944 OBD Project (On-Board Diagnostics)
#406
Hi Charles,
I believe there may be some confusion regarding the Flywheel Sensor "RPM" and how it relates to the Tachometer "RPM". The Flywheel Sensor "RPM" signal only goes to the 8051 microcontroller so it knows the position of the flywheel. The 8051 uses this signal to fire the ignition and injectors at the correct time. A separate signal is generated by the 8051 for the ignition system and tachometer. For the 951, this signal is converted to 0 and battery voltage square wave via a NPN transistor and fed to pin 21. Pin 21 is connected to your tachometer and the KLR input for the ignition signal. The KLR then regenerates this signal and sends it back to the DME on Pin 32 which is connected directly to the ignition driver. The KLR may decide to retard the ignition if it determines the engine is knocking. The tachometer just receives the signal and displays the current RPM. Listening to the audio as you crank the engine, it sounds as if the engine is trying to fire, but it is also possible the engine could be hydro locking a bit. Hydro locking can be very bad and cause engine damage, you may consider disabling the ignition system by pulling the circuit breaker form the SPORT DME then removing all the spark plugs. When cranking the engine everything in the vicinity can be sprayed with fuel, so make sure there are no ignition sources or items that can be damaged by gasoline.
My suspicion at the moment based on the OBD+ Module's tooth count and your O-scope measurement of the RPM signal is the Flywheel Sensor's RPM signal is not correct. There are several direct and indirect paths to diagnose this issue. Because reading the RPM signal directly can be a bit tricky, I recommend an indirect path by isolating the KLR and measuring the 8051 ignition signal indirectly off pin 21. The outcome of this measurement will determine the next step. Note that if your engine is flooded form cranking, it may not fire up until the flooding condition has cleared.
Whenever I do a lot engine cranking while testing with the OBD+ SPORT DME, I always disable the Ignition system (Remove Circuit Breaker), unplug all the injectors and remove the spark plugs. This is for my personal safety and to prevent damage to my engine including excessive wear on the starter motor and battery. If I was diagnosing a no-start condition, during cranking I would expect the FocusOBD software to show a consistent RPM around 200, the flywheel tooth count to be consistently 132 teeth and the Injection Pulse Width to be consistently around 2.5ms. If these conditions where met, only then would I reinstall the plugs, circuit breaker, connect the injectors then try to start the engine.
- Joe
I believe there may be some confusion regarding the Flywheel Sensor "RPM" and how it relates to the Tachometer "RPM". The Flywheel Sensor "RPM" signal only goes to the 8051 microcontroller so it knows the position of the flywheel. The 8051 uses this signal to fire the ignition and injectors at the correct time. A separate signal is generated by the 8051 for the ignition system and tachometer. For the 951, this signal is converted to 0 and battery voltage square wave via a NPN transistor and fed to pin 21. Pin 21 is connected to your tachometer and the KLR input for the ignition signal. The KLR then regenerates this signal and sends it back to the DME on Pin 32 which is connected directly to the ignition driver. The KLR may decide to retard the ignition if it determines the engine is knocking. The tachometer just receives the signal and displays the current RPM. Listening to the audio as you crank the engine, it sounds as if the engine is trying to fire, but it is also possible the engine could be hydro locking a bit. Hydro locking can be very bad and cause engine damage, you may consider disabling the ignition system by pulling the circuit breaker form the SPORT DME then removing all the spark plugs. When cranking the engine everything in the vicinity can be sprayed with fuel, so make sure there are no ignition sources or items that can be damaged by gasoline.
My suspicion at the moment based on the OBD+ Module's tooth count and your O-scope measurement of the RPM signal is the Flywheel Sensor's RPM signal is not correct. There are several direct and indirect paths to diagnose this issue. Because reading the RPM signal directly can be a bit tricky, I recommend an indirect path by isolating the KLR and measuring the 8051 ignition signal indirectly off pin 21. The outcome of this measurement will determine the next step. Note that if your engine is flooded form cranking, it may not fire up until the flooding condition has cleared.
Whenever I do a lot engine cranking while testing with the OBD+ SPORT DME, I always disable the Ignition system (Remove Circuit Breaker), unplug all the injectors and remove the spark plugs. This is for my personal safety and to prevent damage to my engine including excessive wear on the starter motor and battery. If I was diagnosing a no-start condition, during cranking I would expect the FocusOBD software to show a consistent RPM around 200, the flywheel tooth count to be consistently 132 teeth and the Injection Pulse Width to be consistently around 2.5ms. If these conditions where met, only then would I reinstall the plugs, circuit breaker, connect the injectors then try to start the engine.
- Joe
Also make sure your oil is still good. Mine was like water with all the fuel mixed in.
Joe, wondering if you have been tracking any issues with this following what happened to my motor.
Enjoying seeing what progress has been made. Excited to get back into the 944 game some day soon.
#407
Rennlist
Basic Site Sponsor
Basic Site Sponsor
Thread Starter
Update #19 MAF Housing
Lately things have been very busy, but I managed to get some MAF development done. I decided to try additive manufacturing (i.e. 3D printing) and produce my own MAF housings. Originally, I was going to use OTS (Off-The-Shelf) MAF housings and adapters but there were a couple issues. First, the MAF sensor was not perfectly centered in the housing. This is not ideal because the airflow at the center of the housing is the least restricted and only degrades as it approaches the walls. Second, the airflow transition between the square air cleaner box and the round MAF housing was abrupt. This creates air flow disturbance that cause the MAF sensor to read less accurately than it could. There was also an issue where I was not able to source a OTS housing and adapter for the 944 Turbo.
I designed the below MAF housing in Solidworks and printed it using engineering grade ASA filament. This material is commonly used in automotive applications and will easily withstand engine bay temperatures. It bolts into the factory location, accepts the factory gasket and bolts. The brass inserts for the bolts are flanged and designed so they will not pull out or distort the housing if over torqued. The resolution is lower than production units because it takes 2 to 3 days to print and at this point, I just need functional samples.
Here is the MAF installed in my 944 NA, note the correct moisture-sealed connector that properly adapts the factory harness to the MAF sensor.
I designed the below MAF housing in Solidworks and printed it using engineering grade ASA filament. This material is commonly used in automotive applications and will easily withstand engine bay temperatures. It bolts into the factory location, accepts the factory gasket and bolts. The brass inserts for the bolts are flanged and designed so they will not pull out or distort the housing if over torqued. The resolution is lower than production units because it takes 2 to 3 days to print and at this point, I just need functional samples.
Here is the MAF installed in my 944 NA, note the correct moisture-sealed connector that properly adapts the factory harness to the MAF sensor.
__________________
http://www.ftech9.com/
http://www.ftech9.com/
The following 7 users liked this post by Ftech9:
J1NX3D (12-08-2021),
Knucklebuster (05-20-2024),
Millermatic (12-06-2021),
mwc951 (12-07-2021),
riouxc (12-06-2021),
and 2 others liked this post.
#408
Rennlist Member
Be careful with flooding. I had a fuel injector short circuit caused through the DME and caused the injectors to flood the cylinders nearly half full with fuel. This led to a hydrolock and ultimately catastrophic engine damage. The engine never fired and this was purely from starting. The damaged parts include a starter that is cracked and bent connecting rods.
Also make sure your oil is still good. Mine was like water with all the fuel mixed in.
Joe, wondering if you have been tracking any issues with this following what happened to my motor.
Enjoying seeing what progress has been made. Excited to get back into the 944 game some day soon.
Also make sure your oil is still good. Mine was like water with all the fuel mixed in.
Joe, wondering if you have been tracking any issues with this following what happened to my motor.
Enjoying seeing what progress has been made. Excited to get back into the 944 game some day soon.
To bad for that starter... They don't sell enclosure only... You need to buy all, maybe exclude the solenoid, Ask Ian Riley at https://944online.com/index.php/ .
Charles
#409
Rennlist Member
Joe -
If this means that I can do a MAF conversion keeping my original airbox on my turbo... I'll be sending you some more money in the future...
(I know there are restrictions with the stock box... but I just love the way it looks...)
If this means that I can do a MAF conversion keeping my original airbox on my turbo... I'll be sending you some more money in the future...
(I know there are restrictions with the stock box... but I just love the way it looks...)
The following 4 users liked this post by Millermatic:
#411
Thanks for the info. My car is currently in storage in my garage as it's winter over here. Snow and ice this morning ;-) . Will start back working on it around end of March early May.
To bad for that starter... They don't sell enclosure only... You need to buy all, maybe exclude the solenoid, Ask Ian Riley at https://944online.com/index.php/ .
Charles
To bad for that starter... They don't sell enclosure only... You need to buy all, maybe exclude the solenoid, Ask Ian Riley at https://944online.com/index.php/ .
Charles
#413
Rennlist
Basic Site Sponsor
Basic Site Sponsor
Thread Starter
-Joe
The following 2 users liked this post by Ftech9:
Millermatic (12-08-2021),
spencang (12-09-2021)
#414
Rennlist
Basic Site Sponsor
Basic Site Sponsor
Thread Starter
-Joe
The following 5 users liked this post by Ftech9:
jeyjey (12-09-2021),
Millermatic (12-08-2021),
spencang (12-09-2021),
Tiger03447 (12-08-2021),
YODA2009 (11-11-2022)
#415
Rennlist
Basic Site Sponsor
Basic Site Sponsor
Thread Starter
Update #20
I just pushed a new FocusOBD v0.19.219 to the website to fix the following bugs:
a) FQS sensor would always read "0" regardless of switch setting.
b) Graph line color would not retain the last color selection.
If anyone sees a bug please report it to support@ftech9.com, so I can fix it.
-Joe
a) FQS sensor would always read "0" regardless of switch setting.
b) Graph line color would not retain the last color selection.
If anyone sees a bug please report it to support@ftech9.com, so I can fix it.
-Joe
Last edited by Ftech9; 12-08-2021 at 04:51 PM.
#417
Rennlist Member
Re: the MAF, as always the question; is there a roadmap for something similar for the 16V cars?
Maybe I need to go read some more about what you're doing with this; have you posted something on your website about this development? I am very intrigued!
Cheers
Maybe I need to go read some more about what you're doing with this; have you posted something on your website about this development? I am very intrigued!
Cheers
The following users liked this post:
J1NX3D (12-08-2021)
#418
Three Wheelin'
I agree about 16v cars but for another reason. 968's which already have a MAF are getting to that age where MAF senors are starting to cause issues but are NLA from porsche and are unfortunately exotic, only shared with something like a Lamborghini or Ferrari. Ive had recent MAF issues which luckily could be fixed. I feel like it's on borrowed time and would replace it with an aftermarket sensor and housing if I could get one when I needed one.
#419
Addict
Rennlist
Lifetime Member
Rennlist
Lifetime Member
Same here as I'd love a MAF for my S2. However, the S2 uses a different DME compared to the NA and Turbo cars (it has built-in knock sensor), so I think Joe would have to create a new DME for us first, which may not make sense given the limited number of 16V cars produced?
#420
Rennlist
Basic Site Sponsor
Basic Site Sponsor
Thread Starter
A DME for the 16-Vavle 944/968/964 is high on my development list, I believe I can start serous development on it later this year. It will of course inherit my OBD and MAF technology out of the gate. Also, it seems I can easily design and 3D print a MAF housing that will perfectly replace the original 968 MAF, but accept a modern sensor. Reprocessing the MAF signal to emulate the factory MAF is very simple for me at this point.
-Joe
-Joe
The following 3 users liked this post by Ftech9: