that.
My attempt is to get the lambda and TCIP4 logged to csv logfile and display some info on front control at runtime.

ok. Had a quick look to the code. Is it possible to get the 'advance' logged from TCIP4, too? If this could be easily implemented, I would really look forward to this. Then all parameters could be logged which are needed to work on the 3D mapping of the Ignitech ignition in the first time.

[techie66]

Bluetooth and flatbuffers are necessary for the dashboard. Have a look at cb1100f-app.fbs to see what is displayed. The app is not very configurable at the moment so pretty much everything is always shown. It is fairly easy to log the advance from the tcip4, I will start adding that. Please let me know how testing goes with the CAN interface.

[ivanzanzani]

You two are too far ahead ... hats off !

[kurt-k3po]

Jacob, by the way - it's Ignitech WB2 , not Ignitech WC-2 ;-)

[techie66]

Oops. 😆 I was tired.

[techie66]

This discussion has now started to center around CAN support. Let's continue with #4 existing issue.