Page 3 of 8
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Mon Mar 01, 2021 4:45 pm
by Lambo97
That's good news, I purchased the 17323 would be nice to hear about that sensor.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Thu Apr 01, 2021 6:30 pm
by mck1117
AndreyB wrote: ↑Thu Apr 01, 2021 6:30 pm
Does this apply to GD or does this apply to original 48 pad package ask?
probably both, but especially GD since the firmware is less compatible than a different F0
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Mon Apr 05, 2021 3:50 am
by SHOf429
US $13.85 9%OFF | 10pcs/lot STM32F031 STM32F031K6T6 LQFP32
https://a.aliexpress.com/_mMRiurp
Could this be a substitute? Decent availability, at least for the moment. Cheap enough, too.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Mon Apr 05, 2021 4:09 am
by mck1117
SHOf429 wrote: ↑Mon Apr 05, 2021 3:50 am
US $13.85 9%OFF | 10pcs/lot STM32F031 STM32F031K6T6 LQFP32
https://a.aliexpress.com/_mMRiurp
Could this be a substitute? Decent availability, at least for the moment. Cheap enough, too.
IIRC the F031 doesn't have CAN.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Tue Apr 06, 2021 11:10 pm
by Simon@FutureProof
puff wrote: ↑Thu Apr 01, 2021 6:58 pm
But anyway there are no guarrantees that by the time you send your board to fabrication, these parts are still in stock.
Totally this, things seem to be in and out of stock on a daily basis. Predicting what is going to be available is becoming a gamble.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Wed May 05, 2021 9:17 pm
by hugovw1976
mck1117 wrote: ↑Wed Dec 16, 2020 1:48 am
Ok, time for updates! After some basic testing (idling in the driveway) of the first revision of the board, I ordered a batch of a second revision with some modifications to make it a small module instead of a development board:
image.png
left: USB programming adapter center: wideband controller module right: module installed on its carrier board
The carrier board has a 12-pin sealed connector, protected 5v regulator, and CAN interface chip.
Here's a snip of a log from the very first drive:
image.png
Channel "Lambda" (red) is an AEM X-Series inline controller, sniffing the right hand bank of the engine (even number cylinders).
Channel "Lambda 2" (green) is the new custom controller, sniffing the left hand bank of the engine (odd number cylinders).
The engine itself is a GM LS 5.3 liter V8. The Bosch sensors themselves are identical - but I don't expect the two banks of the engine to be perfectly identical. But they are always within a few percent, which is reflected by the log.
It certainly works! The PID loops for the pump control and heater both need some tuning (less bandwidth, mostly), and the output needs some better filtering. but it's a very promising start.
Tune and log are uploaded
here and
here.
How you set Tuner studio for dual wideband sensors, how you set wideband controller ID for dual sensors? Do you change ID direct in firmware before upload to board?
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Thu May 06, 2021 6:52 pm
by mck1117
I need to write that section for the wiki page at
http://rusefi.com/s/wb
The short version is that there are "set ID 0" and "set ID 1" buttons in tunerstudio that will set the ID of all attached rusEFI wideband controllers. Plug in only the first controller, push the ID 0 button. Unplug that, plug in the second controller, push the ID 1 button.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Fri May 07, 2021 3:54 am
by hugovw1976
mck1117 wrote: ↑Thu May 06, 2021 6:52 pm
I need to write that section for the wiki page at
http://rusefi.com/s/wb
The short version is that there are "set ID 0" and "set ID 1" buttons in tunerstudio that will set the ID of all attached rusEFI wideband controllers. Plug in only the first controller, push the ID 0 button. Unplug that, plug in the second controller, push the ID 1 button.
Thanks. One last question, wideband controllers connect to can1 or can2 on proteus.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Fri May 07, 2021 3:54 am
by mck1117
right now can2 is not fully implemented, so everything goes through CAN1
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Wed May 12, 2021 4:37 am
by AndreyB
I am using eBay v2 ST-Link device.
I am able to flash Hellen WBO module using stm32cube but not using ST-Link utility 4.5
I've tried playing with different connect modes in ST-Link utility but no luck. Is that any sort of 042 magic or else?
- image.png (127.44 KiB) Viewed 76861 times
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Wed May 12, 2021 4:38 am
by mck1117
Was the board powered via USB or no?
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Wed May 12, 2021 4:39 am
by AndreyB
mck1117 wrote: ↑Wed May 12, 2021 4:38 am
Was the board powered via USB or no?
Tried both ways, without USB power stm32cube does not connect always, with USB power stm32cube seems to connect more often.
Still did not work with USB power via ST-Link utility.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Fri May 14, 2021 3:09 am
by vvk74
Does the sensor matter for which fuel? i want to apply bosch 0 281 004 133 (lsu4.9) from a diesel mercedes
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Fri May 14, 2021 3:48 am
by mck1117
vvk74 wrote: ↑Fri May 14, 2021 3:09 am
Does the sensor matter for which fuel? i want to apply bosch 0 281 004 133 (lsu4.9) from a diesel mercedes
Shouldn't matter, no. AFAIK the diesel LSU4.9 has the same sensing element, but a different metal shroud to prevent soot buildup.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Fri May 14, 2021 4:00 am
by AndreyB
Poke poke hellen121vag is asking for 4.2 support
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Wed May 19, 2021 6:21 am
by AndreyB
How does one assign which module is which on a Hellen board for instance when both are constantly powered on? Would we have to go by detecting disconnected sensor?
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Wed May 19, 2021 6:24 am
by mck1117
AndreyB wrote: ↑Wed May 19, 2021 6:21 am
How does one assign which module is which on a Hellen board for instance when both are constantly powered on? Would we have to go by detecting disconnected sensor?
https://rusefi.com/forum/viewtopic.php?p=40940#p40940
do we call this a "distributed double post"? are we cloud engineers now?
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Wed May 19, 2021 6:26 am
by AndreyB
Hehe
- image.png (640.66 KiB) Viewed 76648 times
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Wed May 19, 2021 7:22 am
by AndreyB
- image.png (149.96 KiB) Viewed 76646 times
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Wed May 19, 2021 7:22 am
by mck1117
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Fri May 28, 2021 4:11 pm
by AndreyB
I believe some real vehicle validation is happening? Looking forward some sweet log files
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Sat Jun 26, 2021 12:05 am
by Lambo97
Wiring up the O2's to the Proteus. We touched on this earlier in the thread just want to confirm if my BOSCH 17323 doesn't have the Cal-resistor pin on the plug, just ignore and everything will be rosy?
- image.png (161.48 KiB) Viewed 74999 times
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Sat Jun 26, 2021 12:36 am
by mck1117
It has the resistor. Look in the plug, there are 6 pins. That image is super confusing: it shows the _wires_ that leave the plug and go to the sensor (but who cares - they're between the sensor and captive plug!). The plug should still have 6 pins.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Fri Jul 09, 2021 6:44 pm
by mck1117
This controller has now been embedded in to a handful of Hellen boards, which has uncovered some problems with the heater control. Apparently the wiring in the car I used to tune the heater PID is pretty junk - so the heater warmup is sort of lazy, which means for other people, it heats very quickly and overshoots the target temperature.
Here are two example of "good" warmup, one from a car (bold line) and one from a bench test running at 12 volts (lighter lines):
- image.png (14.27 KiB) Viewed 74077 times
And here's the same good in-car trace compared to a "bad" in-car trace:
- image.png (14.91 KiB) Viewed 74077 times
In the bad trace, the engine was started, so not only was the supply voltage ~2-3 volts higher (alternator charging!), there was some extra heat from the exhaust helping the sensor warm up.
Will update with a comparison once it's fixed and warming up nicely!
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Sun Aug 01, 2021 5:48 pm
by hugovw1976
Hi, I try to compile firmware and have a error, do you know what I'm doing wrong?
- wideband_compile_error.png (41.02 KiB) Viewed 73993 times
I download the github files, and the kicad file don't mach with BOM and Zip Gerber files, the Kicad files have U3 REF3033 and BOM and Zip Gerber files don't. wich one is the latest?
Thanks.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Mon Aug 02, 2021 10:05 pm
by mck1117
hugovw1976 wrote: ↑Sun Aug 01, 2021 5:48 pm
Hi, I try to compile firmware and have a error, do you know what I'm doing wrong?
It's looking for the bootloader to be built first, as the bootloader gets cooked in to the main firmware image. The most "bulletproof" way to build it is to use the build_wideband.sh script, found in the f0_module folder.
hugovw1976 wrote: ↑Sun Aug 01, 2021 5:48 pm
I download the github files, and the kicad file don't mach with BOM and Zip Gerber files, the Kicad files have U3 REF3033 and BOM and Zip Gerber files don't. wich one is the latest?
The gerber files/zip represent the last time the board was fabricated. The kicad in master has some edits that have never been fabricated - but have been validated as the module embedded in to Hellen boards. Provided that it passes ERC/DRC (which I think it should), the kicad in master should be fine.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Tue Aug 03, 2021 12:06 am
by hugovw1976
mck1117 wrote: ↑Mon Aug 02, 2021 10:05 pm
It's looking for the bootloader to be built first, as the bootloader gets cooked in to the main firmware image. The most "bulletproof" way to build it is to use the build_wideband.sh script, found in the f0_module folder.
Thanks for answer, I'm very new on STM32, how I build bootloader?
mck1117 wrote: ↑Mon Aug 02, 2021 10:05 pm
The gerber files/zip represent the last time the board was fabricated. The kicad in master has some edits that have never been fabricated - but have been validated as the module embedded in to Hellen boards. Provided that it passes ERC/DRC (which I think it should), the kicad in master should be fine.
Ok but the board of the zip file works find? or it's better the kicad board, I ask because I all ready order the board of the zip file, but still don't solder any part and I wonder if it's best the new board.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Tue Aug 03, 2021 1:19 am
by mck1117
hugovw1976 wrote: ↑Tue Aug 03, 2021 12:06 am
Thanks for answer, I'm very new on STM32, how I build bootloader?
have a look in that script - it'll build the bootloader before building firmware, then combine them in to a single bin image
hugovw1976 wrote: ↑Tue Aug 03, 2021 12:06 am
Ok but the board of the zip file works find? or it's better the kicad board, I ask because I all ready order the board of the zip file, but still don't solder any part and I wonder if it's best the new board.
The files in the export zip should be fine, yes. They aren't perfect but they work totally fine.
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Thu Aug 12, 2021 3:36 pm
by AndreyB
Re: Wideband Controller: Fully Custom, no ASIC!
Posted: Sat Aug 14, 2021 1:57 am
by hugovw1976
mck1117 wrote: ↑Tue Aug 03, 2021 1:19 am
have a look in that script - it'll build the bootloader before building firmware, then combine them in to a single bin image
Thanks, finally I compile firmware (still need to upload to a board and test)
mck1117 wrote: ↑Tue Aug 03, 2021 1:19 am
The files in the export zip should be fine, yes. They aren't perfect but they work totally fine.
OK, but have some question, the board on export zip just have one jumper pad in back connected to pin 10 on stm32 (wath is the purpose of this) and the board on kicad files have two jumper pads sel1 (pin 15) and sel2 (pin18) to select the can id. The firmware works fine in both boards?