Former User

I think that the response of my gas pedal at low throttle positions may be too sensitive (rpm increases too much with only small movements of the pedal). As I have already cleaned the throttle body, adjusted the throttle cable to spec, and replaced the throttle cable, today I decided to make some TPS measurements, even though no TPS or other codes are thrown. Below is what I found.

Resistance measurements of the TPS at closed throttle:
1) Between Grn/Blk wire (SG2) and Red/Blk wire (TPS) - 838 Ohms (spec = 500-900 Ohms)
2) Between Red/Blk wire (TPS) and Yel/Blu wire (VCC2) - 2524 Ohms (spec = ~4500 Ohms)

Voltage outputs by the TPS measured between the Grn/Blk wire (SG2) and Red/Blk wire (TPS):
0.459V at closed throttle (spec = ~0.5V)
4.63V at wide open throttle (spec = ~4.5V)

***My question is whether anybody considers the red highlighted resistance value to be out of spec enough to cause an overly sensitive gas pedal at low throttle?

Former User

I also have a 97 Civic DX (manual transmission) that does not have an over-sensitive throttle, so I made the same TPS resistance measurements at closed throttle on it. Here are the results:

1) Between Grn/Blk wire (SG2) and Red/Blk wire (TPS) - 876 Ohms (spec = 500-900 Ohms)
2) Between Red/Blk wire (TPS) and Yel/Blu wire (VCC2) - 4760 Ohms (spec = 3600-5400 Ohms)

Interestingly, unlike my 00 Civic, both TPS resistance readings for my 97 Civic are within spec.

slowcivic2k

I've seen this usually when the throttle is adjusted too tight. Which doesn't apply here due to the normal closed throttle position voltage. Has the TPS been reclocked? Does the transaxle shift at a higher speed than what you would consider "normal"

I'm sure you would have mentioned cable binding if there was some, just have to ask.

The voltages that your sensor produces will not set the mil, but the part that is harder is the in between 0 throttle and 100% throttle.

If your meter resolves fast enough, have someone very slowly tip the throttle and record the (whoops voltage) readings as it opens. Normally it is a smooth progression, but if you see the resistance jump up or down suddenly, it will indicate transient throttle response, which is what it sounds like you have.

If you have a scan tool with live data recording tip the throttle yourself and monitor the voltage increase. Again a smooth progression, any dips should be noted and gone over again, only slower to provoke it.

Former User

But what do you think about the one TPS reading that appears to be substantially out of spec?

Former User

By the way, these diagrams show the troubleshooting steps that I did (even though code P0122 is not thrown):

slowcivic2k

[QUOTE=RonJ@HT;44698222]By the way, these diagrams show the troubleshooting steps that I did (even though code P0122 is not thrown):

I will assume that the throttle body plate does not bind and operates freely when you replaced the throttle cable.

Is the throttle sheath mounted to the firewall correctly?

I was pointed at an intermittent brush problem inside the TPS itself. Since most of use spend more of our time nearer the closed end, the commutator tends to wear faster on that side. I would replace it with a known good substitute and recheck the driveability portion. Clear the kieep-alive memory and relearn.

Former User

slowcivic2k

Yes, the resistance reading immediately caught my attention.

I would warrant replacement based on your driveability complaints, and retest. Nothing else makes sense.

I would use a NAPA ECHLIN brand sensor if I could not source an replacement OEM one, used or new. I haven't had this problem on many Honda's, (Not including the ones that people break, only one's that legitimately failed) mainly on Chrysler's. I've always used ECHLIN after the bad luck I got with the Oreilly's Master Pro brand.

I would assume you could jimmy the TPS retaining nuts off and source bolts to replace them, they aren't really reusable. I always mangle them off and use bolts. Make sure it has a gasket in the box.

Former User

Thanks. I'd like a new OEM TPS, but I have not been able to find one alone as it appears to be sold in conjunction with the entire throttle body. And I'll look into the NAPA TPS and ask about the gasket.

Former User

Update

Cliff notes - I was misled to believe that the resistance of my 00 Civic TPS was out of spec because I made TPS resistance measurements at ECU connector C rather than at the TPS itself.
-------------------
The "apparent" out-of-spec reading for test 2 below was taken from ECU terminals C27 and C28. When I repeated the equivalent test on terminals 2 and 3 of the unplugged TPS itself, the resistance reading was now different and within spec (5000 Ohms).

Getting a different resistance reading from ECU connector C versus the TPS itself was initially confusing to me because the resistance measurement at the ECU connector of my 97 Civic was within spec (see below).

This led me to compare the TPS circuits of my 00 and 97 Civics. I found that there was one major difference between the circuits. The wire from ECU terminal C28 for the 00 Civic also branches to the fuel tank pressure sensor (see diagram below). The equivalent ECU terminal for the 97 Civic does not because it lacks a fuel tank pressure sensor.



In conclusion, the fuel tank pressure sensor in the 00 Civic led to a false out-of-spec TPS resistance reading from ECU connector C. The resistance measurement taken directly at the unplugged TPS was within spec because taking the measurement this way eliminated effects from the fuel tank pressure sensor. Therefore, the TPS is fine.

slowcivic2k

I still suspect a worn contact patch inside the TPS, there isn't much else logically that makes sense

That will be harder to diagnose, because it requires a meter than can resolve fast. Measure the resistance between pins 1 and 2 with the TPS disconnected and tip the throttle slowly to wide open. A graphing meter would be ideal.

Former User

I plan to replace the TPS anyway, but based on the results, I don't have a high expectation for any major change.

sk8shorty012

iTrader: (3)

I agree with slowcivic2k, sounds to me like there's a "hotspot". The best way would be to verify the linearity as the throttle valve is slowly opened as stated previously.

Former User

Yesterday I replaced the old TPS. All auto parts stores in my area only sell the Dorman brand TPS, so I bought it. It came with two nice replacement hex screws and a new gasket for $45. I removed the throttle body and etched a slot into the top of each original TPS screw using a Dremel tool and cutting wheel. The old screws easily came out with a flat head screw driver. Installation of the new TPS and gasket on the throttle body was a snap, as was reinstalling the throttle body with a new gasket ($2) on the intake manifold. I also made sure to torque the throttle body nuts and bolts to spec (16-lb-ft).

To calibrate the TPS output voltage, I poked two sewing needles into the rear of the unplugged TPS connector in such a way to make electrical contact with either the Red/Black wire (-) or the Yel/Blu wire (+) (=connector back probing). After plugging the TPS connector into into the TPS, I clipped the + and - leads of my multimeter to the corresponding protruding sewing needles, making sure the needles and probes did not make contact with one another. I then turned the key to ON(II) to measure the TPS output voltage at closed throttle. It read 0.36V. I then loosened the TPS screws (just enough so that the TPS could be turned with resistance) and rotated the TPS slightly until the voltage read ~0.45V. After tightening the TPS screws, I verified that the voltage did not change. I then read the voltage at wide open throttle, which registered 4.7V. When done, I disconnected the multimeter leads from the needles and pulled the needles out of the rear of the TPS connector. The final step was resetting the ECU by removing the 7.5A hood Back Up fuse for more than three minutes.

The low-rpm throttle response now seems smoother and less over-sensitive.

sk8shorty012

iTrader: (3)

Very good information for TPS diagnostics and replacement

slowcivic2k

Sounds good man