TPS Diagnostic Help Needed
03-30-15, 05:22 PM
#1
TPS Diagnostic Help Needed
I may have a TPS issue. There are two tests for the TPS in the FSM, and I've seen two outlined in various threads and external sites. One checks what I'm led to believe the ECU sees via the voltage method at the diagnostic plug (substituting voltmeter for the special tool light dongle setup), the other checks resistance at the switch itself where it plugs into the harness. I've checked/set the TPS with both methods now, and the two settings do not overlap. Additionally, it seems the switch is reading out of spec.
Someone please check my work here:
- Regardless the settings/adjustment, it idles like crap. Little backfires, stumbling, and wandering between 500-900 RPM, mostly staying in the 600-800 RPM range. After all my dicking with it, it will now even die if I don't feather the throttle if it's between hot and cold. I can't say it has idled smooth since I've owned it, but I chalked that up to the leaky injectors and the dumbass fuel system modifications courtesy of the PPO. There are still two presumedly fuel pump related wires connected and operated via a rocker switch under the steering column. One is obviously power to the fuel pump, but I don't know what the other is (it's right next to it on the same plug under the steering column).
- If I set the TPS such that the throttle closed (car off/key off) position is 1KΩ, the diagnostic plug doesn't check out (no voltage on either lead) and vice versa. I want to say resistance at the plug is something like 1.4KΩ where the voltage checks out.
- I'm only getting 10 and change volts with the car off/key on versus the 12V stated in the one TPS check writeup I found online. I initially wrote this off as the light method as stated in the FSM would most likely be voltage agnostic. It's just a question of whether or not voltage is available to light the test lights. Maybe it matters, maybe it doesn't.
- WOT at the sensor plug results in an Ω reading that is NOT as expected. I just went and double checked. If I fully depress the TPS with my finger, it reads ~500Ω. Fully out (effectively WOT), it also reads 500Ω, and the max (in the middle of the travel range) is only 1.4KΩ. I did this with the engine warm, not hot, so that may affect it a little, but I'd be willing to bet only a little.
Could someone verify my test methodology and interpret my findings? I think my TPS is shot, but I don't want to drop $175 only to find that I misread something.
Also, and maybe more importantly, could someone clue me in on the inner workings of this particular system? What, exactly, does the TPS do in relation to engine operation. I know it tells the ECU where the throttle is to some degree (the throttle has more travel than the switch, so there's a point after which the ECU doesn't care) , and eventually the injectors fire at some time with some quantity of fuel for some reason. The ECU is a black box in my mind, and I don't like that. Knowing the hows and whys will help me with this and future diagnostic issues.
Also, and related to the black box issue, what are the long term ramifications of driving the car with a faulty TPS. Obviously it still runs (well), it just idles poorly. What negative things could happen as a result of the sensor reading so far out of spec?
Someone please check my work here:
- Regardless the settings/adjustment, it idles like crap. Little backfires, stumbling, and wandering between 500-900 RPM, mostly staying in the 600-800 RPM range. After all my dicking with it, it will now even die if I don't feather the throttle if it's between hot and cold. I can't say it has idled smooth since I've owned it, but I chalked that up to the leaky injectors and the dumbass fuel system modifications courtesy of the PPO. There are still two presumedly fuel pump related wires connected and operated via a rocker switch under the steering column. One is obviously power to the fuel pump, but I don't know what the other is (it's right next to it on the same plug under the steering column).
- If I set the TPS such that the throttle closed (car off/key off) position is 1KΩ, the diagnostic plug doesn't check out (no voltage on either lead) and vice versa. I want to say resistance at the plug is something like 1.4KΩ where the voltage checks out.
- I'm only getting 10 and change volts with the car off/key on versus the 12V stated in the one TPS check writeup I found online. I initially wrote this off as the light method as stated in the FSM would most likely be voltage agnostic. It's just a question of whether or not voltage is available to light the test lights. Maybe it matters, maybe it doesn't.
- WOT at the sensor plug results in an Ω reading that is NOT as expected. I just went and double checked. If I fully depress the TPS with my finger, it reads ~500Ω. Fully out (effectively WOT), it also reads 500Ω, and the max (in the middle of the travel range) is only 1.4KΩ. I did this with the engine warm, not hot, so that may affect it a little, but I'd be willing to bet only a little.
Could someone verify my test methodology and interpret my findings? I think my TPS is shot, but I don't want to drop $175 only to find that I misread something.
Also, and maybe more importantly, could someone clue me in on the inner workings of this particular system? What, exactly, does the TPS do in relation to engine operation. I know it tells the ECU where the throttle is to some degree (the throttle has more travel than the switch, so there's a point after which the ECU doesn't care) , and eventually the injectors fire at some time with some quantity of fuel for some reason. The ECU is a black box in my mind, and I don't like that. Knowing the hows and whys will help me with this and future diagnostic issues.
Also, and related to the black box issue, what are the long term ramifications of driving the car with a faulty TPS. Obviously it still runs (well), it just idles poorly. What negative things could happen as a result of the sensor reading so far out of spec?
Last edited by spectre6000; 03-30-15 at 05:25 PM.
03-30-15, 06:05 PM
#2
Moderator
iTrader: (3)
Join Date: Mar 2001
Location: https://www2.mazda.com/en/100th/
Posts: 30,780
Received 2,565 Likes
on
1,824 Posts
the TPS does two overt things.
the first is to control the two Air Control Valve solenoids. the wires for the TPS test lights are literally the ECU's output to the solenoids, its literally the same wires. this changes where the air pump air goes and at low rpm it seems to really mess with how the car runs.
the second is to tell the ecu the throttle is closed, we've seen a few tables* from the ecu, and they usually start at an rpm higher than idle....
it may also do some other things, but those are the overt things.
oh so set it up so its right at idle, and see where that gets you
*as an aside, the timing is actually a table, but the fuel is actually computed on the fly. there is a fuel correction table, but the ecu comes up with a raw number, and then goes to look at corrections
the first is to control the two Air Control Valve solenoids. the wires for the TPS test lights are literally the ECU's output to the solenoids, its literally the same wires. this changes where the air pump air goes and at low rpm it seems to really mess with how the car runs.
the second is to tell the ecu the throttle is closed, we've seen a few tables* from the ecu, and they usually start at an rpm higher than idle....
it may also do some other things, but those are the overt things.
oh so set it up so its right at idle, and see where that gets you
*as an aside, the timing is actually a table, but the fuel is actually computed on the fly. there is a fuel correction table, but the ecu comes up with a raw number, and then goes to look at corrections
Last edited by j9fd3s; 03-30-15 at 06:11 PM.
03-30-15, 06:18 PM
#3
When you are checking the TPS using the ohm method are you unplugging the TPS as you should or are you doing it plugged in?
How about the voltage method. You drive the car for at least 20 minutes followed by parking the car and w/key to on you measure the voltage of the Green/Red wire and it should read very close to 1 volt. If it doesn't then you use the TPS adjustment screw so it does. Under either method of adjustment the engine needs to be as hot as it can possibly get or you will get a false reading. And under the voltage method the range goes from 1 volt at idle to about 4.5 volts w/the plunger fully extended. When checking for a smooth sweep the analog meter is more accurate than the digital counterpart.
How about the voltage method. You drive the car for at least 20 minutes followed by parking the car and w/key to on you measure the voltage of the Green/Red wire and it should read very close to 1 volt. If it doesn't then you use the TPS adjustment screw so it does. Under either method of adjustment the engine needs to be as hot as it can possibly get or you will get a false reading. And under the voltage method the range goes from 1 volt at idle to about 4.5 volts w/the plunger fully extended. When checking for a smooth sweep the analog meter is more accurate than the digital counterpart.
03-30-15, 06:29 PM
#4
I just came back in from the garage, laptop (FSM) in hand for verification purposes. According to the manual p. 4A-50 I got it hot, turned the engine off, then turned the key back on, and checked the test plug (between the air box and the strut tower). It was still set according to my previous adjustment (30 minutes prior). Jumping one set of leads read ~10.5V, and the other read ~0.1V. Actuating the throttle caused the other set of leads to read ~10.5V as well. This should indicate the TPS set screw is properly adjusted.
Then I turned the key off and unplugged the TPS from the harness according to the FSM. I get (range from idle position to WOT) 1.276KΩ-∞ across the A/B terminals (as labeled in the FSM) and ∞-521Ω across A/C. A/B spec is 1KΩ-5KΩ ±1KΩ, and A/C is 5KΩ ±1KΩ-5KΩ ±1KΩ.
I was not able to go on to set the idle speed or the idle mixture, because this results in a hunting and rough idle.
Then I turned the key off and unplugged the TPS from the harness according to the FSM. I get (range from idle position to WOT) 1.276KΩ-∞ across the A/B terminals (as labeled in the FSM) and ∞-521Ω across A/C. A/B spec is 1KΩ-5KΩ ±1KΩ, and A/C is 5KΩ ±1KΩ-5KΩ ±1KΩ.
I was not able to go on to set the idle speed or the idle mixture, because this results in a hunting and rough idle.
03-30-15, 06:47 PM
#5
Just try the voltage method and see what you get. I can tell by the ohm reading that the idle reading in volts is not going to be 1 volt and this is why you use the TPS adjustment screw to get it into spec range. When measuring the G/R wire you backstab the meter lead into the back of the plug (everything connected) in case you weren't doing it this way.
03-30-15, 07:11 PM
#6
If I adjust it to 1KΩ at idle, the test plug gets no voltage on either terminal set. If I adjust such that the test plug sees voltage like it should, the TPS is at about 1.4KΩ. The two specified ranges are mutually exclusive, and in either case it idles like crap.
How do the TPS switches typically fail?
How do the TPS switches typically fail?
03-30-15, 07:20 PM
#7
How many wires are in your test plug as there should be three of them. The B/W wire gets 12 volts w/key to on. The Blue/Yellow wire should read below 2 volts w/key to on if the TPS is set correctly. The Blue/Red wire should read 12 volts w/key to on and the TPS adjusted properly.
The spec sheet for the ECU pinouts suggests the G/R wire at the ECU should read 1 volt as stated earlier. Measuring this pin (2G) would give you a heads up.
The spec sheet for the ECU pinouts suggests the G/R wire at the ECU should read 1 volt as stated earlier. Measuring this pin (2G) would give you a heads up.
Last edited by satch; 03-30-15 at 07:25 PM.
Trending Topics
03-30-15, 07:21 PM
#8
Green plug. Three wires. Comes off the harness at the same point as the variable resistor (idle mixture adjustment pot), and seems to live underneath it like a little electroll. I get ~0.1V across the one, and ~10.5V on the other when adjusted with the key on, engine off. Actuating the throttle gives them both ~10.5V.
03-30-15, 07:30 PM
#9
I don't know if this has any bearing on anything, but once I get it going, it's fine. Power is good, revs all the way up without any hesitation into the secondaries... I got 21.71 MPG on the only full tank I've driven so far. It's just crappy at idle... Still learning the platform....
03-30-15, 07:39 PM
#10
Is the car a turbo or NA? Going back to the voltage method how about checking it. Don't focus on the Green check connector but the Green/Red wire of the TPS. Your wiring might be aged and showing it. As stated earlier, the ECU wants to see 1 volt w/key to on and the engine as hot as it can get. If the wire between the TPS connector and the ECU is poor then the ohm reading will have to be higher than normal because the higher the resistance the lower the voltage goes. If you set the TPS to 1 ohm and then quickly measured the G/R wire in volts I would tend to think the reading will likely be below 1 volt.
03-30-15, 07:56 PM
#11
86 NA in sig. The TPS is the single switch variety with a total of 3 leads.
Either your mixing Ω/V, your decimals are floating around, or I'm beyond confused.
I'm focusing solely on the terminals and general relationships. I haven't paid attention to the colors of wires at all.
The TPS has three terminals/leads associated with it. One is power, the other two are signal out and vary their output by resistance. The spec for these resistances is 1KΩ at the adjusted idle to 5KΩ ±1KΩ fully extended (WOT), and 5KΩ ±1KΩ at adjusted idle and the same fully extended (WOT) for the other. What I'm getting (assuming I adjust it to 1KΩ at idle) is 1KΩ at adjusted idle and ∞Ω at WOT on the one and ∞Ω - 521Ω (or .521KΩ to keep the decimal in the same place). Further, the total range of resistance is from a little below 500Ω (.5KΩ) fully depressed to a maximum resistance of slightly more than 1.4KΩ somewhere in the middle, then it drops back down to about 500Ω (.5KΩ) at the full extension before losing connectivity entirely. No voltage exists save what my tester is spitting out to measure the resistance because the switch is completely unplugged and does not produce any current of its own.
Meanwhile, at the test plug (where one would conduct the test with the two LEDs), if the TPS is reading the specified 1KΩ, neither lead sees more than a small fraction of a volt (I never measured resistance as it's not important). If I adjust it such that there is an appropriate voltage signal at the test plug, the TPS reads 1.276KΩ.
The two do not seem to overlap. If one is in spec, the other is necessarily out and vice versa. Either way, it idles like crap.
Either your mixing Ω/V, your decimals are floating around, or I'm beyond confused.
I'm focusing solely on the terminals and general relationships. I haven't paid attention to the colors of wires at all.
The TPS has three terminals/leads associated with it. One is power, the other two are signal out and vary their output by resistance. The spec for these resistances is 1KΩ at the adjusted idle to 5KΩ ±1KΩ fully extended (WOT), and 5KΩ ±1KΩ at adjusted idle and the same fully extended (WOT) for the other. What I'm getting (assuming I adjust it to 1KΩ at idle) is 1KΩ at adjusted idle and ∞Ω at WOT on the one and ∞Ω - 521Ω (or .521KΩ to keep the decimal in the same place). Further, the total range of resistance is from a little below 500Ω (.5KΩ) fully depressed to a maximum resistance of slightly more than 1.4KΩ somewhere in the middle, then it drops back down to about 500Ω (.5KΩ) at the full extension before losing connectivity entirely. No voltage exists save what my tester is spitting out to measure the resistance because the switch is completely unplugged and does not produce any current of its own.
Meanwhile, at the test plug (where one would conduct the test with the two LEDs), if the TPS is reading the specified 1KΩ, neither lead sees more than a small fraction of a volt (I never measured resistance as it's not important). If I adjust it such that there is an appropriate voltage signal at the test plug, the TPS reads 1.276KΩ.
The two do not seem to overlap. If one is in spec, the other is necessarily out and vice versa. Either way, it idles like crap.
03-30-15, 08:04 PM
#12
There are two methods to setting the TPS. Ohms or voltage. Just look at the wiring diagram spec sheet of all the pinouts of the ECU and it plainly states the G/R wire of the TPS should read 1 volt (not ohms), thus you can check the TPS while plugged in, key to on, engine as hot as it can possibly get. One meter lead (red) goes to the G/R wire and the other meter lead (black) to the negative battery terminal. It should read 1 volt. In your case either the connector or wiring is not proper thus causing you to have to set the TPS to higher than what the ohm reading should be which means if you measured the G/R wire in voltage it would likely not be the voltage required by the ECU.
And I did leave out the k.
Post #6 explains it in a different perspective:
https://www.rx7club.com/2nd-generati...w-idle-762997/
And I did leave out the k.
Post #6 explains it in a different perspective:
https://www.rx7club.com/2nd-generati...w-idle-762997/
Last edited by satch; 03-30-15 at 08:18 PM.
03-30-15, 08:05 PM
#13
Look at your TPS, follow the wires down to the plug, should land at the front of the engine, near the tstat housing, thats where you check the voltage, and adjust from there
Satch is talking VOLTS, you seem to be talking OHMs
You also seriously over analyzing the whole thing too
Satch is talking VOLTS, you seem to be talking OHMs
You also seriously over analyzing the whole thing too
03-30-15, 08:35 PM
#14
I guess I'll give it another shot in the morning... The switch seems to be pretty thoroughly off spec though... It should read 1KΩ (adjusted, so presumedly there's space below that) up to 4KΩ-6KΩ smoothly and without dead spots. Mine goes from (adjusted) 1KΩ to 1.4KΩ before dropping down to .5KΩ and then ∞Ω at the maximum extension... Doesn't read in the same ballpark (or zipcode) as the spec. Not really a lot the ECU can do with that I wouldn't imagine.
What could happen if it's run like that? It seems to behave fine (at least within my frame of reference), but obviously idles poorly. What damage could occur to the engine?
What could happen if it's run like that? It seems to behave fine (at least within my frame of reference), but obviously idles poorly. What damage could occur to the engine?
03-31-15, 10:55 AM
#16
In what is hopefully a positive sign, I was able to get the TPS to respond and set via the voltage at the sensor itself method. 1.00V at idle set position to 4.6V at WOT. That checked out at the test plug (between the air box and the strut tower) via the simulated light test. It did not check out at all measuring resistance at the TPS itself unplugged. Not even close. That worked out to 1.4KΩ at adjusted idle and swept up to non-conductivity on the one lead, and non-conductivity down to 500Ω on the other lead... Way off, and not even resembling the general relationship that should exist according to the manual. The voltage is right though, so there's that. This was done with the plug up by the battery jumpered.
I tried to take video of this whole exercise, but my phone decided it was more interested in being a paperweight. Despite all my photo management, it tells me the memory is full... I'm backing it up to the computer after deleting all but a few essentials, and then I'll try again.
After some phone management and a few attempts at video recording, I went to start it, and it would fire up, rev to something higher than idle, then die completely. If I feathered the throttle I could keep it running, and it would do so until it got fully hot again, at which point it would hunt and lope. The idle adjust screw had very little (if any) effect.
As soon as my phone finishes its paperweight routine, I'll go out and try to get video again. Hopefully the quality of the poor idle (and maybe some other factors) could help explain what it is I'm experiencing to those with more experienced eyes and ears.
I tried to take video of this whole exercise, but my phone decided it was more interested in being a paperweight. Despite all my photo management, it tells me the memory is full... I'm backing it up to the computer after deleting all but a few essentials, and then I'll try again.
After some phone management and a few attempts at video recording, I went to start it, and it would fire up, rev to something higher than idle, then die completely. If I feathered the throttle I could keep it running, and it would do so until it got fully hot again, at which point it would hunt and lope. The idle adjust screw had very little (if any) effect.
As soon as my phone finishes its paperweight routine, I'll go out and try to get video again. Hopefully the quality of the poor idle (and maybe some other factors) could help explain what it is I'm experiencing to those with more experienced eyes and ears.
03-31-15, 12:37 PM
#17
Could be a vacuum leak. The water thermosensor might be an issue as well. Did you set the TPS w/the engine cold or as hot as it can get? If you set it hot did you test the TPS voltage of the G/R wire to see if it was still 1 volt. And lastly, did you remember to pull the jumper wire from the initial set coupler?
Last edited by satch; 03-31-15 at 12:39 PM.
03-31-15, 01:09 PM
#18
I'm on the phone with Clokker now comparing notes. He watched me do the vacuum lines, and between the second set of educated eyes, a third set of slightly less educated (at least on the specific platform) eyes, and my particular methodology, we are both of the opinion that a vacuum leak is a very remote possibility.
The water thermosensor plug was hit with some electro-grease, but no test was done on it one way or another. Steve says there's some way to test it, but he didn't know off the top of his head.
I did the TPS adjustment fully hot, and the voltage was set at 1.00V hot with a clean sweep to a range maximum of 4.6V at WOT (or wherever the throttle is in its travel where the cam leaves the sensor).
There is no jumper in the initial set coupler.
When I called Clokker up (about the time I started responding to the post), I went back out there with him on the line (not that he could hear anything), and started it up. The temp gauge read just barely above stone cold. It started, revved to 1500RPM or so very briefly, then dropped to 500 RPM and promptly shut off. I started it again, and was able to feather the throttle to keep it going, and it drove fine as long as it didn't idle. If I came to a gate or a stop sign or something, I had to heel-toe it to keep the engine running until it got fully up to temp. Once it was fully hot, it would idle like it has all along: between 800-900RPM, hunting around and backfiring in the exhaust.
It has idled that way (the 800-900 RPM hunting/backfiring) the entire time I've had the car (about 1 1/3 tanks of fuel so far). I initially chalked it up to some stupid fuel system modifications that were intended to compensate for leaky fuel injectors. I replaced the injectors and removed the more egregious modifications this weekend, and it still idled poorly. Once I set about trying to set the TPS, it started dying all together. Now that the TPS is for sure set, it dies cold, and idles like crap hot.
I have a video demonstrating the idle quality (along with some other video documentation) uploading now... My technology is fighting me almost as hard as the car is right now...
The water thermosensor plug was hit with some electro-grease, but no test was done on it one way or another. Steve says there's some way to test it, but he didn't know off the top of his head.
I did the TPS adjustment fully hot, and the voltage was set at 1.00V hot with a clean sweep to a range maximum of 4.6V at WOT (or wherever the throttle is in its travel where the cam leaves the sensor).
There is no jumper in the initial set coupler.
When I called Clokker up (about the time I started responding to the post), I went back out there with him on the line (not that he could hear anything), and started it up. The temp gauge read just barely above stone cold. It started, revved to 1500RPM or so very briefly, then dropped to 500 RPM and promptly shut off. I started it again, and was able to feather the throttle to keep it going, and it drove fine as long as it didn't idle. If I came to a gate or a stop sign or something, I had to heel-toe it to keep the engine running until it got fully up to temp. Once it was fully hot, it would idle like it has all along: between 800-900RPM, hunting around and backfiring in the exhaust.
It has idled that way (the 800-900 RPM hunting/backfiring) the entire time I've had the car (about 1 1/3 tanks of fuel so far). I initially chalked it up to some stupid fuel system modifications that were intended to compensate for leaky fuel injectors. I replaced the injectors and removed the more egregious modifications this weekend, and it still idled poorly. Once I set about trying to set the TPS, it started dying all together. Now that the TPS is for sure set, it dies cold, and idles like crap hot.
I have a video demonstrating the idle quality (along with some other video documentation) uploading now... My technology is fighting me almost as hard as the car is right now...
03-31-15, 01:42 PM
#19
The water thermosensor ECU pin should read between 2-3 volts w/key to on and the voltage progressively drops to about .5 volts fully warmed. Pin 2I of the ECU and the wire is Green/White.
And if you check this ECU pin you might as well test the pin for the TPS at pin 2G (Green/Red wire).
And the BAC should prevent the idle from dropping below 750 rpm.
And check the variable resistor as it can help determine the fuel mixture when the car is idling.
And check pin 2J (Green/Orange wire) as it's for the AFM intake air temp sensor. Should read 2 to 3 volts w/the air temp at 68 degrees. W/key to on and the temp cooler I believe the voltage value drops.
And if you check this ECU pin you might as well test the pin for the TPS at pin 2G (Green/Red wire).
And the BAC should prevent the idle from dropping below 750 rpm.
And check the variable resistor as it can help determine the fuel mixture when the car is idling.
And check pin 2J (Green/Orange wire) as it's for the AFM intake air temp sensor. Should read 2 to 3 volts w/the air temp at 68 degrees. W/key to on and the temp cooler I believe the voltage value drops.
Last edited by satch; 03-31-15 at 01:47 PM.
03-31-15, 02:01 PM
#22
There are 3 ECU plugs 1A is in the largest, 2A in the middle sized plug, and 3A in the smallest ECU plug. Pin 2A would be at the top right position looking at the plug from the back of it. Pin 2B is directly below 2A. Thus 2C would be to the left of pin 2A and so on.
The variable resistor ranges from 1 to 4 volts. The screw only turns from about 9:00 o'clock to 3:00. The initial set coupler needs to be jumpered and engine fully warmed to set it. It is normally set just past the half way mark or close to 1:00 o'clock. Do not attempt to turn the screw further past the two limits or it will break. You might want to mess with it when the engine is warmed and the idle is bouncing a bit.
03-31-15, 02:22 PM
#23
OK. Sweet. Found the ECU. Now I know how the pins are mapped. Looks like a (relatively) quick and easy way to test a lot of things at once. If I'm checking voltages at specific wires (2I, 2G, etc.), what am I checking the voltage with? What else am I probing to complete the circuit? Is there a list somewhere of things that can be checked here and the ranges to be expected?