Another hesitation thread probably TPS related

[metsuki]

Hello everyone ,

From the day I bought my car I had hesitation on light throttle or even at constant speed (throttle seems to fluctuate a bit even if I stay steady )

in two years i changed to solve this:

6x spark plugs (two times)
6x coil pack
2x O2 sensor
SS exhaust manifold
SS cats delete
SS one side single exaust TAITEC

and i have Still the samedi issue.

I searched for a solution since then

Last day I found a fellow primer with a TPS issue sounding like what I endure for so long.
I disconnected my TPS and beside the CEL light coming on, I can tell that the car seems to ride perfectly. I reconnected it and the issue came back

Can i drive with the TPS disconnected waiting for the part to arrive?

I only found TPS for 95/05 nsx wich I bought anyway. I found divergent informations about it. Some says that it turn the wrong way, other says that it is ok all honda's TPS are the same, other says you can invert the polarity.
Has anyone have any real good informations about this stuff?

I will clean my throttle body and check the idle at the same time

 

[Old Guy]

You can probably drive satisfactorily without the TPS. The TPS does a few things, triggers fuel cut during throttle off / de acceleration, probably tells the ECU to go into open loop at large throttle openings and provides acceleration enrichment. The most engine critical function is forcing the engine into open loop at wide open throttle so don't be doing any vigorous driving without the TPS connected.

The TPS is just a variable potentiometer. All you have to do is make sure that when the throttle is closed the original TPS and the new TPS have the same resistance values between the corresponding pins. Same thing when the throttle is wide open.

 

[NSaneXr]

You may find this thread and in particular this post from Kaz helpful for setting up the TPS after replacement

http://www.nsxprime.com/forum/showt...lation-issue?p=1739342&viewfull=1#post1739342

 

[metsuki]

thank you,

i already saw the KAZ post.

my car is a 93 . will the 95/05 TPS we find everywhere be ok?
i do not find TPS for early models, i think there is not from what i gathered

 

[metsuki]

Hi everyone

I work on this today and I really need your help on the voltage check, please.

For other people with problem finding TPS for non DBW throttle body.
I only found TPS for 95 05 nsx.
Those TPS turn the wrong way (like I saw in some thread) but you can easily make it turn the good way by swapping the preload spring from your original TPS. the voltage goes the right way too

BUT

I changed my TPS today (5,02v VCC, 0,48v idle 4,45v wot), I clean my throttle body and check the idle at the same time (little too low eacv disconnected as the engine won't idle )

And for God sakes the issue is still there

I now that the car runs open loop when the TPS is unplug, so what else can cause my problem

the only thing i cannot understand is why TPS unplug everything Come back to normal

what other sensor can i unplug putting the car in open loop, just to check if my problem is solve by unpluging the TPS or just by running open loop?

Actually i already went far beyond what i thought i could troubleshoot

I am sorry but here I feel upset and hopeless

Please help me

 

[Old Guy]

I now that the car runs open loop when the TPS is unplug, so what else can cause my problem

Which open loop? There are at least two control loops on the NSX ECU, the idle control and the fuel mixture control. Idle control inputs are typically throttle position (TPS) and engine speed and on the NSX there appear to be brake and clutch interlocks. Fuel mixture control typically gets kicked out of closed loop if the engine coolant temperature is not up to normal operating spec and when the throttle approaches wide open. Some ECUs may be set up to switch to open loop fuel control if the engine RPM drops too low or if the MAP value is at the very low end (wide open throttle) or high end. It might also be possible that if the O2 readings go way out of limits the ECU goes into open loop; but, that should trigger a mixture error code.

As a matter of clarification, how do you know that the engine is going into open loop when you disconnect the TPS? Assuming that it is going into open loop fuel control an incorrect TPS reading could do that. The best way to confirm that the engine is or is not in closed loop would be to measure the O2 sensor voltage output. To do this, you would need to make a jumper arrangement that would allow you to measure the O2 sensor voltage while the engine is running. You need the jumper because the sensor needs to be connected to the ECU during the test. If the ECU is running in closed loop, the sensor voltage will typically be around 0.5 volts and will be constantly swinging above and below this value. If the sensor voltage remains permanently below 0.5 volts (reading lean) or above (reading rich) the engine is not in closed loop control or the sensor is dead. Sensors commonly fail to reading lean (0 volts) causing the ECU to add fuel and the engine starts running rich.

 

[Mac Attack]

Are you measuring the TPS voltages at the sensor or at the connector going to the ECU? You need to measure at the ECU.

 

[Mac Attack]

I clean my throttle body and check the idle at the same time (little too low eacv disconnected as the engine won't idle )

You need to set the idle correctly first by following the procedure on pages 11-84 through 11-85 of the Service Manual.

 

[metsuki]

Which open loop? There are at least two control loops on the NSX ECU, the idle control and the fuel mixture control. Idle control inputs are typically throttle position (TPS) and engine speed and on the NSX there appear to be brake and clutch interlocks. Fuel mixture control typically gets kicked out of closed loop if the engine coolant temperature is not up to normal operating spec and when the throttle approaches wide open. Some ECUs may be set up to switch to open loop fuel control if the engine RPM drops too low or if the MAP value is at the very low end (wide open throttle) or high end. It might also be possible that if the O2 readings go way out of limits the ECU goes into open loop; but, that should trigger a mixture error code.

As a matter of clarification, how do you know that the engine is going into open loop when you disconnect the TPS? Assuming that it is going into open loop fuel control an incorrect TPS reading could do that. The best way to confirm that the engine is or is not in closed loop would be to measure the O2 sensor voltage output. To do this, you would need to make a jumper arrangement that would allow you to measure the O2 sensor voltage while the engine is running. You need the jumper because the sensor needs to be connected to the ECU during the test. If the ECU is running in closed loop, the sensor voltage will typically be around 0.5 volts and will be constantly swinging above and below this value. If the sensor voltage remains permanently below 0.5 volts (reading lean) or above (reading rich) the engine is not in closed loop control or the sensor is dead. Sensors commonly fail to reading lean (0 volts) causing the ECU to add fuel and the engine starts running rich.

in fact i do not know if it is in close loop, that's what i read here and there about hondas in general "when you unplug TPS you get a CEL and you run open loop" ... i am sorry for my mistake, taking it for granted

the only fact for sure is when i unplug the TPS the car runs great and as soon as i plug it back i have the same issue coming back
i bought two TPS, maybe the first one was faulty?

 

[metsuki]

Are you measuring the TPS voltages at the sensor or at the connector going to the ECU? You need to measure at the ECU

You need to set the idle correctly first by following the procedure on pages 11-84 through 11-85 of the Service Manual.

no, in fact i measured it at the connector because i found nowhere how i can measure it at the ECU without disconnect it

i respected meticulously the procedure from the service manual about the idle

 

[metsuki]

thank you for your help

​

 

[Mac Attack]

Which open loop? There are at least two control loops on the NSX ECU, the idle control and the fuel mixture control. ...

I don't think he means open loop per se. I think he means the Fail-Safe Functions built into the ECU programming if a sensor goes faulty.

For the NSX, there are really two throttle angle sensors. In the SM, they are called the throttle angle sensor and the accelerator pedal angle sensor. Both vary in voltage output between ~0.5V to 4.5V. When you disconnect one of the sensors, the ECU kicks into Fail-Safe and uses the other sensor as input.

 

[Mac Attack]

no, in fact i measured it at the connector because i found nowhere how i can measure it at the ECU without disconnect it

i respected meticulously the procedure from the service manual about the idle

You need a fine tapered probe tip like described on page 23-7 of the SM.

 

[metsuki]

You need a fine tapered probe tip like described on page 23-7 of the SM.

ok will do!

For the NSX, there are really two throttle angle sensors. In the SM, they are called the throttle angle sensor and the accelerator pedal angle sensor. Both vary in voltage output between ~0.5V to 4.5V. When you disconnect one of the sensors, the ECU kicks into Fail-Safe and uses the other sensor as input.

so it will confirm that only my TPS can be faulty?

 

[Old Guy]

For the NSX, there are really two throttle angle sensors. In the SM, they are called the throttle angle sensor and the accelerator pedal angle sensor. Both vary in voltage output between ~0.5V to 4.5V. When you disconnect one of the sensors, the ECU kicks into Fail-Safe and uses the other sensor as input.

I think that is only on the drive by wire cars -actually two accelerator positions sensors and one TPS on a DBW car. I think the pre DBW cars just have the TPS and since the OP was mentioning looking for a pre 1995 TPS I am assuming that his car is not DBW.

 

[metsuki]

My car is a 93 maual car with 90000 miles

 

[Mac Attack]

so it will confirm that only my TPS can be faulty?

Perhaps, more likely, the wiring or connection is faulty.

 

[Mac Attack]

I think that is only on the drive by wire cars -actually two accelerator positions sensors and one TPS on a DBW car. I think the pre DBW cars just have the TPS and since the OP was mentioning looking for a pre 1995 TPS I am assuming that his car is not DBW.

You would think so from the description of the "other" TPS, but no, the pre-DBW's have two sensors. I removed one of them on my '92 when I deleted the TCS.

 

[Old Guy]

You would think so from the description of the "other" TPS, but no, the pre-DBW's have two sensors. I removed one of them on my '92 when I deleted the TCS.

Live and learn!

Out of curiosity I went back and looked at the schematic of the EFI in the .pdf version of the electrical section of the 1991 service manual (I like the 1991 manual because it has all the electrical schematics in one easy to find section - something missing in the paper version of my 1998 + manual). Sure enough, right on the separation line of the scanned pages is something listed as Acce Senso (the rest disappeared in the scan) right beside the throttle angle sensor (which I presume is the TPS). For good measure, the 1991 service manual also refers to a Pedal Angle Sensor and Accelerator pedal opening sensor (same thing) and for extra good measure in at least one spot double names the throttle angle sensor as a throttle valve opening sensor. I also feign innocence since the EFI portion of the 1991 service manual doesn't even mention the pedal angle / accelerator pedal sensor even though it is attached to the ECU.

From the diagrams in the SM the sensors appear to be on opposite sides of the throttle body. In which case I am guessing that the wipers rotate in the opposite direction, although the closed and open have the same test voltages on both switches. The OP has probably not messed up (like I did); but, given the multiple names they would want to make sure that they are testing the correct sensor - particularly since both sensors have the same open and closed voltages. The DBW cars have the great advantage that the accelerator position sensor is in a completely separate location than the TPS and I believe has the reverse open and close voltages so that makes it harder to mess up when testing (I clearly need that).

 

[Old Guy]

You need a fine tapered probe tip like described on page 23-7 of the SM.

This is the most complete and reliable way to confirm operation of the TPS. It confirms the TPS is working correctly and it confirms that the wiring from the TPS to the ECU is intact. It also confirms that the when the 1995+ TPS was modified for use on the earlier car that the pin connection arrangement remains correct.

 

[Old Guy]

As an addition to my post #6 on measuring the O2 sensor voltage. If the OP is going to back probe the ECU connector to confirm the TPS voltage, rather than fabricate jumpers to access the O2 sensor voltages from the engine bay, it would be infinitely easier to back probe the O2 sensor connections at the ECU. The front sensor voltage is on pin D14 and the rear sensor voltage is on pin D16 (on North American ECUs). Measure the voltage from the pin to a good chassis ground using a digital voltmeter. Do not use an old style analog volt meter. The input resistance of a 'meter' style volt meter is too low and will load the O2 sensor corrupting the voltage reading and messing everything up.

 

[metsuki]

This is the most complete and reliable way to confirm operation of the TPS. It confirms the TPS is working correctly and it confirms that the wiring from the TPS to the ECU is intact. It also confirms that the when the 1995+ TPS was modified for use on the earlier car that the pin connection arrangement remains correct.

Hi, i will change the TPS with the second one i bought and check voltage at the ECU connection just to be sure then i will check the wiring and the plug.
then i come back to you guys

if my problem can only come from the TPS it will be easier for sure

thanks A LOT for your help

 

[Old Guy]

Hi, i will change the TPS with the second one i bought and check voltage at the ECU connection just to be sure then i will check the wiring and the plug.
then i come back to you guys

if my problem can only come from the TPS it will be easier for sure

thanks A LOT for your help

So, you are planning to replace the TPS that you installed in Post #5 with another TPS? If that is correct, as a suggestion I would recommend that you check the voltage at the ECU terminals first. If you are getting the correct voltages at the ECU terminal, then I would not change the TPS. Changing things unnecessarily always runs the risk that something gets broken or messed up.

When you do the voltage test, don't just check the full closed and open voltage checks. Include the test in the 1991 service manual where you slowly open the throttle and watch the voltage on the wiper contact. The voltage must smoothly change from 0.5 volts to 4.5 volts as you press down on the accelerator pedal. Most ECUs provide a fuel enrichment function (the equivalent of the carburettor accelerator pump) that is signaled by how quickly the voltage is changing on the TPS output (how fast you push down on the accelerator pedal). If the TPS voltage at the ECU pin jumps up and down as you open the throttle (or is completely missing or reversed because of a wiring problem), the acceleration signal will be messed up or missing which can result in a slightly sluggish response when accelerating or brief periods when the ECU goes into fuel cut if a drop in TPS voltage tricks the ECU into de acceleration mode.

As an observation, the TPS is a pretty simple and reliable device. Unless your car has a very high number of kilometers on it with a lot of stop and go driving (it is repeated opening and closing of the TPS that results in wear), it is unlikely that the TPS is the cause of your problems. In particular, if you are having hesitation problems while driving at steady speed that is likely not a TPS problem.

Have you checked or considered the following?

- Did a previous owner install larger injectors or replace the 'chip' in the ECU? This can lead to idle problems and low throttle opening response problems.
- You said your car is a 1993. Have the injectors ever been cleaned. My understanding is that EU gasoline standards tend to be higher than in North America; but, after 25 years the cleanliness of the injectors might be suspect.
- Have you checked the condition of the grounding of the ignitor module? That is a problem that frequently shows up on older NSXs that have there engines cleaned regularly or are used in damp climates. Bad grounding results in misfires rather than hesitation; however, it does not cost anything to check the grounding.
- Have you checked the fuel pressure? Incorrect fuel pressure can result in all kinds of problems. When was the last time the fuel filter was changed? Even though the car only has 90,000 miles (not kilometers?), at 25 years the fuel pump may be getting tired.
- Has the valve clearance ever been checked? Incorrect valve clearances can result in driveability problems. Excessive valve clearance typically results in increased valve train noise. However, with a Taitec exhaust you may not notice the increased noise.

On my personal 'trouble causing list', all of these things would be higher up the list than the TPS. This does not guarantee that the TPS is not the problem, it is just that it would be one of the last places I would go looking.

Since you replaced the exhaust and deleted the catalytic converters, make sure that you don't have an exhaust system leak that is allowing air to enter into the exhaust system and potentially reach either or both O2 sensors. Atmospheric air reaching the O2 sensor will trick the ECU into running rich which will cause driveability problems. If the problem is really severe, this will normally trigger a MIL code; however, depending on the magnitude of the leak it may not be bad enough to trigger a code; but, still result in driveability problems.

 

[Mac Attack]

Wow Old Guy. Thank you for all of your well-written, thorough posts. You truly are an asset to this community!

Those are all great things to consider when running into this type of issue.

For this particular case, the issue goes away when the sensor is disconnected and the ECU goes into Fail-Safe mode. To me that pretty much says it is a sensor-related or wiring issue... probably either the TPS, a bad electrical connection, or an exhaust leak screwing with the O2 sensor(s) like you noted.

Granted, without looking at the ECU code, I don't know exactly what the Fail-Safe mode does since it isn't described in the SM. Does it bypass the fuel injector resistor and cause the fuel pump to run at full speed therefore masking the symptoms of a clogged fuel filter or failing fuel pump? It probably results in a richer fuel mixture and/or reduced timing to protect the engine, but I don't know.

 

[Old Guy]

For this particular case, the issue goes away when the sensor is disconnected and the ECU goes into Fail-Safe mode. To me that pretty much says it is a sensor-related or wiring issue... probably either the TPS, a bad electrical connection, or an exhaust leak screwing with the O2 sensor(s) like you noted.
.

The only caution here is that the OP noted that in post #9 , that he doesn't know that the ECU is going into open loop (or as you probably correctly identify - fail-safe mode). Fail safe mode for a particular sensor occurs when the signal voltage becomes 'abnormal', typically out of range. Once the ECU detects an abnormality, it substitutes a pre programmed fixed value for the sensor which allows the engine to continue to run. If the TPS has a fail-safe function, then disconnecting it should trigger the fail-safe feature. The more interesting question is if the ECU substitutes a fail-safe value for the TPS sensor, why does the engine work better?

As a side note, my interpretation of the limited description of the ECU functions is that the ECU implements fail-safe on a particular sensor basis. One bad sensor does not cause everything to go into fail-safe. There is a separate back-up function if the problem occurs in the ECU itself. The back-up function probably qualifies as a limp mode since the service manual describes '..permit minimal driving'.

When the OP first changed his TPS, he confirmed that the open and closed voltages (measured at the TPS) were correct. The pre OBDII cars check for voltage too high or too low on the TPS which triggers an error code (and fail-safe mode?). The OP has not reported a high or low voltage error code for the TPS which is not a surprise since his voltage measurements were within the acceptable range. That leads me back to the question of why is the ECU presumably working better with the fail-safe value for the TPS? The only thing I can speculate is that if the Vss and ground connections between the TPS and ECU got switched when a substitute TPS was installed, this would result in the ECU getting a 4.5 volt signal at closed throttle. This would still be an acceptable sensor voltage. However, I expect that the ECU would interpret this as a wide open throttle which would likely switch the ECU into open loop fuel control. If the ECU has accumulated very large idle fuel trims the engine may run sub optimally in open loop. This is my wild assed guess at how the TPS could be inducing a performance problem.

The ultimate test of the TPS will be, as you suggested, the back probe voltage measurements at the ECU with the connector plug inserted into the ECU. If these measurements show that the TPS voltage is 0.5 volts closed and 4.5 volts open and the voltage transitions smoothly as the throttle is opened, then the TPS is not the cause of the problem. Disconnecting the TPS and forcing a fail-safe value for the sensor reading might improve driveability by masking the initiating problem if it prevents the ECU from entering / exiting closed loop O2 control or idle control or throttle cut or who knows what other function is throttle dependent