Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Your chance to introduce yourself and your vehicle
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by AndreyB »

tmbryhn wrote:
Thu Aug 12, 2021 3:46 pm
As far as I can understand, current config setup should't look for a cam sensor, but regardless of settings, the trigger log clearly shows what looks like a camshaft input (one green trace pulse per 720 degrees). It does also correlate with "Trigger angle advance"; the green pulse appears at the same point relative to the missing tooth as the value defined under trigger settings - eg. if i set "180", the green pulse appears right in the middle of the two missing teeth in the log. 287 deg defined during the screen capture above.
Worth mentioning is that the "cam signal" appears just the same if I'm running the internally generated ECU stimulator "Trigger Simulator" feature.
Could this be a mislabeled "TDC" mark?
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by AndreyB »

Reminder that rusEFI console Engine Sniffer is the most reliable way to look at engine cycle activity.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

AndreyB wrote:
Thu Aug 12, 2021 3:57 pm
tmbryhn wrote:
Thu Aug 12, 2021 3:46 pm
given that ONLY "VR1" has been configured as an input.
I believe there is an inconsistency here. Not saying that this is relevant or else but just stating that VR2 is in an interesting state.

image.png
Completely agree, and I made the same observation during testing:
I temporarily activated "VVT" just to access the "Cam sensor bank 1 intake" so I could set it to NONE before disabling VVT again - No change unfortunately.
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

AndreyB wrote:
Thu Aug 12, 2021 3:59 pm
tmbryhn wrote:
Thu Aug 12, 2021 3:46 pm
As far as I can understand, current config setup should't look for a cam sensor, but regardless of settings, the trigger log clearly shows what looks like a camshaft input (one green trace pulse per 720 degrees). It does also correlate with "Trigger angle advance"; the green pulse appears at the same point relative to the missing tooth as the value defined under trigger settings - eg. if i set "180", the green pulse appears right in the middle of the two missing teeth in the log. 287 deg defined during the screen capture above.
Worth mentioning is that the "cam signal" appears just the same if I'm running the internally generated ECU stimulator "Trigger Simulator" feature.
Could this be a mislabeled "TDC" mark?
It certainly seem so with respect to the phase change between the missing tooth and the green trace pulse as a function of the "Trigger angle advance" variable, but if it were a true TDC label or flag, it should've happened every rotation as opposed to every 720°, replicating the characteristic of the camshaft and thus allowing for direct spark/fully sequential without a true secondary input for phase detection.
Last edited by tmbryhn on Thu Aug 12, 2021 4:10 pm, edited 1 time in total.
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

Just came back in. Gave it a quick attempt with wasted spark and batched 2-wire mode.

The good news is that it cranked for < 1 sec before it fired straight up and ran like a charm. :D
The bad news is that the USB instantaneously lost connection and Windows repeatedly reported "USB device not recognized". :?
Re-established link successfully once engine was turned off.

Have to figure out what's going on with the link. My previous ECU has had zero connectivity issues on the exact same setup and cable, and the same applies to this ECU on my test bench with a similar 3m long USB cable and live coils/injectors. Hopefully I can figure it out.
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by mck1117 »

That green mark is the TDC mark as generated by the ECU - it's not a sensor.
tmbryhn wrote:
Thu Aug 12, 2021 4:07 pm
It certainly seem so with respect to the phase change between the missing tooth and the green trace pulse as a function of the "Trigger angle advance" variable, but if it were a true TDC label or flag, it should've happened every rotation as opposed to every 720°, replicating the characteristic of the camshaft and thus allowing for direct spark/fully sequential without a true secondary input for phase detection.
Even when in wasted spark mode, rusEFI still operates on a 720 degree engine cycle. That mark is not just #1 TDC, but #1 TDC on the compression stroke. Everything you're seeing is expected behavior.
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

mck1117 wrote:
Thu Aug 12, 2021 7:05 pm
That green mark is the TDC mark as generated by the ECU - it's not a sensor.
tmbryhn wrote:
Thu Aug 12, 2021 4:07 pm
It certainly seem so with respect to the phase change between the missing tooth and the green trace pulse as a function of the "Trigger angle advance" variable, but if it were a true TDC label or flag, it should've happened every rotation as opposed to every 720°, replicating the characteristic of the camshaft and thus allowing for direct spark/fully sequential without a true secondary input for phase detection.
Even when in wasted spark mode, rusEFI still operates on a 720 degree engine cycle. That mark is not just #1 TDC, but #1 TDC on the compression stroke. Everything you're seeing is expected behavior.
I don't understand how spark mode has anything to do with this phenomenon - I mean, you can just fine run in wasted mode while still having a cam sensor to provide phase information for the ECU. The test was btw. conducted with ignition and injection disabled.

How can RusEFi create a "TDC @ compression stroke flag" when there is no cam sensor or equivalent present to provide the information while
the only engine speed/position input present *should* have been configured as a crankshaft sensor? Are you saying that RusEFi makes a TDC compression cycle flag per 720° arbitrarily?
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by mck1117 »

tmbryhn wrote:
Thu Aug 12, 2021 7:14 pm
Are you saying that RusEFi makes a TDC compression cycle flag per 720° arbitrarily?
without a cam sensor, yes
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by mck1117 »

To expand: if you have a cam sensor, it will start on wasted spark, then "fix" the engine phase with the extra knowledge from the cam sensor.
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

Ok. Then it all makes sense.

Improvement proposal: Make direct spark and fully sequential options unavailable (or to throw an error in TS) when only a crank based trigger is configured. I can see some potential back-fires from users miss-configuring their setups with only a crank trigger present and RusEFi starts on the wrong phase.

I hooked up another USB cable and got stable link. Started it up, but dies after only a couple of seconds as it throws a TPS error in TS and some of the analog sensors change their values - kinda like if the 3V3 ref. or something went out of wack. Gotta do some more logging.

Is there btw. a way to log more raw data like eg. all the TPS inputs? Debug mode?
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

When I first gave it a go yesterday, I couldn't get the TBs to work.
Calibrated TPS just fine, but wouldn't run the PID-cal routine.
From neutral to closed, they would respons just fine, but as soon as I tried to get them open past neutral point, the TPS gauge would oscillate between real value and zero while flashing the TPS error indicator. I assumed an overly sensitive "fault TB shutdown logic" and replaced them with another pair I had lying around - the latter had Hall type pos. sensors as opposed to the old ones who had the conventional carbon pots inside the gearbox.

The DBW calibration routines for both TPS and PID as well as regular operation has worked flawlessly on the bench, but I don't know how RusEFi handles glitches or sudden mismatches between the TPS sensors eg. because of noise on the analog inputs. Given the frantic TPS Error flashing in TS, I could only assume it's rather sensitive.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by AndreyB »

We are entering trabant territory :( we know that your choice of op-ams matters so now we will be confirming what op-amp you have on the board

You need 5% error to get TPS error, I have it on hellen if I use wrong op-amps
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by mck1117 »

tmbryhn wrote:
Thu Aug 12, 2021 7:56 pm
When I first gave it a go yesterday, I couldn't get the TBs to work.
Calibrated TPS just fine, but wouldn't run the PID-cal routine.
From neutral to closed, they would respons just fine, but as soon as I tried to get them open past neutral point, the TPS gauge would oscillate between real value and zero while flashing the TPS error indicator. I assumed an overly sensitive "fault TB shutdown logic" and replaced them with another pair I had lying around - the latter had Hall type pos. sensors as opposed to the old ones who had the conventional carbon pots inside the gearbox.

The DBW calibration routines for both TPS and PID as well as regular operation has worked flawlessly on the bench, but I don't know how RusEFi handles glitches or sudden mismatches between the TPS sensors eg. because of noise on the analog inputs. Given the frantic TPS Error flashing in TS, I could only assume it's rather sensitive.
There are gauges for raw TPS voltages, yes. I don't think we currently log them. Did you calibrate the TPS sensors with 12v applied to the board? If not, your 5v sensor supply rail was probably low, so the TPS calibration was wrong.
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

mck1117 wrote:
Thu Aug 12, 2021 8:18 pm
tmbryhn wrote:
Thu Aug 12, 2021 7:56 pm
When I first gave it a go yesterday, I couldn't get the TBs to work.
Calibrated TPS just fine, but wouldn't run the PID-cal routine.
From neutral to closed, they would respons just fine, but as soon as I tried to get them open past neutral point, the TPS gauge would oscillate between real value and zero while flashing the TPS error indicator. I assumed an overly sensitive "fault TB shutdown logic" and replaced them with another pair I had lying around - the latter had Hall type pos. sensors as opposed to the old ones who had the conventional carbon pots inside the gearbox.

The DBW calibration routines for both TPS and PID as well as regular operation has worked flawlessly on the bench, but I don't know how RusEFi handles glitches or sudden mismatches between the TPS sensors eg. because of noise on the analog inputs. Given the frantic TPS Error flashing in TS, I could only assume it's rather sensitive.
There are gauges for raw TPS voltages, yes. I don't think we currently log them. Did you calibrate the TPS sensors with 12v applied to the board? If not, your 5v sensor supply rail was probably low, so the TPS calibration was wrong.
I view them during diagnostics in TS along with the raw pedal inputs. Nice to have. We don't log them currently (hereby on the wish list ;) ).

My board has separated 12V feed to the H-bridge drivers, so there wouldn't even been movements without 12V applied. It has all worked flawlessly on the test bench. Old footage from back when i did the first test of rev.a:

They also respond perfectly with no visible TPS error indicator when the engine isn't running, so it seems like a noise related problem - strange, given that it all seem fairly robust on the test bench with the exact same ECU. Some logging of the raw data would probably be helpful.
Also, it's strange how readings like temperature and MAP/Baro also got screwed after a few seconds after startup before the engine stopped.
Seems like analog reading in general has some problems. I'll give it another go later and gather some more logs - maybe I'll be able to see some patterns.
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

AndreyB wrote:
Thu Aug 12, 2021 8:12 pm
We are entering trabant territory :( we know that your choice of op-ams matters so now we will be confirming what op-amp you have on the board

You need 5% error to get TPS error, I have it on hellen if I use wrong op-amps
Interesting. I'm using rail-to-rail LM324 opamps: 511-LMV324IDT
https://no.mouser.com/ProductDetail/STMicroelectronics/LMV324IDT?qs=%2Fha2pyFaduibZAP9%252B5%252BCY2oSbPgWZew9Bgw6Sz79D3I8LujhD5rsqg%3D%3D

What's really strange is the fact that neither of my old TB's would even open during the PID-autotune or under regular movement seemingly because of TPS error (yes, I testet them individually in single mode with the exact same result), meanwhile working perfectly on my old standalone DBW controller for all practical purposes. Has this been a problem on multiple setups?
If this is the case, would there be an idea to reduce the frequency of the fault checking routine in order to let actual working TBs pass the test while still covering for eg. faulty sensors or shorts on signal wires? Just a thought.
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by mck1117 »

tmbryhn wrote:
Thu Aug 12, 2021 11:01 pm
Interesting. I'm using rail-to-rail LM324 opamps: 511-LMV324IDT
https://no.mouser.com/ProductDetail/STMicroelectronics/LMV324IDT?qs=%2Fha2pyFaduibZAP9%252B5%252BCY2oSbPgWZew9Bgw6Sz79D3I8LujhD5rsqg%3D%3D
Those are not as rail-to-rail as you think they are. They can only manage about 50mv to 4.5v output. The MCP6004 can hit within about 10mv of both rails.
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

mck1117 wrote:
Thu Aug 12, 2021 11:04 pm
tmbryhn wrote:
Thu Aug 12, 2021 11:01 pm
Interesting. I'm using rail-to-rail LM324 opamps: 511-LMV324IDT
https://no.mouser.com/ProductDetail/STMicroelectronics/LMV324IDT?qs=%2Fha2pyFaduibZAP9%252B5%252BCY2oSbPgWZew9Bgw6Sz79D3I8LujhD5rsqg%3D%3D
Those are not as rail-to-rail as you think they are. They can only manage about 50mv to 4.5v output. The MCP6004 can hit within about 10mv of both rails.
Haven't you opted for the 324s on the latest Proteus because of parts availability?
The datasheet states a typical 40mV - 4.95V for a 10k load and 80mV - 4.91V for a 1k load.

Anyway, I fail to see how it's relevant given that the sensor min/max is well within that range, and especially given that it works on the bench and with engine off. But I might be missing something crucial?

If we could get raw TPS data logged, it would be helpful trying to debug this one.
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

Problem solved.
I noticed from todays logs that most of my analog sensors were acting up all of a sudden. After a bit of back and forth, I checked the voltages in the power supply. I have a polyfuse between 12V input and the 5V reg. and it was just barely holding on, so after, say 30 sec. of on-time, it would "semi-trip", creating a lot of voltage drop and consequently was feeding the 5V reg. with between 5-6 Volts. Jumped the polyfuse for now and it works fine.

Now that the engine is running, I have to figure out the last mystery:
The WBO2 is reading infinitely lean while the engine is running. The sensor itself works just fine when I remove the harness adapter+RusEFi ECU and plug in the old ECU. The most screwed up part is that the WBO2 sensor works as intended on the bench with one of my old stimulators connected straight into the harness adapter and RusEFi-ECU. :P
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by AndreyB »

tmbryhn wrote:
Fri Aug 13, 2021 5:21 pm
Now that the engine is running
We deserve a video, maybe even tune and logs :)
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

I left for work the day after, so I'm out of luck for further testing and documenting until I come home 😅

I did upload the .MSQ to RusEFi online the other day though.

Next step is to figure out the weird WBO2 controller behaviour before I can start driving/tuning.
The rest seems to be working perfect, knock sensor input included 🤓🤙🏼
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

Came back from work the other day. Still trying to make sense of this weird WBO2 characteristic.

It responds perfectly on the test bench or out of the exhaust with my propane torch while the engine isn't running (overriding the FP output state to get the MCU heater logic going). But the moment the engine comes to life, it stops responding completely.

I had a hunch it could've been the heater PID control strategy that had a slightly wrong setpoint, or too high total resistance in the wiring harness with the addition of the plug-in adapter, thus resulting in the nernst temp being out of tolerance. I checked all my numbers, ran some more spice simulations, and had it all set up on my bench for testing. I simultaneously sacrificed the BARO analog input to instead log the discrete analog signal that represents Nernst cell ESR so I could compare the temp data in the log with what I observed on the bench. I calibrated and used the Fuel Press HIGH variable (would've btw been really nice if we could have raw logging capabilities for all ADC ;) )
WBO2_temp_test.png
WBO2_temp_test.png (140.13 KiB) Viewed 9495 times
The log shows some deviation, but nothing remotely critical that would've rendered the sensor useless - besides, a too low nernst temp would've resulted in a lambda = 1 reading anyway, but I had to get that part ruled out.

Because I'm running out of spare analog inputs, I probably need to hook up the 4 channel USB scope and look at the analog signals that flows through the op amp based WBO2 controller and see where the values get screwed up. I suspect the lock-up happens in the MCU that serves as the core of the controller, but time will tell. Strange given that it's been a truly tried and tested circuit for years, but it might be something regarding this particular design that I haven't taken into account. Shall be interesting...

Once I get this bug sorted out, it'll be time for some cruising to finally get properly familiar with RusEFi :)
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

Still no luck with the WBO2 🙃
But I made a stimulator. Testing 12 cyl direct spark and fully sequential. Nice tool to confirm that all HW is functioning properly 🤓👍🏼
Snapchat-1876719403.mp4
(8.12 MiB) Downloaded 154 times
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by AndreyB »

tmbryhn wrote:
Tue Sep 29, 2020 4:12 pm
AndreyB wrote:
Tue Sep 29, 2020 11:27 am
Is that 2x45 or 90 pin connector? What connector is that - is it OEM for any vehicle?
It's part of the FCI China knock-off lineup. You also get them in 24, 39 and 56 types.
Do you have a datasheet for the 90 pin connector? Did you use KiCAD or something else for your boards - any chance you can publish the footprint files?

Do you know the official part number of the original connector by any chance?
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by AndreyB »

Mouser says it's Delphi PPI0001336 but I am still curious if you have KiCAD or not KiCAD footprint :)

https://github.com/rusefi/rusefi-hardware/issues/27
Attachments
delphi_PPI0001336.pdf
(410.38 KiB) Downloaded 110 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
tmbryhn
Posts: 169
Joined: Wed Feb 12, 2020 2:40 am
Location: Norway

Re: Hello from Norway! VW Beetle w. 3D printed intakes, 2x DBW and staged injection

Post by tmbryhn »

Hey. I use altium for my designs.
Attached is the data sheet that I've been using. I think it's a FCI connector originally, but I haven't been able to find the original part#, nor a CAD footprint or 3D model (yet). Anyway, the attached datasheet should get you going with implementing the model yourself :)
Attachments
Connector - FCI 90-way.pdf
(410.38 KiB) Downloaded 86 times
Post Reply