F103 dual channel wideband controller + EGT + 2 x AUX In + 2 Aux out

[Krakert]

On the github of the controller there is this section:

https://github.com/rusefi/rusefi-hardware/tree/main/lambda-x2#esr-measurement-crosstalk-between-channels-rev0-and-rev1

Does this also apply to the one I got from you Andrey?

[AndreyB]

My name is Andrey. I've removed resistors and marked your unit "4.9"

The comment relates to the dream of universal hardware where same exact unit would work on both 4.2 and ADV which is still not the case.

[LemonX]

Hi!

I have questions about dual channel controller.
When TS is set to can EGO sensor, wideband controller sets both heating outputs to 100% duty cycle. This is when engine is still off.
In wideband controller TS both AFR0 and AFR1 changes from red no supply to green ok.
This should not be that right? without can connected it doesn't do that.

Screenshot 2023-08-01 233107.png (268.43 KiB) Viewed 47747 times

Other issue is CRC check error when changing settings. Maybe known problem?

Screenshot 2023-08-01 233132.png (99.52 KiB) Viewed 47747 times

Thanks!

[AndreyB]

First thing is a known issue just created issue https://github.com/dron0gus/wideband/issues/12

[LemonX]

Thanks!

I notice one more thing, ECU tool reset to DFU seems not working.
Reset ECU and Reset to OpenBLT works.

[AndreyB]

ECU tool reset to DFU seems not working.

Please help by creating a github ticket in the same issue tracker.

[Krakert]

Im having trouble with getting the controller to work properly.
Via bluetooth I can connect to the controller, and when I start the engine, the lambda sensor start to heat up. But the readings go all over the place.

I have checked the wiring a couple of times, all good. Could somebody check out the logs I have attached, maybe the tell something.

[Dron_Gus]

Im having trouble with getting the controller to work properly.
Via bluetooth I can connect to the controller, and when I start the engine, the lambda sensor start to heat up. But the readings go all over the place.

I have checked the wiring a couple of times, all good. Could somebody check out the logs I have attached, maybe the tell something.

Hi. Looks like known bug. Already fixed in last firmware.
Please update device with attached FW.

[Krakert]

Uploading the firmware, can that be done via UART, or do I need to break out the ST-link for this? Not a lot of information is available on the Github page

[AndreyB]

Yes UART would work see https://github.com/rusefi/rusefi-hardware/commit/c8f628fe3bf8f1a6a0dce1cfb1599914ba47ed9c

[Krakert]

Ahh oke! DFU mode over UART, what programmer do I need than? FTDI will do, or a ST-link?

[AndreyB]

Two separate options not to mix

* any uart dongle on uart pins
* st-link if you have tc2030 spring loaded cable

[Dron_Gus]

Uploading the firmware, can that be done via UART, or do I need to break out the ST-link for this? Not a lot of information is available on the Github page

Yes. Documentation is weak. Sorry, someday I'll copy following instruction somewhere on github.
I'm not sure about your current FW, if it has OpenBLT enabled or not, so lets flash it using STM32 DFU mode (internal ROM bootloader) over UART.
You will need USB to serial adapter with 3.3V levels (not USB to RS232). I would recommend FTDI based device.

1. Disconnect main connector. If you are going to apply +12V power through main connector - make sure that nothing else is connected to WBO (LSU sensors, any load, etc). But I recommend flash using +5V from USB.
2. connect USB to uart connector to J3 connector: gnd, rx and tx.
Don't forget to cross Rx-`Tx (adapter's Tx goes to WBO's Rx, WBO's Tx goes to adapter's Rx).
If you going to power WBO from USB port - also attach 5V line. Do not connect +5 from USB adapter if you are going to use +12V supply through main connector.

20230923_125247.jpg (721.15 KiB) Viewed 45506 times

3. Download and install STM32 Flash Loader Demonstrator. (Alternative tool is stm32flash - not covered in this instruction)
4. Figure out USB to serial serial port number:

device manager.png (113.35 KiB) Viewed 45506 times

5. Start Flash Loader Demonstrator GUI application, select correct COM port, reduce timeout to 1 second;

stm32 flasher.png (18.24 KiB) Viewed 45506 times

6. Press BOOT0 button on the bottom of PCB. Or short BOOT0 PCB jumper if button is not populated on your board.
7. Apply power (while holding BOOT0 button) to board and press Next in application.
Note: this is a tricky point if you have BT module on your WBO. Seems BT module sends some data over UART to STM32 and this cause STM32 to start listen to other interface and PC app fails to connect with timeout. So you have to be fast, apply power and press "Next" in app while BT module is loading. (this issue will be investigated and fixed in better way)
After app detects chip you should see something similar to:

stm32 flasher 2.png (30.42 KiB) Viewed 45506 times

8. Press next, select "Download to device", select wideband.bin file from archive above (or from github).
NOTE: currently github produce bad .hex files, so use .bin!
Select "Jump to user application" and "Verify after download"

stm32 flasher 3.png (22.76 KiB) Viewed 45506 times

9. Press Next and wait for flash/verification ends.

stm32 flasher 4.png (19.18 KiB) Viewed 45506 times

[Krakert]

Yup that worked great, updated the firmware.

And was able to connect to the controller via bluetooth, after getting three different laptops .
I keept getting a error, maybe somebody knows what's up with this one

Screenshot 2023-09-23 192932.png (108.52 KiB) Viewed 45401 times

But the probes did not heat up, do they only start warming up after the engine is started? Today I was unable to start the engine, it looks like the batteries have died on me, but I was wondering if it was possible to start the cycle of the probe to see if the readings make any sense, and stay stable, not as in the log I shared.

[Dron_Gus]

Yup that worked great, updated the firmware.

And was able to connect to the controller via bluetooth, after getting three different laptops .
I keept getting a error, maybe somebody knows what's up with this one Screenshot 2023-09-23 192932.png

But the probes did not heat up, do they only start warming up after the engine is started? Today I was unable to start the engine, it looks like the batteries have died on me, but I was wondering if it was possible to start the cycle of the probe to see if the readings make any sense, and stay stable, not as in the log I shared.

Good.
Please also update ini file in your project to match FW version https://github.com/dron0gus/wideband/blob/master/firmware/ini/wideband_dual.ini .

WBO should start heating probe(s) as soon as heater voltage rises above 9.5V if not forbidden by CAN message from ECU.
So if you have RusEFI ECU better to disconnect it for test. It may disable heating on low battery or stopped engine. I'm not sure about its logic.

BTW, you can use J3 connector (without +5V line) to connect to TunerStudio, instead of BT for better stability and refresh rate. But BT does not look like your current problem.

What LEDs activity do you see on WBO?
Heater LED ("Heater L"/"Heater R") should start from low light increasing up to bright.
Status LED ("L Stat"/"R stat") blinks fast while heating up, then blinks slow. In case of error it blinks fast N times then off for 2 seconds, where N is an error code number. See https://github.com/dron0gus/wideband/blob/master/for_rusefi/wideband_can.h#L33 .

And debug log from TunerStudio can be very useful. File is named TunerStudioAppDebug.txt and located inside TunerStudioProjects directory, which is located in home directory in linux. Should be somewhere in My Documents in Windows:
1. Close TS
2. Delete this file
3. Open TS
4. Set "Communication" -> "Comm Debug Log"
5. Open your WBO project. Wait 10..20..30 seconds.
6. Close TS.
7. Attach TunerStudioAppDebug.txt to this thread.

[Krakert]

I have disconnected the CAN wires and the module started heating all by itself, and AFR was al good.
With a engine running I have two AFR and they are a bit apart, so need to fix that, but channel 1 has a big delta in the readings (about 0.1) compared to channel 0 (about 0.09).

Via the can bus, is it possible to get both AFR send to the Proteus ECU?

[Dron_Gus]

I have disconnected the CAN wires and the module started heating all by itself, and AFR was al good.
With a engine running I have two AFR and they are a bit apart, so need to fix that, but channel 1 has a big delta in the readings (about 0.1) compared to channel 0 (about 0.09).

Via the can bus, is it possible to get both AFR send to the Proteus ECU?

Yes, Proteus can log both channels. "Air/Fuel Ratio" and "Air/Fuel Ratio 2". AFAIR you will have to set "Enable CAN Wideband" to true in Sensors->EGO sensor.

Your log look like two sensor are in quite different environment. Too different heater PWM and Pump current, and AFR of course. Can you show your setup?

[Krakert]

I don't have a recent picture, but both sensors are placed just after the Y junction where the 3 exhaust pipes meet. The straight section just before the next bent.

Could it be possible that one bank of cilinders is running so much richer/leaner? I don't really have synced all the cilinders yet.

IMG_20210829_152451.jpg (1.51 MiB) Viewed 45245 times

[Dron_Gus]

I don't have a recent picture, but both sensors are placed just after the Y junction where the 3 exhaust pipes meet. The straight section just before the next bent.

Could it be possible that one bank of cilinders is running so much richer/leaner? I don't really have synced all the cilinders yet.

First of all, I do not have too much experience with actual tuning. I'm still in progress with my first car/engine. Maybe some of more experienced forum members will comment this.

Screenshot from 2023-09-26 09-08-09.png (36.68 KiB) Viewed 45112 times

I see 10 vs 18 AFR on your log.
On my H6 I saw 10 vs 15 AFR when two cylinders in right bank had no spark. But my H6 has dual throttle that needs no synchronization.
Once I saw 15 vs 13 AFR when exhaust has leak in left bank and was sucking fresh air.
You can try swapping left and right channels on WBO (swap sensor connectors) This should help sorting out any possible HW/SW issue with WBO controller and wiring.

[Dron_Gus]

7. Apply power (while holding BOOT0 button) to board and press Next in application.
Note: this is a tricky point if you have BT module on your WBO. Seems BT module sends some data over UART to STM32 and this cause STM32 to start listen to other interface and PC app fails to connect with timeout. So you have to be fast, apply power and press "Next" in app while BT module is loading. (this issue will be investigated and fixed in better way)

The root cause of this is issue is floating BOOT1 (PB2) pin. To enter DFU it should be pulled low.
On daul channel WBO rev1 this pin is used as bias current source for LSU4.9 and goes to connector through resistor. So here is dirty workaround:

20230930_110548.jpg (690.09 KiB) Viewed 44905 times

Connect A2 pin of main connector with GND (J4 AUX connector pin 10)

[Dron_Gus]

I notice one more thing, ECU tool reset to DFU seems not working.

Should be fixed now.

[Krakert]

I did some more testing today.

I swapped both sensors and still had a problem on the right side of the engine. Checked the spark plugs, and two of the three on the right side were wet. So I swapped the whole right bank of spark plugs. With great succes, both sensors read about the same (within a few tenth of each other)

Engine also runs better and with more power, great succes!

I notice one more thing, ECU tool reset to DFU seems not working.

Should be fixed now.

Do I need to update the firmware again, haven't used DFU yet. Nice to have.

[Dron_Gus]

I did some more testing today.

I swapped both sensors and still had a problem on the right side of the engine. Checked the spark plugs, and two of the three on the right side were wet. So I swapped the whole right bank of spark plugs. With great succes, both sensors read about the same (within a few tenth of each other)

Engine also runs better and with more power, great succes!

Glad to hear it!

Do I need to update the firmware again, haven't used DFU yet. Nice to have.

DFU is bootloader mode that you been using to flash device over uart. I just fixed a TunerStudio button "Reboot to DFU". So you no longet need to press BOOT0 and RESET buttons in magic order.
OpenBLT bootloader is also included in previously shared binaries. So you can also update over CAN or UART (using OpenBLT specific protocol).
I'm planing to do first official FW release after some testing of current master branch.
But if you would like to be a beta-tester - I can share binaries now.

[AndreyB]

Please update device with attached FW.

This is now https://github.com/rusefi/wideband/releases/tag/20230922

[Dron_Gus]

OpenBLT is supported over BT for f1_dual_rev1 now.

[Dron_Gus]

Now WBOx2 supports AemNet protocol for AFR and EGT. So controller can be connected to MS2/MS3 over CAN interface.
https://github.com/dron0gus/wideband/blob/master/firmware/can_aemnet.cpp

Screenshot from 2024-02-10 13-59-11.png (160.31 KiB) Viewed 22028 times

Working on IO-Box protocol support for AUX analog inputs, AUX PWM and analog outputs.

[kimman1]

I'm just finishing lamba x2
The problem is when I connect the lsu 4.2. Just about 30 second after heating, the temperature notification on tunner studio is sensor overheat.
I attached log below. Please help me
Attached the is latest file
Repost in this thread