Ecu reset procedure.....

[britlude]

With my gearbox woes and whatnot, while I had the firewall covers off, and prompted by another thread, I took my ECU out and resoldered/reflowed the solder joints between the PCB and the connector terminals.

Now I've warmed the car up and am trying the 'restart the the car and not touching anything and let ECU self learn it's basic settings'..... For quite probably the first time the car is actually changing revs and seems to be doing something....

My question is how long is this supposed to take before I decide it's actually an unstable idle bouncing between 900 and 1500 rpm randomly? Does it give you a clue as to when it might be mostly happy??


Ooo idle getting down to 800ish on occasion now.. then back to 1200.... And down, and up......

[Kaz-kzukNA1]

Presumed that your base idle air control is within the targeted spec (must be adjusted with fully warmed up engine but load condition depends on the non-DBW/DBW).

Presumed no CEL, misfire, seized a/c compressor, etc.


During the re-lean process, the only time the rpm stays at higher region is the first 5sec-ish immediately after starting the engine.
Thereafter, rpm would be forced to stay at the specified idle rpm controlled by ECU through EACV (non-DBW) or stepping motor at the TH body (DBW).

If rpm goes up and down during the process, you have a certain issue and the ECU won't re-learn properly.

ECU reset must be done with fully warmed up engine under no load condition.


From what you wrote, you already knew the followings but in case someone searches;

1. Go out for short driving session to fully warm up the engine or just start the engine and wait until the rad fan kicks in.

2. Once fully warmed up, make sure no electronics devices such as A/C CCU, power window, head lights, audio head unit, etc are in use and then stop the engine.
Depending on the spec of aftermarket parts or modifications on your NSX such as DRL, ICE, etc, you may need disabling them.

3. Pull out 7.5A CLOCK fuse and wait for 10sec.
Put the fuse back in place.

4. Start the engine immediately without touching the TH pedal or any of the electronics devices such as the ones mentioned above.

5. The rpm would hung at around 1,200 - 1,500rpm for about 3 - 5sec immediately after starting the engine but would drop to about 1,000rpm and then down to the specified targeted idle rpm (MT/AT dependant) within the next few seconds.

6. Keep running the engine at idle rpm without touching the TH pedal or any electronics devices for the next 10 - 15min.
The rpm will change slightly when rad fan kicks in but that's normal.

7. Stop the engine and re-start.
This time, the rpm should drop to targeted idle rpm within a few seconds.


For the clarification, the ECU can re-learn the long term co-efficient factors without going through the above procedure through multiple driving cycles.
It's just quicker by carrying out the above process.

As a side note, above process is carried out with fully warmed up engine that your CAT, O2 feedback, secondary air, etc flags (depending on the spec of your NSX) won't be set until you re-start the engine after cooled down and go through the driving cycle(s).
Depending on the country spec, you will need multiple drive cycles to set all flags but for JDM and European spec, normally one drive cycle after cooling down the engine will set all of the flags on our NSX.


Kaz

[britlude]

Thanks for that info Kaz, not sure I've seen the complete procedure written in one place before!!

I have no CEL codes, everything as it should be, just the idle high/unstable and the intermittent gearbox losing its TPS signal (code 3) error. Is the sensor 5v reference generated in the ECU? I've seen it on a diagram as ''control unit' but it seemed vague! I have new air control valves to go on when I get a chance to fit them!

[britlude]

https://youtu.be/B12fxIGOBeM

Just when you think it's calmed down it wanders off again....

[goldnsx]

...and the intermittent gearbox losing its TPS signal (code 3) error.
Sounds fishy. I'm not aware of the full story but is the 12V battery good and all ground cables perfect? I've once had a car (of my friend) who lost connection to ... due to a battery being too small for the car. The voltage dropped and the connection between electronic modules went south and it's thrown a code 'lost connection'. Just a basic input.

[britlude]

battery is fine, doesn't seem to be any power issue, but i shall check earths...

... another thing, it does seem to be running rich too.... :(

[Pride]

Having watched your erratic idle video a couple of times is it possible you’ve missed a slight air leak somewhere, I’m sure if anyone on this forum would diligently check all possibilities then it would be you but from what I’ve seen and read there are many places the air leak could come from, O rings, gaskets, joints or even slight hair line fractures.

What a nightmare symptom you have. 😰

[goldnsx]

If you think it's an air leak somewhere you could use a fog/smoke engine in the intake and pull and twist on all hoses. What about the injectors and injector seals?

[Kaz-kzukNA1]

Watched your video. That's not even staying within the idle control region, lol.
ECU will re-learn something but don't know what will happen within the idle control window.

Similar to another factor causing EACV to hit its max/min limit and ECU just compensating it back and forth or you have other issues.

The TH butterfly should be at constant closed position for non-DBW spec.

AT selector signal sent to the ECU has to be constant during the process.

Fuel pressure should be fairly steady even with your rpm behaviour.

You could disconnect both F & R O2 sensor connectors to force the ECU into open loop.
This may or may not prove anything if the engine getting extra air intermittently such as cracked vacuum tube changing the amount of external air with engine vibration.

Just a matter of interest, where about does the base idle air adjust screw at the TH body sit from fully closed position?
I won't touch it as you may have other issues and adjusting it without fixing other issues first will introduce new issue/factor later.
But, with thoroughly cleaned TH body, for both the non-DBW and DBW model, somewhere around 3/4 opened from fully closed position is a good example.
That's how the TH body left the factory.

Probably you have already seen other examples but this is what you would see during the reset sequence on non-DBW model.
DBW one would behave slightly differently at the beginning but as mentioned below, once it reaches the target idle rpm, it will be the same.
Please ignore the temperature gauge as my sender unit has problem (incorrect resistance and slow response) for ages.
Just waiting for the next coolant service.

It's boring once it reaches the targeted idle rpm (depends on the MT or AT) so I just shortened the video as the rpm just sits at the target for the rest of the process.
Couldn't position my gymbal through the steering wheel so apology for the shaky video.


https://youtu.be/PgO0SfcmSF4

Kaz

[goldnsx]

OT and just curious: what is the red symbol in the lower center of the instrument panel. I haven't seen this before.

[britlude]

I shall be looking at it this week when I get a chance, but I can completely close the idle screw and the car will still run. I presume u should be able to stop it as the engine 'shouldn't' be getting any air.....

To that end, I have a new fast idle valve and EACV ready to go on along with pcv valve and AIT sensor... I'll try the easy bits first.....

[Kaz-kzukNA1]

OT and just curious: what is the red symbol in the lower center of the instrument panel. I haven't seen this before.

Mine is JDM so different from European spec.
The beep and that indicator immediately after IGSW P2 ON are the seat belt reminder.
Introduced from Nov/92 model in Japan.


Kaz

[britlude]

right.. latest investigation results....... when i got home from work the idle was about 1000, with the occasional rise and drop, not as bad as the previous vid... with it running i did douse anything that could have a vac leak with carb cleaner hoping to get a reaction, alas no, so disconnected the ECAV to see what happens.... idle nice and stable, adjustable with the idle screw....


https://www.youtube.com/watch?v=9fcEW8JDH1g

if you multiply the Hz on the multimeter by 20 you get the rpm.... so i can wind it down to 37hz or 740rpm and no lower. i would expect the engine to be able to stall on the screw.... i did block the pcv valve, and it dropped a little to 660rpm, ..... but rose with the pcv valve released again,so some air getting in there as expected...

so no obvious leaks, so it could be the fast idle valve not closing completely, which is a possibility, or the ECAV leaking a little air past...

reconnecting the ECAV valve, the ecu controls the engine speed, the idle lifting again to about 1000rpm, so that is doing something. if i alter the idle air screw (so many idle air circuits!!) the ECAV tweaks to keep the revs about the same regardless of whether the screw is closed or not, but with the occasional rise and fall....

a little percussive maintenance on the ECAV (ie. tap it with a metal stick!) and it seems to be a little happier, so suspect that might be being a little sticky too, given the soot that was on the idle screw when it came out, i wouldn't be surprised!

so, to that end, i have a new fast idle valve, new ECAV, AIT sensor and pcv valve from Amayama, along with a full intake gasket set, so it'll get removed, cleaned, internal butterfly screws checked and all put back together!



as it stands at the moment, restarts like this..... with gear selection too....


https://www.youtube.com/watch?v=atJKTuZq1YI

so, better, we'll see how it behaves for the rest of the week!

i shall put a new idle screw in too, just to complete the intake set! just got to find one that's not on a silly lead time! (16016-PM6-005... same as a del sol!!!)

[britlude]

Idle screw should be with me Saturday!!!

[sorepaws]

Thats the problem with idle screws, they are never in a hurry !

[britlude]

Ba-dum tish!

[britlude]

i might have a breakthrough with the idle issue.....

i recorded this today, the TPS signal at the transmission computer....


https://www.youtube.com/watch?v=0PXnAfhjaNg

now, i'm not doing anything, but the signal is unstable... and the engine revs are following the signal, rev rising when the signal drops low...at least i know whats triggering the rev changes.....

the TPS signal only goes to 3 places- the engine ECU, the gearbox ECU and the traction control computer.....


so disconnect the gearbox and traction ecus.... start up and see what happens?...... (you don't need to watch all of it....!)


https://www.youtube.com/watch?v=y-oYp8uB9f4

somewhat more stable, just a bit of noise, but nothing unexpected, cars are electrically noisy places......

[britlude]

plug the gearbox ecu back in.....


https://www.youtube.com/watch?v=9tBZWBUQNVw

well that's a whole lot better, doing just what it's supposed to......


fortunately i have a spares, so next step is to add in my spare traction control ecu....... (and find out that AJ and Kiki are safe, phew)


https://www.youtube.com/watch?v=9stPbAOBfXw



so i might be a whole lot closer than i was... i suspect i was getting false leads as anything i did to the gearbox ecu involved unplugging the traction control computer to get to the gearbox connectors.... not counting my chickens, yet, i've been here before, but does look promising

seems the traction control computer may have been corrupting the TPS signal... i shall swap in the tps computer proper at the weekend once it's proved itself and investigate the 'problem' one!

[britlude]

alas, just to be a pain, the idle is not so bad, but the gearbox computer dropped under its 'the TPS is ok and i can work with it' threshold.... the signal being erratic once more, with the traction control computer connected or not.....

so i connected a wire to the 5v reference supply that feeds all the sensors...... and hooked that up to the oscilloscope too!

with gearbox and TCS computers disconnected.... i switched on, the yellow line being the 5v ref supply, the blue being the tps output....

https://www.youtube.com/watch?v=mPOoHPXW0yI

5v supply solid as a rock, but safe to say there's something up there with the blue line then! even before it's started, and not touching the throttle, it's all over the place.....

this was the tps i swapped on to the car when the gearbox computer was having it's tps issues..... so swap it back then.... ho hum

on the bench i tested it, it's a basic potentiometer 0v, 5v, and the voltage output depending on where the wiper is, a simple resistor arrangement....

so on the bench, end to end resistance 4.8k ohms, ok acceptable, one side to wiper 1.6k, the other side to wiper 6.8k... hmmm something definitely not right there, it means the first reading should have been more like 8.5k, AND the value is changing with the heat of my hand holding it! no wonder that was causing issues!

i'm guessing when i found the signal stabilized after disconnecting the TCS computer, it dropped the load on the tps so it was happier, but them decided it wasn't... or it was just coincidence?? i hate intermittent faults! grrr

https://live.staticflickr.com/65535/33836938228_204f2a5c25_c.jpg (https://flic.kr/p/Ty49SG)IMG_20190427_142548 (https://flic.kr/p/Ty49SG) by jon sutherland (https://www.flickr.com/photos/152809076@N06/), on Flickr

new idle screw, pcv valve in and i still couldn't get the idle down to spec , so new fast idle valve and ECAV fitted....

https://live.staticflickr.com/65535/33836938258_74ffd8c1c1_c.jpg (https://flic.kr/p/Ty49Td)IMG_20190427_145728 (https://flic.kr/p/Ty49Td) by jon sutherland (https://www.flickr.com/photos/152809076@N06/), on Flickr

https://live.staticflickr.com/65535/40747658773_d009cfe866_c.jpg (https://flic.kr/p/255JpjV)IMG_20190427_145703 (https://flic.kr/p/255JpjV) by jon sutherland (https://www.flickr.com/photos/152809076@N06/), on Flickr

now i can stall the car on the idle screw! some progress at least!

with the original TPS in place and it seems happy, output levels where they should be, and the idle is at the correct levels too.... now it's a waiting game to see if the gearbox computer finds it's TPS levels too low..... again.......

[britlude]

of course, why would honda dream of using a conventional style variable resistor, when they can make it overly complicated....

https://live.staticflickr.com/65535/46800869935_f5e4a9d201_c.jpg (https://flic.kr/p/2eiCG7B)2019-04-28_12-12-05 (https://flic.kr/p/2eiCG7B) by jon sutherland (https://www.flickr.com/photos/152809076@N06/), on Flickr

normally the centre connection is a metal track, and a wiper on the carbon resistor element, giving a simple resistance value (a percentage of the overall resistance)

a toiny pic.....

13690

the honda style has 2 carbon tracks, just to be awkward...

[Heineken]

Would say that's not an unusual design, given that the inner track has zero resistance: https://cdn.instructables.com/F2R/0K9S/IFVC97Q2/F2R0K9SIFVC97Q2.LARGE.jpg :-)

[britlude]

It has a resistance of approx 1.5k ohms... So quite unusual!!!! That linked one has a metal centre ring, the TPS one is carbon.....

[Heineken]

Ooops, that's indeed unusual ;)
May I ask how the resulting resistance curve looks like?

[britlude]

Like a 1k5 resistor on the centre wiper, min 1k5, max 6k8 or there abouts... But the test is all voltage based, 0.5v to 4.5v-ish, but the TPS has a mechanical offset when it's mounted so that would account for the required bottom end offset anyway!

On a more positive note the car is back to pretty much how it was originally, but with a different TCS ECU..... And everything is behaving .... At the moment....

[britlude]

Update.... Just to be awkward the issue wasn't resolved as I hoped it would be, everything was pretty much returned to how it was with the exception of the TCS computer and the additional wiring I had added to bypass the TPS sensor loom......

Then though a series of unfortunate events and a technical cockup on my part, I managed to kill my engine ECU.... Spiked 12v into the 5v rail, which tends to upset the chips in the computers..... subsequently having no fuel pump control I had to source another..... after purchasing a replacement and crossing my fingers, the engine is running again......

The oscilloscope was still hooked up and the TPS signal was a lot cleaner, far less noise spikes....


https://youtu.be/LIo_GdCY69g

As a precaution I also swapped in spare AT controller and TCS units, just in case the spiked 5v rail had affected them, and it seems to be running better than ever, and with an idle where it should have been all the time....

So was it the main ECU all along???? Time will tell...

[britlude]

2500 miles later and all is still behaving, so i think we can call that sorted (usual caveats apply!) no gearbox issues repeated, idle is happily in the 750-800 region on the rev counter and the exhaust i running a lot cleaner.... so ecu all along???