Motor dying on steep incline help

shawn said:
Since we don't know the exact circumstances, you might have needed to be in 4lo, too. Too much hill, not enough gearing.
Click to expand...

I keep coming back to this. My sammi would do the same thing with the 1.3, it would never lug it would just die. Think a vaccum gauge could be used effectively to tell when the jeep is under load in this situation?
 
Infamous1 said:
I keep coming back to this. My sammi would do the same thing with the 1.3, it would never lug it would just die. Think a vaccum gauge could be used effectively to tell when the jeep is under load in this situation?
Click to expand...

I understand what you guys are saying about the gearing but all four tires have been spinning and some time it is sitting on the rev limiter when it starts to die. It really feels like I am in a carbed rig and the angle is too steep to fill the bowl
 
shawn said:
A $10 mechanical meter will tell you all you need to know on the TPS.

If it's digital, it's going to have to be something fancy (like a fluke) to get a polling interval low enough to get reliable data. A cheap chinesium digital won't necessarily see a dead spot on a TPS.
Click to expand...

Although that's true if the digital meter has a crappy sample rate or display refresh rate, analog meters can also have their serious shortcomings. A cheap digital will still have an input impedance that's an order of magnitude (or more) higher than an analog meter. Low input impedance is a big deal if you're looking at signal voltage on a variable-resistance sensor, as you're putting a significantly low resistance in parallel with the potentiometer resistance, which can skew the actual output signal.

So no, the crappy characteristics of an analog meter are not necessarily a substitute for the different crappy characteristics of a cheap digital meter.
 
Last edited:
Fabrik8 said:
So no, the crappy characteristics of an analog meter are not necessarily a substitute for the different crappy characteristics of a cheap digital meter.
Click to expand...

Who gives a shit? All you're looking for is a dead spot or irregular progression. The analog meter will show that, the cheap digital won't.

You're not looking at voltage, either. Measure resistance. It reduces the variables that you're testing for.
 
shawn said:
Who gives a shit?
Click to expand...

You should, if you're recommending to use an analog meter as a troubleshooting tool. It's important to understand what you're doing, and not just blindly do it.


shawn said:
You're not looking at voltage, either. Measure resistance. It reduces the variables that you're testing for.
Click to expand...

Resistance is a good place to start, but will tell you half of what you need if you want to assume that the other half is fine without testing it.
Given that this is an issue at idle and at full throttle, I'd want to know what the voltage is at those min/max throttle positions and whether they are within range or need a position adapt (I don't know if Jeeps can do that) or if the sensor needs to be clocked differently if there are slotted mounting tabs..
If you're going to troubleshoot sensors, you should actually troubleshoot sensors.


Mac5005 said:
Have you done a voltage sweep test on the map sensor?
Click to expand...

For example, there's another case of where the low input impedance of an analog meter may become a problem when trying to interpret sensor health.
 
Last edited:
I'd use a scanner. That will show you what the ecm is seeing. Better yet, it might even tell you why it's stalling. (Obd2 live data, codes & freeze frames.... Even if it's 91-95, the drb ii/iii can do most of that) if you don't have a scanner, I'd backprobe the tps to check it with a dmm set to read dc voltage. The free ones at HFT are accurate enough. Not the best but it'll work. Also O2 could possibly cause it to lean or richen up the fuel trims. Just something to check. You could simply unplug the o2 and see if it makes a difference. Also the map sensor or a cracked map vacuum line. Check grounds and wiring with a wiggle test while running. Also check the relay panel for loose/cooked relays & fuses. No/low insertion force is a bad thing. Same goes for connectors.
 
Fabrik8 said:
you're going to troubleshoot sensors, you should actually troubleshoot sensors.
Click to expand...

The sensor doesn't give a shit how much voltage it's getting, and measuring the voltage doesn't tell you if the sensor is bad. If there's a problem with the ECM or the harness, he's going to spend a bunch of time and money replacing sensors that aren't bad. The ohm meter isolates a variable, the voltmeter measures a pile of variables simultaneously.

When you're troubleshooting something, you want to break the problem down to individual, discreet questions, not introduce additional confusion with bad/misleading data. This isnt the time to overcomplicate things.
 
shawn said:
The sensor doesn't give a shit how much voltage it's getting, and measuring the voltage doesn't tell you if the sensor is bad.
Click to expand...

I was referring to measuring the sensor output voltage, not the sensor supply voltage. But measuring the sensor output voltage does tell you if the sensor is bad, obviously. :confused:

While we're on the topic.... Variable resistance sensors (with sensor power and ground, like a TPS) do care very much about the supply voltage, because they're ratiometric (they're just a variable resistor strapped between power and ground) and therefore the output voltage is always affected by the supply voltage. That's just basic sensor function and can't really be argued. I'm sure an older Jeep doesn't have a Hall effect (non-contacting) TPS that some newer vehicles use, and those work totally different (can't measure resistance).

I don't feel like arguing anymore, my original point was that you don't want the usage of an analog meter to be the cause of that bad/misleading data.
 
Last edited:
Fabrik8 said:
But measuring the sensor output voltage does tell you if the sensor is bad, obviously.
Click to expand...

It doesn't tell you that. At all.

It tells you that the voltage is low, which could be for a dozen different reasons, only one of which might be a bad TPS.
 
Fabrik8 said:
Variable resistance sensors (with sensor power and ground, like a TPS) do care very much about the supply voltage, because they're ratiometric (they're just a variable resistor strapped between power and ground) and therefore the output voltage is always affected by the supply voltage.
Click to expand...

So.... What's the most likely failure mode for a bad TPS, and how would it be most likely to present itself?
 
shawn said:
So.... What's the most likely failure mode for a bad TPS, and how would it be most likely to present itself?
Click to expand...

I was correcting your misinformation about how the sensor works.

This is not politics; you can't have your own interpretation of how you think a sensor functions.

I don't disagree with your troubleshooting, I just think it's incomplete.
 
Last edited:
Fabrik8 said:
I was correcting your misinformation about how the sensor works. Was there something wrong about what I said?
Click to expand...

Lots.

See, here's the thing. The sensor is just two little pieces of metal that rub together. That's it. They rub back and forth, and the resistance between the input side and the sense side changes. The most likely failure mode is for the little stationary bit to wear out right where the slidey bit spends the most time rubbing on it. That spot is at about 1/4 to 1/2 throttle - right where the throttle plates spend the most time at cruise. Somewhere right there in the middle, they start to wear out, and they don't conduct electricity as well as they should. When that happens, the computer gets bad info. Depending on the car, it either thinks that the throttle is suddenly WTFO, or snapped shut. So it gives a bunch of gas, or none at all, then the O2 data rolls in, and it contradicts the TPS, so the computer tries to compensate, but its outside of the trim tables, so it doesn't know WTF to do, and the result is surging or stalling. It'll happen all the time, usually riding down the road at part throttle. It doesn't matter what gear you're in, or whether the tcase is in 4lo. It might happen more when it's raining (humidity), or more after the car has been running for a while (heat), but it happens pretty much all of the time, and it gets progressively worse with time.

While the TPS might still be the problem (who knows at this point), we can already see the most likely failure mode for a TPS doesn't coincide with the described problem. But that doesn't mean we shouldn't cross it off the list conclusively.

Since we might have some defective wiring, or a corroded fuse socket, or a bad ECM, marginal battery, bad ground, etc, etc, etc, we want to isolate the problem. Does THIS part (the TPS) need to be replaced?

The easiest way to find out conclusively is to unplug it, probe the connector with a cheap analog meter, and slowly open and close the throttle. If the needle moves smoothly back and forth, the TPS is good. If it moves correctly, then jumps around wildly (likely somewhere in the middle of the sweep), the TPS is bad. Those two little bits of metal we talked about earlier? They give zero shits what the impedance of your meter is.

The problem with using a voltmeter to troubleshoot the TPS is that it doesn't tell you for sure that the TPS is bad. It tells you that the TPS voltage is low, or too high, or that it changes sometimes for an unknown reason. If you're Mr. supersmart mechanic guy, that might be okay with you. You've been around the block, you know it's garbage in, garbage out, and you've probably already watched the scanner data and seen the same behavior. But if you're trying to describe the troubleshooting process to somebody else who may or may not know anything about how any of this shit works, why confuse them with any of this extraneous shit? Put a meter on the TPS, cycle it, check it off the list, and move on. I mean, fuck, I've had unrelated problems randomly change the voltage on the 5V rail. So you're out there troubleshooting the TPS output voltage, and meanwhile a charcoal canister evap sensor or some stupid shit is shorting out and dropping the 5V to 4.2V. But you don't know that, because you went out and spent $300 on a fancy meter so that you could watch the TPS output voltage, because somebody told you there was no way you could properly troubleshoot two little bits of metal with a $10 meter.
 
shawn said:
See, here's the thing. The sensor is just two little pieces of metal that rub together. That's it. They rub back and forth, and the resistance between the input side and the sense side changes. The most likely failure mode is for the little stationary bit to wear out right where the slidey bit spends the most time rubbing on it. That spot is at about 1/4 to 1/2 throttle - right where the throttle plates spend the most time at cruise. Somewhere right there in the middle, they start to wear out, and they don't conduct electricity as well as they should. When that happens, the computer gets bad info. Depending on the car, it either thinks that the throttle is suddenly WTFO, or snapped shut. So it gives a bunch of gas, or none at all, then the O2 data rolls in, and it contradicts the TPS, so the computer tries to compensate, but its outside of the trim tables, so it doesn't know WTF to do, and the result is surging or stalling. It'll happen all the time, usually riding down the road at part throttle. It doesn't matter what gear you're in, or whether the tcase is in 4lo. It might happen more when it's raining (humidity), or more after the car has been running for a while (heat), but it happens pretty much all of the time, and it gets progressively worse with time.

While the TPS might still be the problem (who knows at this point), we can already see the most likely failure mode for a TPS doesn't coincide with the described problem. But that doesn't mean we shouldn't cross it off the list conclusively.

Since we might have some defective wiring, or a corroded fuse socket, or a bad ECM, marginal battery, bad ground, etc, etc, etc, we want to isolate the problem. Does THIS part (the TPS) need to be replaced?

The easiest way to find out conclusively is to unplug it, probe the connector with a cheap analog meter, and slowly open and close the throttle. If the needle moves smoothly back and forth, the TPS is good. If it moves correctly, then jumps around wildly (likely somewhere in the middle of the sweep), the TPS is bad. Those two little bits of metal we talked about earlier? They give zero shits what the impedance of your meter is.

The problem with using a voltmeter to troubleshoot the TPS is that it doesn't tell you for sure that the TPS is bad. It tells you that the TPS voltage is low, or too high, or that it changes sometimes for an unknown reason. If you're Mr. supersmart mechanic guy, that might be okay with you. You've been around the block, you know it's garbage in, garbage out, and you've probably already watched the scanner data and seen the same behavior. But if you're trying to describe the troubleshooting process to somebody else who may or may not know anything about how any of this shit works, why confuse them with any of this extraneous shit? Put a meter on the TPS, cycle it, check it off the list, and move on. I mean, fuck, I've had unrelated problems randomly change the voltage on the 5V rail. So you're out there troubleshooting the TPS output voltage, and meanwhile a charcoal canister evap sensor or some stupid shit is shorting out and dropping the 5V to 4.2V. But you don't know that, because you went out and spent $300 on a fancy meter so that you could watch the TPS output voltage, because somebody told you there was no way you could properly troubleshoot two little bits of metal with a $10 meter.
Click to expand...

I agree with what you're saying, and with most of the logic behind it.

That's why it's baffling when you start throwing stuff out there about sensors not caring about how much supply voltage they have, and how measuring sensor output voltage doesn't tell you if the sensor is bad. Clearly you know better than that, even if you ultimately think those things should be ignored in order of run-and-gun importance.
 
Last edited:
To be fair: you can check the tps fine with a (d/a)mm by voltage. If you measure supply from ground that's almost max output (it's like .5v or so off either rail so the ecm can detect a short failure) watch for the voltage to sweep smoothly. No need to check resistance on it. Just sweep the throttle slowly enough that the meter can respond. You can do the same thing with a scanner, and it's got a faster sample rate than an average dmm. Yes, you can also check it by resistance. I prefer the scanner method, as you see what the ecm sees. Although you can check resistance on the Jeep tps, you must voltage test the MAP sensor. Use a vacuum pump too actuate it or let it run and compare readings with a mechanical gauge or look for severe non-linearity or discontinuities in a graph on the scanner.. another reason the scanner is better is it can, for analog sensors, stand in for a oscilloscope. I would definitely try a wiggle test on the harness with it running. Including the ignition coil connector.. I've seen many connectors have problems with losing contact tension, especially if the harness isn't secured well or the motor has/had weak or broken mounts.

Good luck narrowing your problem down.
 
Fabrik8 said:
That's why it's baffling when you start throwing stuff out there about sensors not caring about how much supply voltage they have, and how measuring sensor output voltage doesn't tell you if the sensor is bad. Clearly you know better than that, even if you ultimately think those things should be ignored in order of run-and-gun importance.
Click to expand...

I've given you specific examples where reading the sense voltage would incorrectly lead you to believe the sensor was bad. The supply voltage has no correlation to whether or not the sensor is bad, and the output voltage is measuring the voltage on an entire system, of which the sensor itself is only one small part.
 
I haven't read the all replies but have you checked for a short on your battery positive cable or power wire into your PCM? The "only on steep inclines" has me wondering if something changes position and causes a momentary short or strained connection. Maybe your motor and trans mounts are bad and you're getting movement at steep angles.
 
Do you have a local spot to recreate the problem?

Several times I have taken .023” welding wire and back probed sensors while plugged in thru the plug. Typically can get .023” wire between the wire and weathertight seal on the harness side.

I then twist the wire around my volt meter leads and run it into cab. Then I can watch live sensor data. Just get the correct signal wire of each sensor.

Analog gauge illustrates a dead spot better as you notice the gauge fall to zero or peg out.

Digital gauge shows it to, it just briefly shows an anomaly at the dead spot. It’s definitely more difficult to spot than analog gauge but isn’t impossible.

I suggested map sensor as it reads engine load.

If going to 4 low fixes the issue, by reducing load look at what sensors are affected by load changes. Ie map sensor.

Map reads high vacuum at idle,no load, and no vacuum at wot/high load.

I would rule out that the engine is twisting under high load and causing electrical issues directly by position.

Foot brake it on hard surface like a burnout but don’t spin the tires. Have someone watch to see if engine is moving a lot (bad motor mounts) or pulling/smashing the harness against firewall somewhere.

As far as testing the map sensor, you can attach a small hand vacuum pump directly to sensor.

You should see a nice smooth sweep as voltage or resistance changes thru sensor position as described above.

Anything erratic would be a suspect.
 
Err....I am,kinda lost on all the terminology and engineering but I got a fancy scanner that might help.

1537712625151992852964.jpg
 
BigClay said:
You around next Sunday for me to come borrow it? :)
Click to expand...
Unless I'm welding on a CJ. Should be. Definately doin church. Homecoming, coverdish, and bluegrass band. Welcome to attend that as well.......and any other God fearing blue grass lovers!
 
So a little update. I was driving the jeep on the road the other day, and I had to get on it to merge into traffic, I found a dead spot in the throttle, and lost all power for a second. Did a little research and this seemed to be TPS related. Ordered a new TPS and installed it, also cleaned the IAC while I was in there. Jeep now fires up faster than it ever has, doesn't over rev at idle, and the throttle is much more responsive. I am hoping this was the culprit, just have to wait and see now.
 
BigClay said:
So a little update. I was driving the jeep on the road the other day, and I had to get on it to merge into traffic, I found a dead spot in the throttle, and lost all power for a second. Did a little research and this seemed to be TPS related. Ordered a new TPS and installed it, also cleaned the IAC while I was in there. Jeep now fires up faster than it ever has, doesn't over rev at idle, and the throttle is much more responsive. I am hoping this was the culprit, just have to wait and see now.
Click to expand...

Make sure all of the wiring harnesses to those sensors on the intake area aren’t rubbing in any places or chafed.

The 99XJ DD had a dead spot/stutter until I saw the wiring harness was rubbing against the intake. (Thanks to the help of some gurus here) :)


Sent from my iPhone using Tapatalk
 
Back
Top