[Eth207]

Remember I mentioned that Gear Postion looked a bit flakey, DRodman etc mentioned the neutral position switch being a known contributing fault?
Is your gearbox a 6mt?
So in Cruise Log- External power to Both OCV, in this section of the drive where you gently accellerated to 50mph, on and off the throttle whilst cruising, did you touch the clutch or gearshift? View attachment 70471

Gear box is a 5MT, That is hard to say I don't remember exactly but I dont see why I would, wouldnt the RPM drop if I did?


Edit: I'm looking at the raw data, It appears something happened in the log at 479303ms, It looks like a button was accidentally pressed on rom raider or maybe a glitch, but the log appears to have updated mid log (the parameter's reappeared in the data section, and from that moment forward CL/OL fueling parameter was added.)

 

[bigBADbenny]

You can try the free version of mlv for yourself, many of my parameters are scaled in the app, which can hide glitches, but the evidence is apparent.

Datazap is good but there’s more control for me in mlvhd on a laptop

The next step is to isolate the test mode, 12v applied to each OCV, to see if the test mode created the fault.

Please review my comments & questions above as well.
We’re getting way closer to your diagnosis.

 

[bigBADbenny]

Mlvhd is much like Datazap, x & y axes, x being time and y being data.

But using tools in the (paid) app, like xyz scatter graphs, filters and custom parameters, we might see the exact conditions the fault mode occurred.

Just more control and custom to play with, and for me more familiarity & flexibility.

So we can see that cam and crank are good, but many other parameters glitch.

So for now, review my suggestions and questions above, we’re getting way closer to a diagnosis, solution or place to seek expert advice.

I love a difficult online diagnosis, it’s way harder than doing it in person, so thanks for keeping on, it’s always a learning experience for everyone.

Whatever the highest max possible on any parameter, the glitch is hitting that, so it’s likely to be open circuit on one or the other main plugs?

 

[Eth207]

Okay cool that makes sense.
I’ll check those main connections tomorrow as well as the other stuff you mentioned, and I’ll get some new logs

Does the glitch in the spreadsheet that I found line up with the fault you’re referring to? Or is that not what you’re talking about

 

[bigBADbenny]

I’ll take a look at that later, glitches can be eg kicking the tactrix so that the connection on the plug is lost, or the usb mini plug is strained or loose on the tactrix.

20yo cars can get weird faults eg my oe nav & hu would cut out, the glitch caused IAM to drop, fixed with reseating the units rear plugs with some dielectric grease.

I did reach out regarding the potential fault mode, here:

Based on that comment, I’d be leaning towards intermittent ecu power or ground, or a connection issue.

The neutral position switch seems to be glitchy, I’m going through logs looking for a known good though.

Gear position is more flakey than a known good I found.

 

[Eth207]

I’ll take a look at that later, glitches can be eg kicking the tactrix so that the connection on the plug is lost, or the usb mini plug is strained or loose on the tactrix.

I did reach out regarding the potential fault mode, here: View attachment 70474
Based on that comment, I’d be leaning towards intermittent ecu power or ground, or a connection issue.

Very interesting. Does this glitch only happen once? I have a feeling the person holding my laptop may have accidentally selected the CL/OL parameter mid log, (those values only start showing up at that moment). It is a touchscreen laptop and he was scrolling through the list of parameters while I was driving, maybe he accidentally selected one lol. Maybe those max values are just unknown values because they're non-numerical in the spreadsheet... I can go in the spreadsheet and delete the row of interrupting parameters and re uplaod to see if the fault is still there, or maybe just re-log. But it is interesting how not all of the parameters were effected.

 

[bigBADbenny]

I see what you mean indeed!
Two headers were added at the glitches and one added cl/ol and the other added cl fuel target.
So that’s a red herring, blue duck thing fortunately!
Try get another log, with the last set of parameters, and try to find a spot to cruise for a minute or more with just steering and steady throttle input:
We want to see if the NPS is activating by itself.
I’d definitely also log a cold engine ecu reset, idle with accessories off, no haptic or control input, until warm +75C/167F , then do the 20 minute gentle drive log.
The reset will let us see all the AVCS parameters prior to any recalibration attempt.
I’ve added the known good NPS & gear position logs to the Dropbox for comparison.

Todays required reading, NPS AVCS:

Google Search

 

[Eth207]

I see what you mean indeed!
Two headers were added at the glitches and one added cl/ol and the other added cl fuel target.
So that’s a red herring, blue duck thing fortunately!
Try get another log, with the last set of parameters, and try to find a spot to cruise for a minute or more with just steering and steady throttle input:
We want to see if the NPS is activating by itself.
I’d definitely also log a cold engine ecu reset, idle with accessories off, no haptic or control input, until warm +75C/167F , then do the 20 minute gentle drive log.
The reset will let us see all the AVCS parameters prior to any recalibration attempt.
I’ve added the known good NPS & gear position logs to the Dropbox for comparison.

Todays required reading, NPS AVCS:

Google Search

Okay I will do those logs and research about the NPS. On the topic of the NPS, I have tried resetting the ECU, warming the car up, and driving the car to test for AVCS with the NPS both open (unplugged) and closed (connector wires jumped) and at the time there was no change. I also tested the NPS wire voltage to the ECU. FSM says it should have 12v when closed and 0 when open (or the opposite, I forget) and I remember the values were slightly off, like it would go to 11v and 0.4v, not quite 0.

Anyways Ill log it and we'll see

 

[bigBADbenny]

Ok was the Torqued Performance tune done straight after the swap, was Eric aware that it’s on a WRX 5mt?

Reason I ask is NPS sense can be patched according to the sensor/box combo in the tune. If it’s been setup the wrong way around, perhaps that’s the cause of the hard DTC?

On the other hand, we’ve seen that RHS VVT can work, at least till the system gives up due to operation being single sided, so in addition to logging, try the FSM DTC diagnostics on the harness.

Quote:FSM says it should have 12v when closed and 0 when open (or the opposite, I forget) and I remember the values were slightly off, like it would go to 11v and 0.4v, not quite 0.

Surely a cause for further investigation?
Could be your flakey NPS sensor?

 

[Eth207]

Ok was the Torqued Performance tune done straight after the swap, was Eric aware that it’s on a WRX 5mt?

Reason I ask is NPS sense can be patched according to the sensor/box combo in the tune. If it’s been setup the wrong way around, perhaps that’s the cause of the hard DTC?

On the other hand, we’ve seen that RHS VVT can work, at least till the system gives up due to operation being single sided, so in addition to logging, try the FSM DTC diagnostics on the harness.

Yes I was on the same basemap for a while before the issue started. He hasn't asked if it is a 5mt but he says there is nothing on his end that is stopping the function of the AVCS.

Okay Ill do those too

 

[bigBADbenny]

OK so that's something to keep in mind if there's still no resolution. Poking around the FSM:

The diagnostic voltage & resistance ranges: for the NPS DTC can be found in EN(H4DOTC)(diag)-242

Neutral position switch B134 8
Ignition SW ON (Engine OFF) Engine ON (Idling)
ON: 12±0.5 OFF: 0
Switch is ON when gear is in neutral position.

So your sensor is in range voltage wise, potentially.

NPS resistance check: 5MT-49

1) Turn the ignition switch to OFF.
2) Disconnect the connector of neutral position switch.
3) Measure the resistance between neutral position switch terminals.

Turbo model:

Gear shift position:
Neutral position

Terminal: No. 1 and 3

Specified resistance:

Less than 1 Ω
Other positions More than 1 MΩ

So far it all checks out apart from the NPS being active during that particularly cruisy section of the cruise log where you were on off the throttle.

 

[Eth207]

OK so that's something to keep in mind if there's still no resolution. Poking around the FSM:

The diagnostic voltage & resistance ranges: for the NPS DTC can be found in EN(H4DOTC)(diag)-242

Neutral position switch B134 8
Ignition SW ON (Engine OFF) Engine ON (Idling)
ON: 12±0.5 OFF: 0
Switch is ON when gear is in neutral position.

So your sensor is in range voltage wise, potentially.

NPS resistance check: 5MT-49

1) Turn the ignition switch to OFF.
2) Disconnect the connector of neutral position switch.
3) Measure the resistance between neutral position switch terminals.

Turbo model:

Gear shift position:
Neutral position

Terminal: No. 1 and 3

Specified resistance:

Less than 1 Ω
Other positions More than 1 MΩ

So far it all checks out apart from the NPS being active during that particularly cruisy section of the cruise log where you were on off the throttle.

Okay I ran some logs:

Before I get into it, after the first ECU reset, the CEL is now gone. Completely gone. Hasn't come back after 3 logs and about an hour of driving... very weird, however the problem still remains.


-This one is an ECU reset, cold start, warm up, and then cruise. With OCV's connected.
(You'll notice the NPS jumping around while in gear (I saw it in real time as well). It was cool to see the AVCS wake up once warm, you can see once it gets to operating temp, the left one starts jumping from 0 to 1, like it wants to function.

-Next is the exact same process, except I jumped the NPS wires once warm (I didn't want to warm up the car with them jumped, could've effected the recalibration process) Jumped NPS

-This final log is just a redo of the external power to both OCV's since we had that glitch last time.

I haven't looked deep into the data but so far nothing really stands out to me. While I was driving with the OCV's connected, I noticed the Duty cycles would randomly go back to the 9.41% that they always seem to be at for some reason.

I attempted the NPS resistance test but wasn't able to get a good angle with my probes, Ill try again tomorrow, although the jumping test and the fact that one side of the AVCS is functioning makes me feel like its not the issue, but who knows.

 

[bigBADbenny]

Awesome, I plotted a scatter graph of rpm throttle pos and nps, yours does show significantly more nps activity off idle speed than another car’s log… tbc…
Uploaded to Dropbox.

 

[Eth207]

Awesome, I plotted a scatter graph of rpm throttle pos and nps, yours does show significantly more nps activity off idle speed than another car’s log… tbc…
Uploaded to Dropbox.

Okay interesting

Just a random observation, it is weird how the left current appears steady during warm up on the first log and jumps around in the second. There was no difference between the two tests while warming up (I jumped the NPS after getting to operating temp) there is definitely a lot of inconsistency with the current. This could have been due to the headlights being on? (if you remember, I noticed the current fluctuates when power is applied to another circuit)

What exactly is the current parameter reading? is that the amount of current the ECU is sending to the OCV's?

I think the current has given us the biggest leads, ill refresh our memory:
-Both OVC currents use to fly all over the place (I had p1309/p1307, and neither side functioned)
-I added grounds to the engine (Right side current stabilized, p1307 went away, and the right side started to function)
-I then backprobed the left OCV (ground wire) and attached a ground directly to it (The CEL went away temporarily, and the current was more stable but not perfect)

After all, we know the left side ground wire is sending voltage to the OCV....

I might try taking the ground wire out of the OCV connector, and putting a temporary ground from the engine straight into the OCV connector, so the connector will have power and a known good ground,

Maybe the ground wire is just shorting to power in the ECU. So if i give it its own ground, and assuming the other wire is supplying power sufficiently, maybe it will work


Edit: If I added a ground wire from the chassis straight to the ECU (maybe bolt it to the casing) would this ensure the ECU has a good ground? I've been looking in the FSM and I am unsure how it is grounded or how to check it

 

[bigBADbenny]

I’ll get back with some info about OCV current & duty, apparently if there’s an AVCS fault the OCV duty will default to 9.36/9.4%. Will edit for clarity later.

I haven’t gotten to whether OCV duty is ground or positive switched yet.

Quote: After the ECM sends the target duty cycle signal to the OCV, the ECM monitors the OCV current to establish an actual duty cycle. The ECM determines the existence of a malfunction and sets the DTC when the actual duty cycle ratio varies from the target duty cycle ratio.

All the test logs suggest not enough current at the lhs OCV. We’re trying to find out if that’s the harness, a faulty sensor in a related harness or the ecu.

I overlaid two of the recent test logs, intriguingly, the NPS activity during cruise goes away with iirc powered OCV’s?
That’s the cruise log OCV’s connected overlaid with cruise log external power OCV’s in gray…. So external power makes NPS cruise anomaly go away: it’s a clue.

I’ll go look at the harness wiring diagrams & ecu block diagrams to find just where NPS and OCV circuits get cozy, eg shared harnesses & plugs, grounds etc.

That’s just sanity checks, though.
In the meantime I’d recommend eg voltage drop/resistance testing the lhs OCV wires, between ecu & OCV plugs, include press/shake tests if possible?

Kudos to you for the repeatable logs!

 

[Eth207]

I’ll get back with some info about OCV current & duty, apparently if there’s an AVCS fault the OCV duty will default to 9.36/9.4%. Will edit for clarity later.

I haven’t gotten to whether OCV duty is ground or positive switched yet.

Quote: After the ECM sends the target duty cycle signal to the OCV, the ECM monitors the OCV current to establish an actual duty cycle. The ECM determines the existence of a malfunction and sets the DTC when the actual duty cycle ratio varies from the target duty cycle ratio.

All the test logs suggest not enough current at the lhs OCV. We’re trying to find out if that’s the harness, a faulty sensor in a related harness or the ecu.

I overlaid two of the recent test logs, intriguingly, the NPS activity during cruise goes away with iirc powered OCV’s?
That’s the cruise log OCV’s connected overlaid with cruise log external power OCV’s in gray…. So external power makes NPS cruise anomaly go away: it’s a clue.

I’ll go look at the harness wiring diagrams & ecu block diagrams to find just where NPS and OCV circuits get cozy, eg shared harnesses & plugs, grounds etc.

That’s just sanity checks, though.
In the meantime I’d recommend eg voltage drop/resistance testing the lhs OCV wires, between ecu & OCV plugs, include press/shake tests if possible?

Kudos to you for the repeatable logs!

View attachment 70512

Wow good observation. I did notice while driving the car after the tests, the NPS was acting normal so maybe its just an intermittent issue, but we could be onto something with that.

Is there a way to find what the resistance of the ECU pins should be? maybe I can check right vs left pins and see if there's a difference. That would be an easy way to rule out the ECU I think

Ill do those checks tonight

 

[bigBADbenny]

That last question is a bit out of my league, but I reckon you have enough data in the logs to present to an expert, I’ll ask some tuner friends Sick Chips aka Sean Regan and ThrottleHappy aka Matt McLeod, both on FB.

You could ask Eric’s opinion: these tuners have years of experience and have seen some odd stuff for sure.

Eg send a link to your summary post and my/your Dropbox folder that has all the logs in one place.

I’d also like to confirm what’s going on with your NPS, if it’s a contributing factor is there any danger to swapping out the ecu?

 

[bigBADbenny]

Do you have any logs from before your engine swap and or last tune? If so please share a link

 

[Eth207]

Do you have any logs from before your engine swap and or last tune? If so please share a link

Okay so i tried to do some resistance checks. OCV connector to ECU connector- one wire was 0.6ohms and the other wire is 0.7 ohms. Besides that I was not able to get a clear reading on the resistance to ground (FSM procedures) I am not sure if my multimeter is broken or if I just don't know what am doing, it gives me a new reading every time.

Been driving the car, NPS is not acting up anymore, strange... The fact that one side of the AVCS works makes me look past the NPS, I did a test with the NPS unplugged and neither side worked, so I think we can outrule that?

CEL still hasn't come back, but I tried a test where I put a direct ground wire straight to one of the OCV terminals, and then connected the other OCV terminal to one of the wires in the OCV connector (I tried both wires in the connector)... the first orientation had no results, the second one had no results either however it gave me a p1308 (OCV 2 OPEN), I guess that would make sense

Here is an old log, same engine but this was before my rebuild/upgrades, pretty much a stock 207... However It does not have any AVCS parameters, idk how much help that will be

 

[bigBADbenny]

Keep sharing your driving logs, I want to see what’s going on
NPS is something to keep in mind, but not my current focus which is more grounds related.

Interesting about the ECU plug to ground, the results shouldn’t be vague? Is there at least continuity to ground where you need it?

Going through the fsm wiring diagrams, the ecu has more than one ground, chassis and engine…

So I wanted to ask regarding the swap & your harness specifics, to make sense of the lhd/rhd diagrams, what’s what?
Is the engine harness JDM & the rest USDM?
Did AVCS ever work in this car?
Be specific with the mods/works/schema/timeline as it’s probably relevant.

I’ll put the relevant diagrams in the Dropbox folder.

From googling around the issue, apparently any V7/8 era jdm auto or manual wrx ecu with AVCS can be reflashed for your car, potentially a more affordable option than a specifically STi ecu, source: NASIOC - View Single Post - 02-03 wrx ECU AVCS?