Throttle Position Sensor installation issue

[NSaneXr]

Hey Primers, I'm looking for some help or suggestions with this issue I am having. I have some information in the this thread here, but my question is more specific to my problem. For reference the car is a 97 NSX-T, 6spd - it is a drive by wire car.

I installed a new throttle position sensor today and I'm having problems getting into the recommended specs for voltage. For reference - specs call for .45v throttle closed and 4.5v full throttle. My original TPS measured .39v closed and 4.1 open. After installing the new one I can only get one of the numbers in spec. If I adjust the closed voltage to .45v the best I can do on the full throttle voltage is 4.23v. If I adjust the full throttle to 4.5v then the closed voltage is .8-.9v. I have looked at the plate position in either case and it is closed (small gap) and fully open (I can't move the plate to anymore open). I opted to go with putting the closed value in spec. The throttle response feels much better and it appears to be easier to drive in stop and go situations, but since I reset the ECU I think I have to wait until the idle is back to normal before a final judgement (on reset it seems to start at 1100rpm and eventually get down to 900rpm).

I should also mention that I verified that the accelerator pedal sensor at the front of the car was also in the correct position (couldn't turn anymore at WOT).

Should I be worried that the WOT voltage is out of spec? The howto on the site only seems to worry about the closed voltage - in install document that came with the TPS it says to make sure both are in spec.

If I should be worried, any suggestions on how to fix the issue?

Thanks,
Edward

 

[Larry Bastanza]

Although somewhat speculative, I would take it out and see how it feels. I think it will be fine, but cannot 100% say, since I have not experience TPS failure/replacement.

HTH,
LarryB

 

[NSaneXr]

Thank you for the response. I had it out and it feels fine at WOT and idle has returned to normal. It fixed one of my issues (inability to hold 2800rpm) but not the other (abrupt deceleration when getting off the gas, specifically at 3000rpm in 1st and 2nd). I'm working on that issue on the other thread.

Thanks again.

 

[Larry Bastanza]

Are you refering to depressing the clutch and having the RPM's dip to almost 0, and almost stall?

 

[NSaneXr]

No. Doing that everything seems fine.

In either 1st or 2nd if I take it up to 3000 rpm and release the accelerator pedal but keep the car in gear the deceleration is very very rapid for a moment - its like the brakes were applied. It's very jarring and if you press the accelerator at time it will lurch forward. When it gets really bad it feels like the engine or transmission just stops spinning for a second.

The car has just recently undergone a lot of maintenance and none of it help (TB/WP, valve adjustment, oil change, transmission flush, spark plugs, fuel filter, air filter, cam plugs, new TPS, rear engine mount, transmission engine mount, new 02 sensors).

 

[jaffaz32]

Your symptoms sound very similar to how mine started off and progressively got worse. I do wonder if we have the same issue.

When you checked the WOT voltage how did you do this? get a friend to press the throttle firmly or by hand twisting the butterfly? because if that's the case, does the throttle body actually open fully when fully pressed on pedal, by hand are you opening further than the car actually does and why the voltage will not match? Just a thought. I don't know how it works exactly.

Hey Primers, I'm looking for some help or suggestions with this issue I am having. I have some information in the this thread here, but my question is more specific to my problem. For reference the car is a 97 NSX-T, 6spd - it is a drive by wire car.

I installed a new throttle position sensor today and I'm having problems getting into the recommended specs for voltage. For reference - specs call for .45v throttle closed and 4.5v full throttle. My original TPS measured .39v closed and 4.1 open. After installing the new one I can only get one of the numbers in spec. If I adjust the closed voltage to .45v the best I can do on the full throttle voltage is 4.23v. If I adjust the full throttle to 4.5v then the closed voltage is .8-.9v. I have looked at the plate position in either case and it is closed (small gap) and fully open (I can't move the plate to anymore open). I opted to go with putting the closed value in spec. The throttle response feels much better and it appears to be easier to drive in stop and go situations, but since I reset the ECU I think I have to wait until the idle is back to normal before a final judgement (on reset it seems to start at 1100rpm and eventually get down to 900rpm).

I should also mention that I verified that the accelerator pedal sensor at the front of the car was also in the correct position (couldn't turn anymore at WOT).

Should I be worried that the WOT voltage is out of spec? The howto on the site only seems to worry about the closed voltage - in install document that came with the TPS it says to make sure both are in spec.

If I should be worried, any suggestions on how to fix the issue?

Thanks,
Edward

Where did you get the Voltage figures from?

I'm trying to look for that section in the service manual.

I can find the Throttle body and TPS section but not the Voltage part, do these apply for the motor/Sensor up front in the black box?

 

[NSaneXr]

I connected the multimeter to the TPS. One wire (+) goes to the middle wire on the TPS plug and the (-) goes to a ground. The specification were in the install that came with the TPS and also in the thread here on Prime.

Check out this thread: http://www.nsxprime.com/forum/showthread.php/140528-Throttle-Position-Sensor-(TPS)-replacement-howto

The wires were long enough on the multimeter that I could sit in the cabin and push the accelerator to see WOT. However, when I wanted to check the plate position at WOT I used a hockey stick (being Canadian and all) to push the accelerator while I looked at the plate. It did not look like the plate could open any further at WOT. The only way I think I could get both positions into spec would be to somehow adjust the angle of the plate in the closed position (i.e. set WOT to 4.5v let it close to the .8v and then adjust the angle so it was .45v). I'm not sure how to do that and from what I read, it's probably a bad idea.

I didn't take any measurements from the accelerator pedal position sensor in the front of the car. Although I did manually turn it to make sure it wasn't binding. I wouldn't try to manipulate the plate by hand if the car is "on" (with the engine off) as I believe that the electric throttle motor is engaged.

 

[toomanymodz]

From my Turbo Buick experience, the voltage at idle is more critical than the WOT voltage. For our Buicks, the idle volts should be between .39 and .46. As far as WOT voltage is concerned, anything above 4.0v is acceptable. Not sure how picky the NSX is concerning the WOT voltage though.

 

[Old Guy]

If you have access to an OBDII scanner which supports the extended Honda code set, you should be able to read the actual throttle position reading that the ECU is getting from the TPS in the Freeze data. Honda specifies that the throttle position signal that you read should be approx. 10% at the closed position and approx. 90% at WOT. If you get those values, then your problems do not reside with the TPS. Is your problem with abrupt deceleration a recent issue or something that you have experienced for a long time? If it is not a recent change, I offer the observation (based upon my 2000 6 speed) that with a moderately high compression ratio, the NSX has a fair amount of compression braking. Combined with what appears to be a fair amount of lash in the drivetrain on the 6 speed (when I asked the Acura mechanics about the lash, their comment was that they are all like that!), the compression braking makes for a very jerky ride if you lift the throttle completely at lower speeds. Lifting the throttle at higher speeds doesn't result in the jerkiness because increased drag probably smoothens the transition and you are in a higher gear where the braking torque has less effect. If I am in slow speed traffic, I try to avoid lifting the throttle completely or if I do, I feather the clutch to avoid the jerkiness. Just a thought!

 

[NSaneXr]

Old Guy - thank you for the suggestions. I'll have to grab a different scan tool. The one I have says there is no freeze data available.

The jerkiness has been a problem that I have been experiencing for a while. I wasn't confident enough that I would have been able to isolate the issue as the car needed a lot of maintenance when I got it. It wasn't until this summer that I felt I had addressed all the maintenance issues and could move on to this challenge. I replaced the TPS as during my testing I found a spot at about 2800RPMs where the RPMs would drop. The old TPS also gave some odd readings - the voltage would fluctuate when measuring at idle. It does not do this with the new TPS so I feel I at least fixed that.

I did consider engine braking but it seemed strange that it is very noticable at 3000rpm - a little below and a little above that and it really isn't bad. Since everything else seems to be checking out, it would appear that this is normal behaviour.

Thanks all for the help.

 

[jaffaz32]

I have bought a 1995 service manual and it is saying something completely different.

The manual is saying 0.3V fully closed and 4.5V fully open.

This might be why your having problems making the two match.

 

[NSaneXr]

Actually that is even worse since in order to get to 0.3v fully closed I would have to rotate the TPS which would result in the fully open voltage to be more like 4.1v. I'm basing this on the fact that when I aligned the fully open to 4.5v my closed voltage was 0.8v or so. I think I am going to take Old Guys suggestion and see if I can find a scan tool to read what the ECU is seeing and make the adjustments from there. I'll let you know what I find.

 

[jaffaz32]

I've fixed my issue (i think)

Please see my last post in my original thread, hope it helps.

I do have a question. Please explain how you test your TPS. Idiots guide, I spent ages on trying to work this out. I assume you test it with the plug plugged in to the TPS? rather that the actual terminals on a unplugged TPS otherwise there will be no power to the TPS to get any readings. So, ok test the back of the plug while plugged in to the TPS sounds easy! but... you can't get the red prob of the tester past the insulator on the middle wire on the plug. They have like a grommet around each wire on the plug. Then the earth you run to any earth on the car? And the actual tester itself? would this be set to Volts? but which Volts as i have multiple settings. Yes, i don't know how to use a Voltmeter all new to me. It actually came with a small manual and said for a Toyota TPS which was the example given set to something other than Volts! but the manual reads in Volts.

 

[NSaneXr]

PM sent.

 

[Old Guy]

Just a follow up to my comments on the inherent jerkiness of the NSX 6 speed drivetrain. I did a little experiment on the weekend. A fast lift off of the throttle at 3000 rpm in first gear resulted in a sizeable jerk in the drivetrain. Doing the same thing in 2nd gear the jerk was noticeably less and a fast lift off of the throttle at 3000 rpm in 3rd gear (somewhere around 70 km/hr) resulted in no noticeable jerk, just a relatively smooth, albeit quick reduction in speed. If you are still getting a real jerk in the drivetrain when you lift the throttle at 3000 rpm in 3rd gear, then you probably have something more at play than what seems to be the normal NSX compression braking and drivetrain slack.

 

[NSaneXr]

Old Guy, thank you very much for verifying this. I really appreciate you taking the time to do so. Those results are in line with what I am experiencing. I have been slowly restoring this car for the last two years and this is one of the few times that an issue has come back as being normal.

I still have to measure the TPS on the scanner as a last check, but it looks like at least the jerkiness can be crossed off the list.

Thank you again.

 

[NSaneXr]

I finally got around to getting a scan tool and doing some testing. The readings I get are interesting.

With the voltages set as indicated in an above post - 0.45v closed and 4.2v WOT the scan tool showed 7% throttle at closed and 84% throttle at WOT with a cold engine. With a warm engine I get 4.5% throttle at closed and 83% WOT.

Looks like I need to make a few more adjustments.

 

[Kaz-kzukNA1]

Hi, NSaneXr.

Unless you got CEL, ideally, you don’t want to touch the TH angle sensor.
It’s been calibrated on the bench and hence, Honda normally supplies it only as part of the TH body.
However, things can happen resulting in broken connector, etc and you may be forced to replace this sensor.
Again, Honda normally forces you to replace the entire TH body and not just the sensor for the above reason.

When adjusting the sensor position, ideally, you want to setup your own bench system with just the TH body without any other factors influencing the voltage reading as well as the physical position of the butterfly angle but if you are measuring the sensor output voltage on the car, please measure it at the ECU connector and not at the sensor side.

There is almost no current flowing so the voltage drop is at minimum but still, the ADC (Analog to Digital Converter) component is within the ECU and not at the sensor.

DBW model like your 97 NSX doesn’t have dashpot linkage or extra stepping motor linkage for TCS so from mechanical point of view, you are in better condition than the non-DBW one that has all sorts of extra linkages that will change the reading of TH angle sensor depending on many conditions.

Also, did you check the Vref (or Vcc) at the sensor connector?

All of the voltage spec for TH angle sensor is based on the Vref (Vcc) at 5.00V.

Again, the voltage drop is at minimum over that thin wire but still one must measure it before talking about the sensor output voltage.
There is no point in measuring the Vref at the ECU side because ADC has its own Vref inside the ECU.
For this one, you should measure it at the sensor connector.

As the ADC inside the ECU has its own Vref, if the sensor supply voltage is not exactly at 5.00V, you will need to adjust the spec accordingly.
This is one of the reason why the TH body was calibrated on the bench to minimise the outer influence.

As mentioned, using the OBD2 reader is a very good advice for those with US Acura 95+ models.
For other markets, you may not see OBD2 port on your NSX until 2000+ models.
You don’t need expensive scanner.
Smart phone with cheap OBD2 adaptor and application such as ‘Torque’ are really useful tool.
If you don’t mind spending a little bit more, invest in the adaptor for the extra features and faster data rate. OBD2 data stream itself is not fast enough and the data rate is not consistent that it can’t be used for detailed data analysis but good enough for majority of DIY purpose.


For DBW model, the spec of TH angle sensor voltage measured at ECU when the Vref (Vcc) at the sensor was 5.00V is as follows;
Closed: 0.43 – 0.57V, OBD2 reading: about 10%
Opened: 4.14 – 4.82V, OBD2 reading: about 90%


If Vref was not exactly 5.00V, you should convert the closed/opened spec accordingly.

If Vref measured at the sensor connector was 4.80V, then the spec will change as follows;
0.43 x 4.80 / 5.00 = 0.41V, 0.57 x 4.80 / 5.00 = 0.55V so the above spec (0.43 - 0.57V) will become 0.41 – 0.55V for closed position.
Same goes to the opened position and the spec will become 3.97 – 4.63V.


As you measured the TH angle sensor voltage at the sensor side and not at the ECU, it’s hard to tell but since the voltage drop for both the sensor output and Vref across the wire would be minimum, you seemed to be bit too close to the minimum side of the above spec.


Also, whenever you made any changes at the TH body, you must double check the base idle rpm (in open loop).

On DBW model, ECU will force the idle rpm close to 800rpm (for MT) by adjusting the count command to the stepping motor even when you have wrong base idle setting.

After fully warmed up the engine, stop the engine and you need to short the blue 2pin SCS terminal before re-starting the engine in order to force the ECU into base idle rpm mode.
Please follow the manual for the detail and this is only for the DBW model.

For non-DBW model, you will need to disconnect the EACV.

Without adjusting the base idle rpm, it could (but not always, depending on how far you are off) cause hunting idle rpm, sudden drop of rpm when releasing TH pedal, etc.

Kaz

 

[NSaneXr]

Kaz - wow, thank you for detailed reply. Yes, hindsight being 20-20 I probably should have left this one to the experts.

Here is where I am at currently:

Engine not running:
Vref = 5.002v
Closed = 0.495v; 9.4% TH (from OBD)
Open = 4.250v; 84.7% TH (from OBD)

Your comment about the idle is interesting and I was wondering if this is what is giving the odd throttle % reading when the engine is running/driving - the % drops to 4.5% when I'm off the gas and is this possibly causing the car to kick like a mule when I release throttle?

Thank you again for the help.

 

[NSaneXr]

Well, after 2.5 years (that's how long I have owned the car) trying to troubleshoot a horrible throttle response it looks like Kaz solved it. I can't thank you enough for the help Kaz. Actually, thank you to everyone that jumped in to help. I was quite literally days away from getting rid of the car entirely as it was so horrible to drive.

Removing and installing the new TPS was pretty straight forward but as Kaz wrote above you really shouldn't. Good advice. I probably would have gotten away with the replacement as it ended up being within spec but the car still performed really poorly. If you have to swap the TPS out follow the instructions Kaz gave above. I reinstalled the TPS following those instructions just to be sure, but the car still performed poorly.

The missing piece of the puzzle for me was the idle. The idle on my car was always within spec and changing the TPS never changed that, but as Kaz mentions the ECU tries to compensate to get to within spec. I followed the instructions and the base idle was very out of spec. After adjusting that I was very surprised at how much smoother the car was at idle.

The big test was the 3000rpm throttle lift off. Normally doing this caused the car to immediately lurch forward. With all the proper adjustments I would guess that this has decreased by 98%. It is only slightly noticable in 2nd gear and it doesn't seem like there is any in the higher gears. It certainly seems to be more closely aligned with what OldGuy mentioned above. I realize that earlier I said I was in agreement with what his results were but I just didn't know any better. Coasting in gear is now very smooth and there is no rapid deceleration anymore.

Hopefully we will get a few more nice days and I can continue to test and confirm the findings, but so far things look good.

Thanks again for the help and suggestions.

Edward