Engine codes 24 and 15

Tiny
BKAEMPFE1989
  • MEMBER
  • 1995 JEEP WRANGLER
  • 4 CYL
  • 4WD
  • MANUAL
  • 125,000 MILES
I currently have code 24 TPS sensor and 15 Vehicle speed sensor. When we start it for the first time its starts easier still sluggish but doesn't want to idle right. Really low idle and you have to be really light on the gas to get rpm's up. If you give it lots of throttle at once it bugs down and wants to die. When it is running after we get rpm's up and keep my foot on the throttle its super rough with a lot of motor shake. So far I started with cleaning the throttle body really good and took apart anything that needed to to get clean sensor and a cover on the throttle body. I have replaced and thoroughly checked TPS sensor. I will try and post the link of how I tested it. I used volt meter and got a clean sweep of the needle up to 4.5 volts at wide open throttle. Had.8 volts at idle position. Key in engine off while doing this. And checking for power I had 5.4 volts. Also changed spark plugs and injectors. I have spark tester and every plug is getting a bright spark. Timing seems good. I actually did spark plugs and injectors first. Along with dropping the tank and draining old gas putting fresh stuff in. Drained the lines. It's running on good gas now for sure. I see good amount of gas coming out of check valve on fuel rail. Also didn't have one spark plug seated correctly and it was spraying fuel. (Fixed that issue) but I'm pretty sure it's getting correct amount of fuel. I don't not have a fuel tester. If you have any ideas on why I still have a TPS code maybe it will help problems. I shouldn't have code after all test on TPS checked out plus new sensors. Had a new senor already so installed it. And yes, I did disconnect the battery to erase codes.
Thursday, June 18th, 2020 AT 10:21 AM

3 Replies

Tiny
CARADIODOC
  • MECHANIC
  • 33,873 POSTS
You're a number of steps ahead in the testing of the throttle position sensor, however, there is one test you overlooked. The only way you can get the proper signal voltage range of roughly 0.5 to 4.5 volts as you sweep the throttle from idle to wide-open-throttle is the 5.0 volt supply has to be good, the ground circuit has to be good, and the TPS has to be good. Be aware any of those things can be intermittent. The clue is the intermittent defect will be detected, the fault code will set again, and the Check Engine light will turn on when the problem occurs. That could be minutes or days from now. When you have a solid defect, the fault code will return as soon as you turn on the ignition switch and / or erase the code.

The one test you missed is to determine whether that 0.5 to 4.5 signal voltage is getting to the Engine Computer. The easiest and fastest way to do that is to use a scanner that displays live data. If there is a break in the signal wire, you'll still see the normal range of signal voltage at the sensor but not at the computer. In fact, when the signal wire is broken, due to the interconnected circuitry inside the computer, the voltage seen on that terminal can "float" to some random value. If that random value happens to fall within the normal range, 0.5 to 4.5 volts, it will accept that as correct and try to run on that. To prevent that, there is going to be a "pull-up" or "pull-down" resistor inside the computer to place 5.0 or 0.0 volts on the signal terminal of the computer. That resistor is so huge electrically that under normal operation, it has no effect on the circuit. It's only once the signal wire is broken that the resistor comes into play and puts the desired voltage on the signal terminal. 0.0 and 5.0 volts are both indicative of a defective condition that triggers a fault code. For teaching theory, it is anything outside that 0.5 to 4.5 volt range that triggers a fault code.

Rather than tear into the large connector at the Engine Computer, I'd put my effort into finding a scanner to read the TPS voltage. That doesn't have to be a $5000.00 scanner. There are some small ones that cost as little as $35.00. They have very limited functionality, but their main drawback is they're slow. They aren't practical for reading the TPS sweep when you're looking for a momentary dropout. You'll have swept past the dropout long before it shows up on the screen, and since these take a reading about once every three to five seconds, most dropouts will be missed. This is where an analog, pointer-style voltmeter is the best tool.

What the cheap scanner is perfect for is showing when the signal voltage you found at the TPS is different than what the computer is seeing. Most Chrysler computers use pull-up resistors, so you'll find 5.0 volts at the computer when there's a break in the signal circuit from the sensor. When the circuit is working properly, you found 0.8 volts at the sensor. That required the sensor to be plugged in. If you unplug the sensor, all three circuits, the 5.0-volt feed, the ground, and the signal circuit all have a break in them. We're interested only in the signal circuit, but you can measure all three for fun. You'll still find 5.0 volts on the feed wire. The ground wire will have 0.2 volts, not 0.0 volts, because the ground circuit goes through monitoring circuitry inside the computer. The key is the voltage on the signal wire. When there is no defect in the wiring, you'll find 5.0 volts there due to the action of the pull-up resistor. A lot of import models use pull-down resistors, so you'll find 0.2 volts on them. Either one gets detected as a defective condition, and a fault code is set.

I have to describe this carefully to not confuse the issue. With the TPS unplugged, you can expect to find 5.0 volts on the signal wire, thanks to the pull-up resistor. When there's a break in the signal wire, you will find 0.0 volts at the signal wire. That is proof there's a break. The clinker is when that break is caused by something that is very high in resistance. A perfect example is a badly-corroded mating pair of connector terminals. In theory, digital voltmeters do not need current flow through the circuit to take their readings. They are simply an electrical pressure gauge, just like the pressure gauge on a compressed air line in the shop. No air actually flows through the gauge for it to do its job. Imagine if there was a restriction in the air pipe so it was 99 percent blocked. The gauge at the end of the line would still show full pressure, ... Until you tried to run an air tool, then it would drop off to near 0 psi. Same with a digital voltmeter. That blob of corrosion that used to be a pair of terminals could measure thousands or millions of ohms, but as long as there's something there, the voltmeter will still see 5.0 volts. (This is where the cheap test light can be a lot more accurate for many electrical tests. It requires current to be able to flow through the circuit for it to work).

The Engine Computer requires a small amount of current to flow through the signal circuit, so that is when the corroded terminals will show up as a break, and you'll see two different voltages at the sensor and at the computer. All of this can be boiled down to using the scanner to see what the Engine Computer is seeing. Base your test results on that instead of what you found right at the sensor.
Was this
answer
helpful?
Yes
No
Thursday, June 18th, 2020 AT 4:36 PM
Tiny
BKAEMPFE1989
  • MEMBER
  • 36 POSTS
Okay, this is my reference to how I checked my TPS. https://troubleshootmyvehicle.com/jeep/4.0L/tps-multimeter-test-2
I found 5.4 on signal wire with is over the 4.5 it should of gotten. So I think the further testing you mentioned would help me out. I will get out there soon and give it a go and get back ASAP and let you know the verdict. Thank you
Was this
answer
helpful?
Yes
No
Friday, June 19th, 2020 AT 7:48 AM
Tiny
CARADIODOC
  • MECHANIC
  • 33,873 POSTS
The way they had you do the test for the ground circuit is kind of odd. It will work, but you have to add in the variable of battery voltage. A good, fully-charged battery will read 12.6 volts, but if it has been freshly charged, it is going to have some "surface charge"; that's free electrons in the electrolyte that haven't been absorbed into the plates yet. Surface charge can produce a false battery voltage reading that's too high, often as high as 13.0 volts.

Suppose you start with that 13.0 volts, then you find 12.4 on the TPS ground terminal. That means there's actually 0.6 volts, (in reference to ground), on that terminal, which is not acceptable. That circuit should have 0.2 volts.

If you have 12.6 volts on the battery, meaning any surface charge has been removed, and you find 12.4 volts, as before, on the TPS ground terminal, you now have 0.2 volts on that ground circuit, which is perfect. This might sound like I'm picking nits, but in such a low-current circuit like this ground circuit, to have that extra 0.4 volts indicates there's a serious high-resistance problem in it. That is not enough to trigger fault codes for any of the sensors that share that ground circuit, but it will sure cause some weird and elusive driveabilty issues. In particular, the MAP sensor uses that ground circuit. It has the biggest say, (only on Chrysler products), in the fuel metering calculations. A tenth of a volt error can cause hesitations, stumbling, poor fuel mileage, and other miserable problems, all with no fault code to direct you to the cause.

I also didn't like their comment about it being okay to take the readings with the sensor unplugged. Remember my comment about the compressed air line in the shop that is 99 percent blocked? If you start with 100 psi at the compressor, you'll have 100 psi at the end of your hose, as long as you're not trying to run an air tool. As soon as you turn on the tool, almost no air volume can get through, and pressure after that blockage will be near 0 psi. You need to cause air to want to flow by turning on the tool, then, if you take multiple pressure readings, you'll find 100 psi all the way up to the blockage, and 0 psi after the blockage. Those pressure readings are what you'll use to find the location of the blockage, (high resistance to air flow).

Electrical circuits work exactly the same way. Electrical pressure is voltage, and flow, or volume, is current. Please forgive me if this gets too technical, but suppose there is corrosion in a splice that's coming apart in the 5.0-volt feed circuit. That's the blockage equivalent in the air line, and we'll say it's 5,000 ohms. The exact resistance of any throttle position sensor is irrelevant, but a typical value is also 5,000 ohms. With the sensor plugged in, current will flow equally through that corroded splice, the sensor, and the wires. Half of the 5.0 volts will be "dropped", or used up, across the sensor and half will be dropped across the undesirable resistance in the splice. You started with 5.0 volts at the computer, lost 2.5 volts across the corroded splice, so you end up with the other 2.5 volts at the sensor. That means you'll measure 2.5 volts at the 5.0-volt feed terminal at the sensor, and that will instantly tell you there's something wrong in that circuit. If you were to measure the voltage with the sensor unplugged, as they suggested, no current would flow through that circuit, so none of the 5.0 volts would be dropped across the high-resistance splice. That leaves you with the entire 5.0 volts at the sensor, which would falsely tell you the circuit is okay.

By telling you it's okay to take the voltage readings with the sensor unplugged, it is the same as saying you have 100 psi at the end of the air hose, so the pipes have to be okay. Neither test this way takes into account the blockage or high resistance.

As a side note, this is why I often recommend using a test light instead of a digital voltmeter, in most circuits. The voltmeter just takes a pressure reading. The test light requires current to flow through it to work, and it will show up as a dim light when there's too much undesirable resistance in the circuit.

Regardless, in your original post, you were way ahead by testing the signal voltage. As I mentioned, the only way you can get the correct sweep from 0.5 to 4.5 volts that you found is if the 5.0-volt feed and ground circuits are okay. The time to move backward and do the individual tests on the other site is when the signal voltage does not go through the proper sweep, then you have to use those individual tests to figure out if the defect is in the 5.0-volt feed circuit or the ground circuit. Since we're past that, the only thing left is the signal wire running back to the computer. Unless I overlooked something, there's only two things that can happen to cause the TPS fault code to set. There has to be a break in the signal wire, in which case the reading you get at the sensor will not agree with what the computer is seeing, (this is where you need the scanner to see that data), or there is an intermittent dropout in the signal voltage. That glitch only has to last a tiny fraction of a second to be detected by the computer. It could be caused by a speck of dirt getting caught under the wiper / movable contact inside the sensor. That would create an open circuit, then the pull-up resistor would place 5.0 volts on the signal circuit that would be detected as outside the acceptable range of 0.5 to 4.5 volts. That doesn't happen very often. Even less common would be a loose rivet that connects the metal terminal to the carbon strip inside the sensor. If that loose connection was on the ground terminal, that too would send the signal voltage to 5.0 volts momentarily. If the rivet, or the connector terminal for the 5.0-volt supply had an intermittent connection, the loss of the 5.0 volts would send the signal voltage to 0.2 volts, also detected as a defective condition.

Now we're back to observing when the fault code sets. If you erase it, and it comes back instantly, that's the easiest to find because the defect is there all the time. As with the pressure tests on the compressed air pipe, a couple of voltage readings will locate the break in the signal circuit. It's the defects that don't show up for hours or days that are the most frustrating. Often they will show up from wiggling wiring harnesses and putting pressure on connectors. The goal here, if possible, is to do something that makes the defect show up, then be careful to try to keep it in that defective state so you can find it with the voltage readings.
Was this
answer
helpful?
Yes
No
Friday, June 19th, 2020 AT 8:06 PM

Please login or register to post a reply.

Sponsored links