Welcome to the DoItYourself Forums!

To post questions, help other DIYers and reduce advertising (like the one on your left), join our DIY community. It's free!

Trouble Tracker


Evil_Knievel's Avatar
Visiting Guest

Posts: n/a

02-07-04, 07:49 PM   #1  
Evil_Knievel
Trouble Tracker

Hey, I have a really irritating problem with my 1993 Geo Tracker (standard, 4WD, TBI) and it would be great if you guys could point me in a direction here because I've pretty much exhausted every other resource. Took it to a local mechanic, he had no idea. Then I took it to the Suzuki dealer in town, they made it 10 times worse (changed the TPS setting) but when I cranked the TPS all the way clockwise the car would at least have enough power to move again. Anyway, my problem started a few months ago around mid fall when I first noticed brief hesitation under hard acceleration. It only happened once in a while though. There was also slight ping under hard acceleration every so often. This was all annoying, but tolerable. Over the last couple months the slight hesitation then progressed into moderate hesitation, then intermittent complete loss of power (usually around sharp left turns???), then it would idle rough, that progressed into not idling at all. About a two weeks ago it was at the point where it would start, but I had to put my foot half way to the floor (any further and it would stall) and feather the clutch just to be able to move at all. Second gear wasn't even an option because it didn't have enough power to move. Funny thing is though, it runs PERFECTLY when in open loop mode when you first start the engine. You can mat it and it'll pick up just like when I bought the thing. You can feel the second it warms up and goes into closed loop. Just like somebody simultanuously pulled off two of the spark plug boots. Then you can shut it off, then start it again and it'll go briefly then crap out. If you start the car and let it idle from a cold engine, it will idle forever. If you drive 100 feet then stop it will idle for about 20 seconds, then start to sputter (at which point maximum engine output has been decreased to about 5 horses) then stall 30 seconds later. There's big time surging as well if you try to give it any gas to keep it from stalling if you've left it idle. I checked the spark plugs after it has stalled and they look beautiful. Anyway, it's to the point where it's not drivable now and is sitting at the mechanic's place but he has no idea so I figured I'd ask the online gurus for advice.

I know you're supposed to treat any work previously done as suspect, but here's a list of what has been done or checked. As to the thing you've probably got singled out in your mind as the problem right now, the TPS is fine. The mechanic tested it and so did I. Everything is dead centre according to the tolerances given in my Haynes manual. The resistance increases smoothly and linearly with no hickups at all, and that's with pretty extensive testing (I work on electronics for a living). The EGR has been tested, then doubted and completely disconnected. Fuel pressure is good. Valve timing is dead on. Spark appears to be strong. I pulled the computer and tested as many componants as I could, they're all good. But that's not to say one of the ICs isn't bad because I can't test those. Coolant sensor is good. The distributor has been twisted in every direction with no difference. New plugs, wires, distributor. IAC gasget was blown then replaced, no difference. MAP sensor was replaced. The exhaust was removed by the dealer and it's not blocked. ECU is giving the all's well... I tried grounding just about everything under the hood to the battery with a set of jumper cables, nothing. I think it's the ECU but my mechanic and some guy I know that eats sleeps and breathes these Trackers both swear that it can't be the ECU. So basically if you had any new directions to point me in that would be excellent and GREATLY appeciated. That is, unless you would happen to agree with my mechanic; "Got any dynamite?" Anyway, thanks in advance and I apoligize for the long winded e-mail.

Joel Sullivan
[email protected]

 
Sponsored Links
mike from nj's Avatar
Visiting Guest

Posts: n/a

02-08-04, 12:02 AM   #2  
mike from nj
having never worked on one,

nowhere did you mention anything about the oxygen sensor, at the least you could backprobe it with a push-pin(sewing style) and hook the digital voltmeter to that, run the wires inside the vehicle and watch the voltage as you drive. i've done it before. it will tell you whether the exhaust is rich or lean when the problem is occuring and narrow your search down just a little.

second, i got burned (not literally) by a GM tps last week, and i couldn't find it bad by a simple voltmeter backprobe either. it came down to switching it with a known good one, and the problem went away, switch it back, and the problem came back, yet it always showed good on the dvom.
the ideal test for a tps, is with an oscilloscope, but you knew this already.

and the fuel pressure needs to be checked 'while' the problem is happening, not just at idle.

i'm guessing oxygen sensor, based on the open loop/closed loop difference.

did you ever unplug the battery, to erase all the (incorrectly) learned values? (long term adaptive)

which brand name is on the ecu?

this should generate a good discussion, and let us know whatever fixes it.

 
toyotaman11769's Avatar
Member

Join Date: Jun 2002
Posts: 1,531

02-08-04, 04:59 AM   #3  
I agree with Mike, when testing suspect parts or pressure readings it has to be done when the problem occurs. Look into the O2 sensors readings.

I would also back probe the coolant temp sensor when the problem happens.

 
Evil_Knievel's Avatar
Visiting Guest

Posts: n/a

02-08-04, 06:51 AM   #4  
Evil_Knievel
Sorry guys. I forgot to mention that I already replaced the O2 sensor as well. I haven't checked the output to determine if it's too rich or too lean when it stalls though. That's a good idea. I used to have a set of 10 LEDs on the dash of my Nissan pickup that output the state of the O2 sensor in real time. It was really handy in diagnosing problems. I should build the same circuit again and watch the mixture as I'm driving. Good idea.

As for the TPS, no I didn't know that. The type of TPS I have is one that changes it's resistance based on throttle position. I think the one you were testing was one of the type that changes it's frequency based on throttle position. I believe my Nissan used to have that as well.

I have no idea what the dealer did when they checked the fuel pressure though. They were working on the thing for the whole day though and I made sure to ask them several times if the fuel pump was good. They said they had ruled it out. But you're right, the fuel pressure still has been on my mind ever since it got back from the dealer. As you've said, they probably just checked it at idle.

I did try to unplug the computer a couple of times. No go. I also tried unplugging just about every sensor that I could find to see if the computer would substitute expected values for possible bad ones. Tried the MAP, TPS, O2, uh... and something else that I forget right now.

The ECU is a Mitsubishi. I heard terrible things about these computers. I was reading on the net there about these things and one page went so far as saying that if you owned a Tracker/Sidekick your computer WILL fail.

 
carguyinva's Avatar
Member

Join Date: Nov 2003
Posts: 1,147
VA

02-10-04, 09:32 AM   #5  
just a few thoughts...

...mike is right about testing that TP with a labscope...preferably a DSO. I'm sure you can find example patterns of good/bad ones online...if not, let me know here and I'll send you something.

as for fuel pressure testing, i saw nothing about replacing the fuel filter in your post...fuel pressure and volume are very important.

as for the 02 test when the problem occurs, this is a great way to determine the condition inside the engine (rich/lean) and will help point you in the right direction. for example...you would expect that given the symptoms that the engine is starving for fuel and if this is the case and you see consistent low voltage on the 02 (less than 100Mv), get a small can of propane like for a portable grill...run a long hose out the window and under the hood to the air induction system and when the problem is occuring, give it some propane and see if things improve. (you'll need something to regulate the propane with...a soldering tip from the hardware store works well...just cut the end off and screw it on the tank and put the hose over the cut off end) if things DO improve then you can focus on fuel related items. if you want to see if the computer is trying to compensate...use a labscope connected to an injector drive circuit...get the pattern stabilized and then force the computer to make changes in the pulse width (on time of the injector) ie...an air (vacuum) leak should cause the PW to lengthen...adding propane should cause it to shorten....if these things don't happen, the computer may well be just about finished.

one last thought...computers need excellent grounds to work correctly...take your DVOM and do some voltage drop testing at the computer on the computer grounds and sensor grounds. asian cars are notorius for driveability probs associated with ground resistance. definately do this if you find that the computer isn't controlling the injectors properly...asian computers can be very expensive and you want to be sure.

 
dirty dan's Avatar
Member

Join Date: Mar 2002
Posts: 438
LA

02-10-04, 10:49 AM   #6  
What will the engine do if warm and in open loop?

 
Desi501's Avatar
Member

Join Date: Oct 2003
Posts: 2,207

02-10-04, 12:22 PM   #7  
Are you looking at any of the data while this is occuring? Are you comparing the PID values with what they should be. Is you 5v reference staying at 5v. The fact that it goes sour in closed loop points to the idea that one or more of the sensor values is wrong. I had one once that had a shorted low coolant sensor that happened to share 5v reference with the other sensors and in turn robbed half of their reference voltage, changing return signals from everything using that reference signal. I had another GM that had no codes but had a shorting problem with the "signal return" wire from the MAP and acted the same way. Make sure all the PID data is normal. Look at your adaptives and see what direction they are traveling when this happens. Can you force a code by unplugging a sensor? Is the computer getting warm? Is the O/2 cross counting normally? Does the baro match the MAP with engine cold, key on? Do other temps seem correct?

 
dirty dan's Avatar
Member

Join Date: Mar 2002
Posts: 438
LA

02-10-04, 12:35 PM   #8  
"I had one once that had a shorted low coolant sensor that happened to share 5v reference with the other sensors and in turn robbed half of their reference voltage, changing return signals from everything using that reference signal."


If engine running (closed loop) with the problems active, wouldn't the affected (sensors) variation show up on the DVOM or scan tool?
On the GM you reference, did the scan tool show a bad MAP voltage? I'm fighting a similar problem on '89 5.7. No one can find problem, but runs well in open loop. All sensor values are showing up good (factory manual) on scanner.

 
mike from nj's Avatar
Visiting Guest

Posts: n/a

02-10-04, 11:41 PM   #9  
mike from nj
i'll second carguy's answer, i've seen quite a few mitsubishi(and hitachis too) modules come through the shop, never replaced one. i've spent hours and even days(embarrassed to say) hunting down weird things happening, only to find a ground corroded. it wasn't even enough to see. i went through the manual, sanded every ground to bare metal in the wiring diagram related to electronics, and the problem mysteriously went away.

this was with a scan tool hooked up and reading, it never showed a glitch.

yes, an oscilloscope is used to read even an 'analog' tps(resistance style). the glitch is so quick, it won't show up on a conventional dvom, but the computer can see it plain as day.

 
billys68ss's Avatar
Member

Join Date: May 2003
Posts: 1,459
WV

02-11-04, 04:44 AM   #10  
DIfferent car, same symptoms....... I finished a 93 Toyota Corolla yesterday that was acting exactly how you are describing. I did not know the history of the car so I had to start from scratch. I first tested the TPS and replaced it. No Change! The I tested the fuel pressure and found a leaking fuel pressure regulator and replaced it. No Change! Purely by accident I found that the gas tank was somehow full of water, so I flushed the little bit of gass/water out, put in an additive to get rid of the water with fresh premium gas. The problem seemed to get better. I drove it ans suddenly it came back again. So I decide to investigae and test the ignition components. The plugs werent too bad, the cap and rotor had some pretty bad carbon tracking, but I didnt think they would cause it to do this. Then I noticed a crack in the clear insulator along the top on the ignition coil with carbon traces as if it were arcing out. I replaced the coil, wires, plugs, cap and rotor and the problem was solved. The car ran like a new one.
Sometimes you have to replace things that are bad just to rule them out. I dont like doing it like this, but if its bad it cant be helping the situation anyway so might as well replace it.
Billy

 
dirty dan's Avatar
Member

Join Date: Mar 2002
Posts: 438
LA

02-11-04, 05:14 AM   #11  
Desi,
On '89 5.7 GM
Block Learn running around 120.
Checked for exhaust with vacuum gauge and it shows OK. Checked pressure differential before/after cat (put hole in tubes) shows ok.
No backfires.
Fuel Pressure @ 13. (GM reccomends 9-13)

 
carguyinva's Avatar
Member

Join Date: Nov 2003
Posts: 1,147
VA

02-11-04, 07:29 AM   #12  
methinks...

...you guys have started another discussion in the middle of an existing thread...kind of confusing, don't ya think?

 
dirty dan's Avatar
Member

Join Date: Mar 2002
Posts: 438
LA

02-11-04, 08:23 AM   #13  
True... Sorry about that.

Desi,
My readings were posted on the old thread "TBI Worn". I just added a reply to it so it will be one of the more recent.

 
Desi501's Avatar
Member

Join Date: Oct 2003
Posts: 2,207

02-11-04, 12:25 PM   #14  
Your right, I deleted the ones about the other vehicle.

 
Evil_Knievel's Avatar
Visiting Guest

Posts: n/a

02-11-04, 03:38 PM   #15  
Evil_Knievel
Distributor cap, rotor, plugs and wires have all been changed. My mechanic is ordering in a bunch of sensors (probably won't work) this week though. We'll see. Thanks for all the input by the way you guys. It's really appreciated.

 
carguyinva's Avatar
Member

Join Date: Nov 2003
Posts: 1,147
VA

02-12-04, 04:53 AM   #16  
sounds like...

....your mechanic is throwing parts at it and potentially wasting your money. rule out the basics first...

 
Evil_Knievel's Avatar
Visiting Guest

Posts: n/a

02-12-04, 05:05 AM   #17  
Evil_Knievel
I would tend to agree with you on that one.

 
Search this Thread