Note: Allpar does not take responsibility for the veracity of any information or opinions here, does not claim expertise, and is not responsible for any consequences. Please proceed at your own risk.
By Allen Middleton
Also see this general driveability troubleshooting guide
I have a 1985 Dodge Omni GLH Turbo, and last weekend decided to spend my annual ritual of eliminating driveability problems. The most irritating was the engine cutting out under hard acceleration, virtually breaking motor mounts. The second problem, I've been losing acceleration, about 10%, though I generally keep the car tuned.
The first problem, the engine cutting out syndrome, baffled me. I only could think of the worst, severe engine wear, since there is 156K on the motor. Specifically, the engine tended to buck when accelerating after entering the freeway. One of the systems I lacked knowledge, was the barometric reed solenoid. Most the information I came across just said that the computer read the atmospheric pressure occasionally. I read all technical info. I could at the library, and finally came across a description in the Chrysler shop manual. The point that struck me is that the barametric reed solenoid is in line with the MAP sensor. I thought there was another sensor somewhere I didn't know about. It goes onto say that the computer reads the atmospheric pressure at a specified rpm (no doubt low rpm, when there is a vacuum in the intake manifold, so why not read the atmospheric pressure for added information that might allow for tailoring fuel/air mixture and turbo boost levels), when the accelerator is closed (I guess not depressed), and no more than every 30 seconds. Another symptom of the engine bucking, is that it only occured after I started accelerating again, it would occur about that same time, same speed, and same frequency (it would happen twice), then go away.
Now I was onto something. A few years back, I had a severe cold start up problem. First, it would hardly start, then the motor would idle at 3K rpm, for a few minutes. Then, things returned to normal. I took it to the dealer, they did work recommended in a TSB, they added the external MAP sensor kit. I found out the next cold snap (-15) that didn't fix the problem. Brought it back, they finally found a crack in a vacuum line, which sucked in water, which froze...
Anyway, after learning about the reed solenoid function, I immediately inspected this MAP sensor circuit. Low and behold, with my new knowledge, I immediately noticed that the MAP sensor was hooked to the solenoid, and the other ends connected back to the computer (expected) and to another vacuum line (not expected). I would expect the third port of the solenoid to be unconnected, so when the solenoid is grounded, the MAP sensor would read atmospheric pressure. Instead, it was sucking on this mysterous vacuum line. It turns out it went only about 1ft, and was capped. When I got home, I tested it for holding vacuum, and it did. The bucking hasn't returned.
I still don't understanding the bucking phenominon, was the turbo over-boosting? I didn't get any weird error messages.
Well I still couldn't beat an old 0-60 benchmark of 8 sec. Something was still wrong. I re-checked compression, 115 lb+. No vacuum leaks, EGR works, new cap/roter/plugs/air filter/fuel filter, un- plugged cat/exhaust (had this tested - btw, the original exhaust, MN salt has not eroded the stainless yet!). The only error code I'm aware of is some fault with the EGR solenoid (I had previously looked at this, I checked everything, vacuum lines, new back pressure transducer - another story, but concluded the computer is confused). So I researched where the heck the charge air sensor was. The theory being that I knew the plugs looked light grey/brown (perfect according to charts, but when the car was new, the plugs would always look black/sooting, from a rich mixture). One would assume if the computer thought the air was warmer than it really was, that it would lean the fuel/air mixture some, and perhaps alter the turbo boost curve.
On the GLH, it is located on top of the intake manifold, right next to the detonation sensor. It is a "2-way" sensor that screws in. I belive it is identical to the coolant sensor. I unscrewed it, and it was filthy. It also was bubbling on top, looked as though it was melting at some point. Lots of heat sneaks up that path, since the exhaust manifold is right below it.
I found Napa sells one for $17, is called out for all other c-cars other than Omni's. I bought one. This looked like a better sensor - it was mostly brass. The stock unit was plastic. I had trouble screwing the new one in, I don't think it got mis-threaded, I'm not leaking fuel anyway.
Net result - seems like boost comes on quicker, and I have more acceleration. I suspect I'm back to where I was several years ago, will have to verify when I get the chance. I also noticed that when the engine is cold, rpm's stay up longer. Another driveability problem went away - the tendancy to stall after going out of gear while still moving.
Conclusion - another example that some Dodge service technicians can't read pictures, and sensors can change calibration enough to affect driveability/performace, yet not trigger an error on the diagnostic computer (read this soft-head auto firmware engineers). Since my computer is no longer capable of driving the power loss lamp (does anyone have a clue how the computer can tell that it can no longer drive the lamp? It can't, but the shop who tested the cat/muffler restriction diagnostic tool indicated the power loss circuit failure...) I can't verify if the charge sensor fixed the EGR message, it is possible somehow the computer can infer the EGR valve operation based on the charge temperature. Instead of shot-gunning the computer malfunction (no doubt a transistor/driver somewhere), I may break down and purchase the Mopar modified computer, for $159, which if I recall, dropped about $100 in price since I looked last.
Also see these driveability tips
We strive for accuracy but we are not necessarily experts or authorities on the subject. Neither the author nor Allpar.com / Allpar, LLC may be held responsible for the use of the information or advice, implied or otherwise, on this site. This page is offered “AS IS” and without warranties. By reading further, you release the author and Allpar, LLC from any liability.
Maserati LevanteThe luxury crossover finally appears
Don Prudhomme: talking with “the Snake”Drag racing, funny cars, and the Snake and Mongoo$e movie
This page is in-image-ad-free, 50% of the time. Support Allpar by using our Amazon link
All Mopar Car and Truck News
2018 Jeep Compass
2007-10 Jeep Wranglers
2016 Allpar show-meet
41 years in Chrysler Engineering