'98 TJ Bucking Bronco

I would be looking at cruddy connectors, myself, but when I first got mine it was doing this and it took the whole arsenal and several months to get a decent idle. Cleaning the TPS and throttle body made the biggest difference - especially the port for the Idle Air Control

- but it took a new O2 sensor to get the last little dip out. Mine is the older version MFI and it still wants to die when I hit the brake just after starting and easing to the bottom of the hill. Now code retention for the old beast but I'd go with the info you have and check the connectors for the TPS.

Slightly different, but this may be a clue for some others: it now appears that mine dies WHEN I HIT THE BRAKE PEDAL. That points to a vacuum leak but it cures itself after the first time so it would almost have to be a sticky check valve in there. Anyone know of a check valve in the brake booster that can stick allowing a large vavuum leak until it frees itself?

Will Honea did pass the time by typing:

No check valve that I'm aware of. You could block the hose off and test to see if it's electrical.

Have you stuck a vac gauge on the intake rail and watched it as you drive/brake?

One of these comes in real handy for diagnosing vac problems.

I haven't done anything - yet. I just noticed the brake/stumble correlation this weekend and after learning to live with that "it's gonna die at the first stop sign when the weather is just right" for close to 60K miles it's not a priority. I remember chasing a couple of Chrysler problems where the engine would die if you eased down on the brake pedal and held it in just the right spot. That was a leak in the brake booster but it was repeatable (once you figured out WTF it was doing!) and it did not go away like this one does. GM's of the

80's were also bad about that leak but this one has real character so I'll get around to it one of these days...

Will, I kinda doubt a vacuum leak would be to blame for this, unless it was an instant HUGE leak, but that would show up at other times as well. Normally on a speed density system, a vacuum leak will just cause a higher idle.

One thing I'd check would be the EGR/EVAP solenoid. If that thing doesn't shut off right it'll cause problems, especially cold. Should be energized (vacuum off) at idle when hot, and at all times when cold. If it's unplugged, vacuum is always on.

You say it happens cold, at the bottom of a hill? Is it a manual trans?

It's a manual. I hadn't thought about the EGR - thanks. I did have to do a complete soak to get that sucker clean when I first got this truck but I haven't looked lately. When I got it the CCV was plugged up and the whole intake system looked like it had been pickled in crud

- finally took a gallon of Carter's and a big ultrasonic tank to get it all clean. One of the unexpected side effects of blowby and a plugged CCV. The EGR was all carboned up as well due to a really weak O2 letting it idle rich but all of those things finally came clean.

Like I said, after 3 years and 60K miles with the th> Will,

That's probably the fix right there. Going downhill using the engine to slow you down, the computer sees closed throttle but still a higher than normal idle speed. That'll make the computer shove the IAC out all the way to try to get the idle where it thinks it belongs. The computer has no idea the vehicle is moving or that the clutch is engaged, so it doesn't know any better. Then you disengage the clutch and the idle drops too much. Being that it's also cold just adds to the problem, and it can't recover in time to keep from stalling.

Today's computers are a little smarter and have more info to work with. :-)

Well, I don't know if this will help you any; I'm feeling lazy and not reading through all the responses you've already gotten to this. Anyway...

I had similar problems, only I was getting two check engine fault codes. One said the MAP sensor reading was bad, the other said the TPS and MAP reading were inconsistent with each other.

So, I replaced the MAP sensor. Problem solved...for half a day! Engine light came on--MAP code was gone, but the MAP/TPS disagreement reading was still there. So, I replaced the TPS. Problem solved--for good. That was 3 or 4 months ago, and I haven't had a problem since, no matter the trip distance or the alititude (done some high pass trails in Ouray since then, too!).

So, you might check the MAP sensor, too. I think I saw you mentioning some issues that you thought sounded like vacuum pressure, too--would the manifold absolute pressure sensor be applicable to that sort of diagnosis? (I'm asking, I don't know).

Good luck.

/Bob

I've spent the last 40 years in and out of control system design - usually a lot more complex than a consumer fuel system - and pretty much decided that was the situation. That old Bendix computer is early 80's technology (damn, I hope I never have to replace it!), it's still in the open loop mode, and the moving parts are moving as slow as I do when cold. I figure that a lot of the responses are put out-of-band if conditions are even slightly abnormal for that old, slow, processor. I think they used the Motorola 6800 in there - speed demon it ain't.

I got hold of the pinouts and signal descriptions at the computer plug a couple of years ago to build a breakout plug. I was (am) going to profile all the signals with my laptop just to see if I could clone the control system but got sidetracked before I got around to mashing the disk full of data.

Yup. Crank position sensor. Once I found it and removed it, I discovered that the wiring was badly burned at the mid point. I'm guessing that when the guy who put on a header replacing the bad exhaust manifold a couple years ago didn't bother about where the wiring from the connector to the sensor was routed.

Mark '98 TJ '98 XJ

M. E. Bye did pass the time by typing:

Owtch.

You might be able to recover the sensor by splicing new wire but that one is a bit tricky. Hopefully it didn't harm the computer.

About replacing that computer. I was shocked to find out we had one in stock the other day. From the looks of the box though, I think it's just been sitting around forever. I don't know if they're still available from DC or not.

I remember reading somewhere in an AMC manual once what processor they used. Damned if I can remember what it said though. For what I do, it's not really important. The box is either broke or it's not. :-)

You are correct about out of band conditions. They can do some strange stuff if things don't look right to it.

I have a 98 TJ and I'm getting the same thing. check engine light and bucking when taking off from first gear. It happens when it rains. Moisture has something to do with it.

My mechanic suggested changing the spark plug wires. Any comments?

I know I'm a newbie to theis NG, but I think I'm a fair shadetree mechanic. In my opinion the best way to check plug wires is to takle the vehicle to a dark area or garage and with the engine running to spray the wires down with water bottle. If you see sparks replace the wires.

Hope the low tech helps,

Eric Shade