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

[Dron_Gus]

Rev1 was just released. https://github.com/rusefi/rusefi-hardware/tree/main/lambda-x2/export/rev1
Any feedback is welcomed.
Short changelog is:
* Switch to 64 pin STM32/GD32
* Add 8MHz HSE oscillator (better clock stability for CAN over wide temperature range)
* Use DAC (instead of PWM) outputs for driving Ip
* Move BT to UART3/USART3, keep USART1 for bootloader/debug
* Drive BT's EN signal from MCU
* Fix RC filters on heater outputs for better voltage measurement
* Use separate ADC input for Vbat measurement
* Use ADC channels for AUX outputs monitoring/diagnostic
* Use PWM noise cancelation circuit on AUX analog outputs
* Removed useless/non-functional protection of Vm outputs
* Fix board outline to fit two types of enclosure ("plastic top" and "alloy brick")
* Add testpoints for TC2030 JTAG probe
* Add buttons for Reset and Boot0
* Add Cfg1 and jumpers
* Replace fuse holder with resettable 0.2A 33V fuse
* Two status LEDs for separate status indication for left and right channel
* Silkscreen labels around testpoints
* Silkscreen labels at uart connector on both sides of PCB
* Un_3x_sense formula is now 0.247 + 3.15 * (Un - Vm)
* Un_sense (not gained) is also routed to ADC input
* Protection from possible Ip current when MCU in reset/bootloader
* While "power on" LED replaced with green

[Dron_Gus]

Main connector pinout added to silkscreen.

[AndreyB]

Between https://github.com/mck1117/wideband/pull/157 and https://github.com/mck1117/wideband/issues/215 looks like there is enough logic in the wideband repo to justify some sort of automated testing. https://github.com/mck1117/wideband/issues/218 for further discussion.

[AndreyB]

party time!

image.png (20.08 KiB) Viewed 51563 times

[Timbelom]

controller hard-wired to battery
Stock ECU
few times engine restart
actually looks good

[Timbelom]

everything the same
but with RusEfi ecu

[Steve-F]

Rev1 was just released.

Hey, has the Rev1 PCB been tested yet?
I'm considering ordering 2 assembled from JLC to have a play with in the garage.

REF3033 is OOS everywhere at the moment, what do we think of ZRC330F02TA as an alternative.
JLC have it in stock currently.

Cheers.

[Dron_Gus]

Hey, has the Rev1 PCB been tested yet?
I'm considering ordering 2 assembled from JLC to have a play with in the garage.

REF3033 is OOS everywhere at the moment, what do we think of ZRC330F02TA as an alternative.
JLC have it in stock currently.

Cheers.

Not yet. It is on the (two) ways to me. Hope to get in two-three weeks.
FW needs significant adjustment, so add +1 week for testing.
About REF3033 - you can always populate L1 and save few cents... Accuracy possibly will degradate (never tested). But main LDO (XC6206P332MR) is close to ZRC330F02TA in accuracy.

Screenshot from 2023-03-23 21-21-35.png (13.92 KiB) Viewed 51324 times

[Dron_Gus]

New hw is just arrived.

[puff]

Metal enclosure for a board with bluetooth?

[Dron_Gus]

Metal enclosure for a board with bluetooth?

Optional BT. For BT fans there is plastic case. See first post in this thread.

[AndreyB]

Look dual WBO went to a cold place to do some engineering work!

[Dron_Gus]

Just a funny picture of my test bench.
Too different eras and markets...

[Dron_Gus]

Accuracy is not so bad by the way.

Screenshot from 2023-04-26 23-16-20.png (194.2 KiB) Viewed 50901 times

[Steve-F]

I've seen on GitHub there is the issue if you populate all the sensor ESR resistors.
If I just populate the 4.9 ones, is the Rev1 board useable?

I've a V6 I'm building up over the next few weeks, and I'd love to have a play with per cylinder measurements, this would make that possible without costing a fortune.

[Dron_Gus]

I've seen on GitHub there is the issue if you populate all the sensor ESR resistors.
If I just populate the 4.9 ones, is the Rev1 board useable?

I've a V6 I'm building up over the next few weeks, and I'd love to have a play with per cylinder measurements, this would make that possible without costing a fortune.

Dual Rev1? This issue affects only dual channel boards...
Yes. If you have only 22K populated it is ok.

[Steve-F]

I've seen on GitHub there is the issue if you populate all the sensor ESR resistors.
If I just populate the 4.9 ones, is the Rev1 board useable?

I've a V6 I'm building up over the next few weeks, and I'd love to have a play with per cylinder measurements, this would make that possible without costing a fortune.

Dual Rev1? This issue affects only dual channel boards...
Yes. If you have only 22K populated it is ok.

Yes, dual channel Rev1.
I was planning on using 3, so I can have a sensor per runner.

I'll just populate R24 and R71 with 22K.

Are there any other issues that have cropped up that will result in a Rev2?

Thanks

[Dron_Gus]

Are there any other issues that have cropped up that will result in a Rev2?

None were found... yet.

[Krakert]

Hey I just bought this lovely Wideband controller from the RusEFI shop.

Can you tell me a bit more about setting up the controller, how do I setup the CAN bus and all?

And Andrey told me, that for now, only channel 1 is supported in the firmware, could you tell me more about that?

[AndreyB]

I think it's just this one? maybe both just in case.

image.png (29.07 KiB) Viewed 50015 times

image.png (2.99 KiB) Viewed 50015 times

[Krakert]

Further no changes needed in the `CAN Bus Communication` menu. I assume the firmware is already flashed on the unit?

[AndreyB]

I am experiencing something new - overhear error message. Just swapped the sensor and got same exact message again. Attached are two logs both are same second sensor

[Dron_Gus]

I am experiencing something new - overhear error message. Just swapped the sensor and got same exact message again. Attached are two logs both are same second sensor

Log look weird.
Heater in WarmupRamp state. Duty is 51%. Temperature rises up t o 730 (780 - 50). Heater switches to ClosedLoop (PID regulation), duty is 62%. Temperature rises to 778 (close to target 780). Duty drop to 0% and instantly rises to 100% and stay until heater fails with overheat.
100% at 14V of supply is more than allowed 11V for heater. We have clear check for this.

Code: Select all

            // Limit to 11 volts
            if (heaterVoltage > 11) {
                heaterVoltage = 11;
            }

Have no idea. May be some bug in PWM code or newly introduced Sampler returns incorrect GetInternalBatteryVoltage().

[Krakert]

I'm trying to get the controller to work, but with little luck so far.

When it tries to read the sensor it all goes wrong. I'm not sure what is causing this.
Could some with a bit more knowledge look at the log file?

[AndreyB]

A photo of the wiring would help, a few words describing your setup would work - how do you power your unit?

Have you tried on bench? Do you have a regulated power supply to see current consumption?

[Krakert]

The probes are already mounted in the exhaust of the 911E, and they are just feed by 12V from the battery.
I connect to them via bluetooth, and they seem to warmup.

mck1117 also looked at a previous log, and he suggested that maybe Vm and Ip may be swapped.
"the warmup goes OK but when it tries to use the sensor it goes off in to oblivion'"

[AndreyB]

I guess I am sick or just at the end of my rope with customer support just connected directly to battery? That would be strange. I specifically ask for details but I can only treat your answer as a not-answer.

[Krakert]

image.png (21.41 KiB) Viewed 49389 times

https://rusefi.com/docs/pinouts/lambda-x2/

Could you help out with matching the pins? Between this diagram and the pinout online of the Lambda x2?
I often get confussed with the naming of the pins.

And there is really not much to it right? I bougth the controller so I hope some testing is done right?
6 wires and some power, maybe I just made a simple fault myself in the wiring. How about the firmware? maybe a bug there, I doutb tho

[AndreyB]

https://github.com/rusefi/rusefi-hardware/blob/main/lambda-x2/readme.md now mentions https://github.com/rusefi/rusefi/wiki/WBO#naming-convention

[Krakert]

Just checked the wiring, that is all good, the suggestion of mck1117 was incorrect of that Vm and Ip may be swapped.