Cranks, but only starts when manually choking the Intake

Recently, my 2001 Chrysler T&C (3.8L) started sputtering and acted like it was losing power while going down the road. I was able to make it home. I turned off the engine and tried to start it again and it would crank, but wouldn’t start. I checked the codes with a scanner and it reads P0171 (lean)/P0137 (O2). I removed the air box cover and blocked the intake hole with my hand and had a friend turn the engine over and it started. Once it’s started it runs pretty good (especially at higher RPMs). Would this be a vacuum leak? Maybe a MAF sensor? Thanks.

Could be either. Check for obvious vacuum leak sources, like split hoses or hoses that fell off. A split in the boot between the air cleaner and the MAF could cause this too. Next use a hand held vacuum pump to check vacuum operated devices, especially the brake power booster, for vacuum leaks. You might try just disconnecting the MAF as an experiment. If that makes the engine start or run better, it probably needs to be cleaned or replaced. The p0137 is for a problem with the post-cat o2 sensor, and could be related to an MAF problem, or could be a vacuum or exhaust leak, or a failing cat. When you did that experiment partially obstructing the air intake to get it to start, did it then run fine even at higher speeds and during rapid accelerations? If so, unlikely to be a plugged cat.

Yes, it runs well during rapid acceleration. It also idles fine. I’ve been able to drive it, once started. It is a little sluggish when first gaining speed though.

Unlikely to be a plugged cat. Sounds more like some kind of air leak is causing it to run a little lean. That would show up more at idle and slow speed driving than faster speeds and rapid accelerations. One idea, spray a little starter fluid in the air intake area, intake manifold, etc, at idle. If the engine speeds up, there must be an air leak.

I also have the follow blinking codes:


I tried the starting fluid, but didn’t really notice a big change. Maybe have been slight, but hard to tell.

EVA means “electronic voice alert” right? That seems unlikely to be involved w/the engine running properly. I’m not sure what you mean by “blinking codes”. If you mean the check engine light is blinking, that usually indicates a severe misfire. Which could be caused by running too lean. pcv system and egr system problem can cause lean operation among other problems.

Sorry, these codes are blinking at the top of my diagnostic tool. The full O2 code reads "O2 Sensor 1/2 Circuit Low. My vehicle will also start with a couple shots of ether.

What’s the EGR “p” code number? Does it say?

EVA is EVAP System Monitor. Only 2 “p” numbers P0171/P0137. The other codes are blinking red circles at the top of my INNOVA 3100j reader.

I don’t know how to interpret the blinks on your code reader. But the two p codes you mention and the symptoms are consistent with too-lean operation. If you enter p0171 Chrysler into google, you’ll see a link to obdii website, which will give you a list of possibilities why the engine is running lean. Basically it is either too much air getting in, or not enough fuel. Is your code reader able to output fuel trim data? If so, that might be helpful.

Your fuel pump is failing, measure the fuel pressure while driving.
A vacuum leak would cause the engine to idle faster, not poor acceleration.
There is no mass air flow sensor on your Chrysler.

Those are monitors, your device may show that they have passed, have not passed or in process.

The blinking codes are tripped system monitor codes. Blinking red means a trip and solid green is A-OK

Freeze Frame data for the P0171 code

Fuel Sys 1: CL
Fuel Sys 2: N/A
Calc Load: 7.1%
ECT: 194 F
STFT B1: 24.2%
LTFT B1: 24.2%
MAP: 10 inHg
Eng RPM: 800
Speed: 0 mph

The fuel trims are showing very lean operation. LTFT 24.2% means the computer is having to inject that much more fuel than it thinks it should based on the MAF in order to satisfy the O2 sensors. One interesting point, 10 in Hg is pretty poor vacuum at idle. My cars tend to be in the 17-20 inch Hg at idle. The poor vacuum reading at idle leads toward a vacuum leak of some kind. Or a compression problem. Presuming this data corresponds to warm idle. Nevada’s idea above about the fuel pressure is worth considering too. But that wouldn’t explain the poor intake manifold vacuum.

I’m not having any acceleration problems once started. My sluggishness is when running a constant speed at lower RPMs. You are right about the MAF. I was mistakingly referring to the Intake Temp Monitor as the MAF.

If the intake manifold vacuum was really 20 in, but the MAP said 10, this symptom could result. Due to a faulty MAP. Faulty MAP’s as reported here is pretty rare, but happens occasionally I expect.

Guy at the local shop did a fuel pressure test and said it read 20psi at idle and when he accelerated. So he ordered a pump and filter. If that is the case though, why would the vehicle not have any trouble accelerating. Is the motor just compensating? Perhaps that’s why I get the high 24% Long Trim?

Low fuel pressure will definitely cause a positive fuel trim. That’s b/c the computer presumes each time it pulses the injector, that will result in a certain (known) amount of gas injected. The amount injected depends on the fuel pressure and the injector spec (in pounds of fuel per hour). The computer knows both of these. But if the fuel pressure is actually below the rated fuel pressure (and 20 pounds is indeed low), the computer pulses the injector ok but not as much gas gets squirted out as the computer assumes. Then the o2 sensor complains the mixture is too lean (not enough gas in this case, rather than too much air), so the computer increases the injector pulsing time until the O2 sensor stops complaining. That extra pulsing time corresponds to the 25 % fuel trim number.

You pose a good question, why doesn’t this result in poor acceleration? One theory, even with low fuel pressure , as long as the fuel demand isn’t too great, the computer can still compensate for the low fuel pressure by increasing the injector pulse time. At some point it could no longer compensate, but that might occur only at really high acceleration or really fast freeway driving. You might try an experiment, if it is safe to do, see what happens when you really step on the gas while going up a steep hill at freeway speeds.

In any event, with low fuel pressure, the computer is having to meter the fuel according to the O2 sensor, rather than the usual way using the throttle position and map sensors, that would affect drivability as the o2 sensor method doesn’t allow a quick dynamic adjustment; for example when you step on the gas pedal it might stall out before it figures out what to do. Starting the engine , likewise might be difficult.

Hopefully fixing the fuel pressure will solve all the problems. If not, make sure your shop is aware of the MAP measurement I mentioned above. Best of luck.

I think you have your answer but what does the fuel pressure say? The problem with trouble codes is that they can report symptoms rather than the cause of the symptoms.