Another hesitation thread probably TPS related

[Mac Attack]

As I eluded to earlier, I think for this particular sensor, if it goes bad, the ECU simply uses the (virtually identical) backup sensor. That's why there's no hiccups in operation and why the car runs better with it disconnected. It's not getting conflicting resistance readings with the other TPS.

I've invoked the fail-safe mode in the ECU once before. I had a hesitation around 2500 RPM and simply disconnected an O2 sensor. The ECU went into fail-safe and there was no more hesitation. Just appeared to run a bit richer and with reduced engine timing I assume. Replaced a faulty O2 sensor and it was back to normal.

 

[Old Guy]

As I eluded to earlier, I think for this particular sensor, if it goes bad, the ECU simply uses the (virtually identical) backup sensor. That's why there's no hiccups in operation and why the car runs better with it disconnected. It's not getting conflicting resistance readings with the other TPS.
.

That should be moderately easy to check. If the OP measures the wiper voltages on the accelerator position sensor and the throttle position sensor with the throttle closed and open, that would indicate whether there is a conflict between the two sensors which he is resolving by disconnecting one of them. Both wiper voltages are accessible on the ECU plug so should be fairly easy to do.

If both sensor voltages are reading within 1-2 % of one another, I think we are back to something else is causing the problem.

I am not particularly familiar with the pre DBW throttle; but, from the pictures in the service manual isn't the APS directly linked to the gas pedal via the throttle cable and the TPS directly linked to the throttle plate? Under normal conditions the two would follow one another; but, when TCS engages isn't there a mechanism which decouples the throttle plate from the gas pedal which would result in a position difference between the two? That is the part that has me not totally buying into the idea that the APS is a the fail safe for the TPS.

 

[metsuki]

Men!!! you are great

i do not understand everything precisely but both of you had just the same understanding of the thing as i was.
i think my problem as been narrowed to the TPS because when i unplug it the car runs fine. if i had an injector clogged it will be clogged even when i unplug the TPS
if i try to explain my issue more clearly:

- when the TPS is plug in: i can feel like a very very light stop and go thing when i accelerate moderatly (not when i am steady at speed). i am the only one to feel it in the car, if i do not tell it to my wife she do not bother. i fell kind a lack of power but i think it is only due to the intermitent throttle or at least it is what i think. at WOT i have no problem

-when i unplug it i have the feeling the car just runs great, more torque, i cannot tell anything else (but i did not try WOT, i don't know how trigering the CEL can arm the engine in WOT condition)

i only have a digital multimeter and it has a tendency to jump around a litlle bit. i can take a good reading for static points but i don't know if i will be able to see the smooth variation (a needle multimeter would be great for this type of use)
you are right old guy i will check voltage before changing the tps one more time


injector are standard as ECU
injector not cleaned (but i will do it one day)
fuel filter changed (and i already bypass the fuel regulator with no effect at all, but did not go forward checking fuel delivery precisely)
valve clearance done with timing belt 3 years ago can't say if it as been done the right way but i think so (and you are right, the one side TAITEC exhaust with no cats is hell of loud)
i have to check for leaks but i really don't think it is my problem as it was already there before i changed all the exhaust thing (and i changed everything just beacause of my issue...)

i can tell that when i first plug back the new tps i did not adjust it to test. and it trigger the TCS. when i adjusted it correctly the TCS went away. so i think it cannot be reverse polarity.

i will work on the car this week end

sorry i write this quikly, i hope you will understand me

 

[metsuki]

Hi guys
news are not so good actually
today i worked on this and it turned out to be more and more confusing

- test TPS AT THE ECU, got VCC 5V , min 0,53 max 4,37V, very linear all across no peak at all (i have a video to confirm)
- Test second TPS (under the trottle body, for TCS use i think), VCC 5V, min and max ok (more close to the 0,5v 4,80v than theTPS i changed though) and linear too
- if i disconnect TPS: Engine CEL and TCS light up but car runs good just high idle (1200 rpm)
- if i disconnect the second TPS and the top TPS stay connected: engine CEL+ TCS light car get the issue

We can maybe consider that there cannot be conflict between the two as it would have runs good with the second one disconnected.
we can consider that The ENG CEL do not triger a limp mode solving the issue every time. is there multiple limp mode????

- if I disconnect an O2 sensor and both TPS stay connected: engine CEL car get the issue

I tried to see if another type of engine CEL/limp mode will act like the one i get from disconnecting the TPS but No.... big disappointment there

- if everything is connected properly: Very good idle 800 rpm steady and all no cell but same issue as before


so what that TPS limp mode can change to blur my issue like this


I am really lost there

 

[metsuki]

do you think i should find a way to get both TPS more coordinated?
do they read to far from each other?

 

[Old Guy]

To summarize:

- The TPS voltage measured directly at the ECU pin appears to be close to correct. 4.37 volts at full open is a little low; but, I don't think that is causing your current problem. The important part is that the voltage is 0.5 volts with the throttle closed and increases as the throttle opens.
- If you disconnect the TPS, the car drives without the hesitation problem; but, the idle speed is too high (1200 RPM)
- The APS voltages also go in the correct direction (low voltage closed, high voltage full open); but, the 4.8 volt full open is higher than spec.

I return to my original thought that the TPS is not the problem. The high idle speed with the TPS disconnected could be caused by the ECU putting the idle control in to a fail safe mode to make sure that the engine does not stall when being driven. Why the hesitation when driving disappears when the TPS is disconnected is a mystery.

The O2 sensor voltages can be measured at the ECU connector. With narrowband O2 sensors, when the car is warmed up and operating in closed loop the sensor voltage will be around 0.5 volts. It should be continuously moving from slightly below to slightly above 0.5 volts which will make getting a measurement with a digital meter difficult. Measure both O2 sensor voltages at idle with the TPS connected and the TPS disconnected. If the voltage measurements are not around 0.5 volts that tells you that there is a fuel mixture problem. Close to 0 volts means the engine is running lean and close to 1 volt means that it is running rich. If the O2 sensor voltages change when you disconnect the TPS that tells you that the TPS fail safe mode is doing something to the fuel mixture. If the O2 voltages do not change when you disconnect the TPS, then the problem does not appear to be related to fuel mixture.

Have you tried driving the car with the TPS connected and the traction control turned off? I am wondering if the slight voltage discrepancy between the TPS and APS might be causing some traction control operation which appears as hesitation. I have a DBW car so I have no idea what TCS operation feels like on a pre DBW car. I think this is a low possibility because I would expect voltage errors to cause an error code. However, I am now grasping at straws for a solution.

 

[metsuki]

Hi,
It is to late today for me to test the O2 voltage. Moving out the ECU to test there is a little pain in the a..
But i drove the car traction control Turned off TPS pluged in and the issue was still there.
I ll try to check for O2 voltage this week
Thanks again

 

[metsuki]

The O2 sensor voltages can be measured at the ECU connector. With narrowband O2 sensors, when the car is warmed up and operating in closed loop the sensor voltage will be around 0.5 volts. It should be continuously moving from slightly below to slightly above 0.5 volts which will make getting a measurement with a digital meter difficult. Measure both O2 sensor voltages at idle with the TPS connected and the TPS disconnected. If the voltage measurements are not around 0.5 volts that tells you that there is a fuel mixture problem. Close to 0 volts means the engine is running lean and close to 1 volt means that it is running rich. If the O2 sensor voltages change when you disconnect the TPS that tells you that the TPS fail safe mode is doing something to the fuel mixture. If the O2 voltages do not change when you disconnect the TPS, then the problem does not appear to be related to fuel mixture.

I will work again on the car tomorrow.
when i unplug the O2 sensor the car runs as bad as before. we can conclude that unplugging the O2 sensor do not trigger the same fail safe mode as unplugging the TPS. and i think if unpluging the TPS was acting on close or open loop, unplugging O2 sensor would do it too.
i will test the O2 for rich VS lean condition as you told me
i read elsewhere on prime that the car has 5 different map for different issue but really do not know much else.

if you have any other ideas on what i can troubleshoot tell me, i ll have some time for it tomorrow.

thank you sooooo much for helping me on this

 

[Old Guy]

I will work again on the car tomorrow.
when i unplug the O2 sensor the car runs as bad as before. we can conclude that unplugging the O2 sensor do not trigger the same fail safe mode as unplugging the TPS. and i think if unpluging the TPS was acting on close or open loop, unplugging O2 sensor would do it too.
i will test the O2 for rich VS lean condition as you told me
i read elsewhere on prime that the car has 5 different map for different issue but really do not know much else.

if you have any other ideas on what i can troubleshoot tell me, i ll have some time for it tomorrow.

thank you sooooo much for helping me on this

According to the service manual, the fail safe mode is unique to the sensor that failed. When the ECU detects a failure on the sensor, it substitutes a value for the sensor that allows for safe operation (not necessarily good operation). Therefore, it is reasonable that disconnecting the TPS versus disconnecting one or both of the O2 sensors will cause completely different things to happen in the ECU.

Unplugging the O2 sensor might just cause the engine to go into open loop operation. I know that on my other car which has an aftermarket ECU with a wideband O2 sensor, unplugging the O2 sensor just causes the ECU to ignore the O2 signal and default to open loop operation. I really don't know what the Honda ECU does, so I can provide no advice on the usefulness of unplugging the O2 sensors as a test.

 

[metsuki]

i performed some check on th O2 sensors.
found something strange:
first seems to be a problem with the heater from the back O2 sensors; i get 8 ohms between terminal C and D.
service manual tell me i need something between 10 to 20 ohms
i changed those O2 sensor 6 month ago with denso one (not from honda) i do not know if they have the same resistance as the original one?

as for the voltage it is even more weird:
-car is hot and idling
-testing voltage on the ECU
-voltage between A28 and D14 is around 0,1v
-voltage between A28 and D16 is around 1V

i got the same voltage TPS plug and unplug

it seems i have one bank running too rich, and one too lean

guys i am more and more lost with this car, finally i am not even shure i am doing this correctly. taking any voltage on the ECU is very hard, there is literally no room at all and the cables are very tight fitting, making this so tricky

PS: car idling in my parking, it did jump two or three time from 800 to 1200rpm and come back to 800 all alone for no reason (TPS plugged in)

 

[Old Guy]

I am going to guess that you meant A26, not A28 because according to the North American service manual connector A only has 26 pins. A26 is ground and D14 to A26 would give you the front sensor voltage and D16 to A26 would give you the rear sensor voltage. Assuming that the measurements are probably correct the engine is running rich on the back and lean on the front or you have a dead sensor on the front. If you had measured a voltage above 1 volt, I would say that you had picked the wrong pins. The fact that your measurements are 1 volt or less suggests that you connected to the correct pins.

Those O2 sensor voltages are at the outer limits for a narrow band sensor and my opinion is that they should be triggering mixture out of range error codes. Have you tried using the service check connector to see if you have any stored error codes? Does you check engine light work (not burned out or burned out fuse #6 )? Does the check engine light come on during the self-test check when you first turn the ignition key to run?

I don't think the slightly lower resistance on the Denso O2 heaters is a problem. In the worst case, the lower resistance might draw too much current which could blow fuse #5 . But; if fuse #5 failed the engine would not even start so I don't think the heaters are a problem.

Do check for stored error codes because I don't know how you could have those voltages without generating error codes.

 

[metsuki]

Hi,

You are right, it was A26. There is a typo on the service manual it says A28 on the graph and A26 in the writing.

I used the service check connector yesterday but as I had made so much test disconnecting stuff , it triggered many things. today I am going to pull out the clock fuse and check for real codes.
Yesterday I had CEL42 AN CEL7 and TCS37 TCS36 TCS51
Seems my CEL is doing right though
D14 was moving between 0,06v and above
D16 was more steady around 1v. I tested this one first and it came up from 0,7v or so to1v (car had some time to cool done before I take measurements and start the engine I think)
You are right, with such voltages I would think the CEL would come on
Another strange thing is the car passed perfectly emission test last month without cats.
I have only one side exhaust. If there was such mixtures problem I don't think it would pass

 

[goldNSX]

Another strange thing is the car passed perfectly emission test last month without cats.

What country is so generous?

 

[metsuki]

so today i took time too get good read from my O2 sensors.
car ran for nearly 20mn before to be sure it was hot
here are the video from both sensor:

[video]https://photos.app.goo.gl/inQeV5y1fgXFXBsL7[/video]

[video]https://photos.app.goo.gl/oUiZPJkyGZfGwRqD7[/video]

voltage are constantly jumping from 0v to 1v on both sensors (yesterday maybe the car was not that 's why reading was different), i don't know how to interprete that.
i check for exhaust leak and saw one between the back headerand the block and one between the two last parts of the exhaust
I drove the car with both O2 sensors disconnected this time and my hesitations went away .... so not the same result as when I unplug just one O2 sensors.

hesitations are very soft, but i feel it and i hate that

 

[Old Guy]

42 is a rear O2 heater problem. Check the connection to the O2 sensor in the engine compartment to see if there is a problem. If not, then your ECU may not be playing nicely with the Denso O2 sensor and might be triggering an O2 sensor fail safe mode. You may have to try a 'Honda' O2 sensor to see if that makes the codes go away.

7 is a TPS error code which could be high or low voltage (pre OBDII does not separate the two). Did you get this error with the TPS connected? If you did, you may still have some TPS adjustment issues.

I have the 1997+ service manual for DBW cars. You will have to find a 1991+ manual to get the codes for the earlier TCS system. Getting those codes may provide some additional information as to what might be going on with the ECU, or you may just have triggered them when you were working on the APS and TPS. If you are continually triggering the 7 error code on the ECU because of a TPS problem, then you may be triggering similar TPS / APS codes for the TCS.

When you disconnect the battery or remove the clock fuse, you erase all the stored error codes and you erase all the accumulated fuel mixture corrections. You may need to drive around for some time (I have no recommendation for the number of hours or kms ) to re acquire the error codes and re acquire the fuel trims. Repeat the check for error codes after you have done some driving. If you have a less restrictive exhaust system and no catalytic converters, you may require a fair amount of fuel trim to provide the correct fuel mix and it will take some time for the ECU to accumulate that fuel trim value after you have erased it. However, if the ECU is ignoring the rear O2 sensor because of the heater problem, then you may never get the trim re established. So, it would be good to sort out that O2 heater error.

The 1997+ service 'implies' that acceptable voltages for the O2 sensors (as in the sensors are working) at idle are above 0.1 volts and below 0.9 volts. If the voltage stays at 0.1 volts or less you check the fuel pressure first, check the O2 sensor next and then check do an ECU replacement. If the voltage stays at 1.0 volts or higher they don't list fuel pressure as a problem; but, include the O2 sensor and ECU checks. Since you have both 0.1 volts and 1.0 volts at the same time, I'm ???????? Are you absolutely sure that you don't have an air leakage problem on the front O2 sensor which is driving the ECU to go rich and create the 1 volt reading on the back sensor?

My opinion is that any emission test that you can pass with the catalytic converters disconnected may not provide any engine diagnostic value.

In the list of things that you have done to the car, I did not see any injector maintenance or a check on fuel pressure. A clogged injector(s) or a leaking injector could cause your strange O2 voltage readings at idle. Your car is a 1993 and if the injectors have never been checked or cleaned, having them cleaned and flow checked would be a good way to eliminate them as a possible cause. I don't think the fuel pressure is causing your strange O2 readings; but, since its relatively easy to check (if you have access to a fuel pressure gauge) its one of those things I like to check off any time I am dealing with strange fuel mixture issues.

 

[Old Guy]

What country is so generous?

Is the Euro emission standard constant across the whole Euro zone or do they have testing variances for different areas?

 

[goldNSX]

If I understand correctly the exhaust gas leaks are between the engine and the o2-sensors. Get them fixed first.

 

[goldNSX]

Is the Euro emission standard constant across the whole Euro zone or do they have testing variances for different areas?

For new cars it should be identical. But there might be individual exceptions for older cars. In Switzerland (which is not part of the EU for god's sake), you could not pass the emissions test without cats if the car came with them form the factory.

 

[Old Guy]

If I understand correctly the exhaust gas leaks are between the engine and the o2-sensors. Get them fixed first.

Exhaust system leaks 'upstream' of the O2 sensors will definitely cause problems for accurate O2 readings under just about all operating conditions. At idle, an exhaust gas leak that is a short distance 'downstream' from the O2 sensor may allow enough atmospheric air to reach the O2 sensor causing inaccurate O2 readings. A really open exhaust system can also have problems with atmospheric reversion influencing an O2 sensor located close to the end. Since metsuki has a common exhaust exit and one O2 reading 1.0 volts and the other reading 0.1 volts, I don't think atmospheric reversion from the open end of his exhaust is causing this particular mystery.

 

[goldNSX]

Since metsuki has a common exhaust exit and one O2 reading 1.0 volts and the other reading 0.1 volts, I don't think atmospheric reversion from the open end of his exhaust is causing this particular mystery.

I'm with you on this. But my priority would be to fix the leaks first before going any further. Parasitic air is very much undesired by the o2-sensors.

I've aftermarket Denso o2-sensors on my car (Europe 91), not sure which one, but they're working absolutely fine since more than 6 years now. So I think they're a perfect substitute for the very expensive OEM o2-sensors. No need to try with OEM ones.

 

[metsuki]

Can you watch the videos just to give me your opinion on the reading?
Now both bank read the same

I made a real reset ECU today, 5 hours without clock fuse then 12mn idling then driving the car for 15mn
for now no CEL codes are back
TCS codes are still there , I think they do not erase when you pull out the clock fuse

 

[Old Guy]

I had a look at the videos. I can't see the decimal place; but, if the voltages are swinging back and forth from around 0.1 volts to 0.9 volts (sort of looks like that) I would say that your O2 sensors are probably working (when you did that test) and that your fuel mixture control is probably normal. That sort of puts you back at square 1 in terms of trying to figure out what is going on.

It might be helpful to rule out dirty injectors and fuel pressure as a problem because I am running out of possible causes.

 

[Mac Attack]

Hi,

You are right, it was A26. There is a typo on the service manual it says A28 on the graph and A26 in the writing.

I used the service check connector yesterday but as I had made so much test disconnecting stuff , it triggered many things. today I am going to pull out the clock fuse and check for real codes.
Yesterday I had CEL42 AN CEL7 and TCS37 TCS36 TCS51
Seems my CEL is doing right though
D14 was moving between 0,06v and above
D16 was more steady around 1v. I tested this one first and it came up from 0,7v or so to1v (car had some time to cool done before I take measurements and start the engine I think)
You are right, with such voltages I would think the CEL would come on
Another strange thing is the car passed perfectly emission test last month without cats.
I have only one side exhaust. If there was such mixtures problem I don't think it would pass

....
I made a real reset ECU today, 5 hours without clock fuse then 12mn idling then driving the car for 15mn
for now no CEL codes are back
TCS codes are still there , I think they do not erase when you pull out the clock fuse

The TCS codes stay in the TCS memory and do not trigger a light unless a new TCS problem happens. You can erase them by unplugging the TCS computer, or the 20A ALB 23 fuse from the main relay box. The following codes probably occurred when you were unplugging the sensors:

TCS code 36 is abnormal TPS signal.
TCS code 37 is abnormal TCS angle sensor signal (the other TPS)
TCS code 51 is abnormal fail-safe relay, TCS control valve actuator, or TCS control valve assembly.

Did you try to turn off the TCS by disabling it with the button on the dashboard? Maybe go for a drive and test it out with TCS off?

PS: car idling in my parking, it did jump two or three time from 800 to 1200rpm and come back to 800 all alone for no reason (TPS plugged in)

This sounds like the problem - An intermittent electrical issue with a bad connection.

 

[Old Guy]

This sounds like the problem - An intermittent electrical issue with a bad connection.

Are you thinking an intermittent ignition problem rather than fuel problem or still on the fuel delivery side?

 

[Mac Attack]

I know I'm stubborn but I still think it's TPS-related. Either the sensor and wiring resistances, or intermittent sensor/wiring fault.

The 1200 RPM is similar to what someone else noticed here:
http://www.nsxprime.com/forum/showthread.php/144415-Lugging-issues-after-1-yr-storage

Also, not having experienced this issue myself, I can't figure out why one of the most respected engineers/technicians for our NSX infers the TPS replacement is temperamental himself here:
http://www.nsxprime.com/forum/showthread.php/176554-Throttle-Position-Sensor-installation-issue?p=1739342&viewfull=1#post1739342

NSaneXr's follow-up post on resetting the idle afterwards is apparently important too.

Of course, it could still be an exhaust leak, bad injector(s), failing fuel pump, etc.