PO171 97 silhouette strange intermitent problem

[midwestinavan]

I have a 97 silhouette that has a problem that I just can't figure out. When you drive it on a long trip eventually it doesnt want to idle anymore. Sometime it'll go 200 miles before it does it, sometimes 500 but when it does do it it won't idle anymore and my average mpg goes down from 26 on the highway to 13. It still seems to run fine, the idle will just jump all over and die unless you drive two footed. If you let it sit overnight, start it up and it runs great again. I can stop and go all day, but when I start pushing it continuelly eventually it'll always do it. I got a po171 last time (bank 1 too rich or too lean). I changed the fuel filter, fuel pressure regulator and cleaned the mass air flow sensor. Last night I went on a trip with it hopeing it was fine and 250 miles later it started doing it again, only this time I feel a miss as well, and for some reason my fuel gauge didn't want to work right. It continued to read full till I stopped it and shut it off and restarted it then it went down. Anybody got any idea's on this one?

[dewaynep]

I guess my first question would be: do you have access to a scan tool with live data? You need to see what the computer is asking for and what it is getting. No idling could be IAC, EGR valve stuck, MAP, MAF, TPS, plugged Cat., etc... I would want to know what was out of spec or just odd looking before throwing parts at it.

[midwestinavan]

No I don't have access to a live data scan tool. Just genaric obd II . I went and checked it today from my trip last night and I got the po171 again and now a po300 (random mutiple misfire I believe). But once again now that it sat all night it's been running great and probably won't do it again till I drive it 300 miles stright. I just recently got layed off from my job so I'm just lookinh for the most cost effective solution to troubleshooting it, then I can go from there. Thank you for any advice.

[LMP]

Just looking for simple things before, I'd disconnect the injectors one at a time. One sticky injector blocked open might explain a lot....
Of course, if it is blocked open at the time you investigate, it might stay opened ...may be tapping after disconnecting ...or may be just stopping the pulse inputs might release it after some time...
IN any case, each disconnect should lower the RPM. If one does not, it is blocked open...and if that does not reveal anything, at least, it cost nothing...
The fuel gauge problem seems to be unrelated.
.

[midwestinavan]

I'll definatly keep that in mind, it's kind of hard considering it won't probably do it again till I run it down the highway but definatly gives me a direction to go in. What about an ignition problem? Could that be possible to give me a rich/lean code? I'm just stabbing. Oh and the van has 108k on it. I forgot to mention that earlier. We just got it a couple months ago so I'm not sure what's been done on it prior. Since I've gotten it I put in a fuel filter, air filter, fuel pressure regulator and cleaned the mass air flow sensor. Thanks again for the time

[LMP]

...well, what I suggest can explain low fuel mileage, rich burn, high idle, spark plug fouling and misfire....which alltogether would not have another probable simple single cause. MOnitoring fuel rail pressure for rapid pressure loss after stopping engine can tell of blocked injector (not sure where the port is on a 3.4...my 3.8 is easy reach) ...but Waiting for the problem to show before investigation is frustrating though....but if the computer could hint on which plug misfires..like P0303, then I would jump at that and change THE injector to that cylinder...hoping it is one of the front bank... Of course, being there, I'd be tempted to change all three on a bank....then wait and see.
Indirect injection as used here can impair more than one cylinder though (even by only one faulty imjector) , but the plug close to the culprit would surely be more troubled...

[midwestinavan]

ok, well I'm happy that it does tell me it's the front bank, (whatever it is. lol). As far as the idle, it was never high though. Once it starts to do it, it doesn't idle at all. It bobs up and down till it dies witch is usually right away, in fact if I don't give it any gas it'll usually die before I get stopped as I'm getting off the highway. I really wish it would just do it all the time instead of having to drive it hundrads of miles first

[LMP]

Originally Posted by midwestinavan

... As far as the idle, it was never high though. Once it starts to do it, it doesn't idle at all. It bobs up and down till it dies

..with closed throttle and then no air to burn the unwanted fuel...I might indeed expect that..it would be fun to solve the case with such a simple cause. Keep us posted...
There has been a reported case of an injector driver transistor gone dead (in the ECM) and the replacement was possible (in an earlier model)..but as I remember, it (the transistor) was shot open, not shorted (as could also happen, even intermittently with heat)....but this is probably a very remote possibility. The sticky injector is more common.
.I might also be tempted to try generate the trouble , instead of waiting for, that is testing the hypothesis....by keeping one injector opened by applying constant 12V, at idle,...and see what follows.....easy

[midwestinavan]

So are you recomending I just put stright 12V from the battery to the injectors one at a time?

[LMP]

I would apply 12 V to ONE injector, any one, to keep it opened while the engine runs and see if this will replicate symptoms that are already known, like that erratic idle......If the symptoms do replicate, it would confirm the hypothesis that ONE injector...we do not know which...does in fact block opened at times and causes the same conditions.

I would use some means to limit the current as injectors are trigged normally just a fraction of the time....using a series resistor of 4 ohms will limit the current. Using a headlight bulb as a substitute resistor would be fine, even better as the lightbulb would allow a high inrush current making sure the injector opens and as the light gets brighter, it lowers the current to a safe level and this is sufficient to keep it opened.

In a condition like yours, I would try that. I would hope it proves the hypothesis, hence solve the problem; otherwise, I would then look for other options.

[midwestinavan]

ok, well it'll take me a couple days to be able to try it but I will and will let you know what I find. I noticed the injectors plug in facing into the valley of the engine makeing the plugs somewhat difficult to get at. Any ideas on this one? Again thank you for your time, at least now I have some direction to go in.

[LMP]

Originally Posted by midwestinavan

... I noticed the injectors ... plugs somewhat difficult to get at. Any ideas on this one? .

..well.. maybe my plan was easy in theory....I've looked at the pictures in the post about the intake manifold gasket..not easy ...indeed, this is far more difficult than with my 3.8 series 1, where they are exposed right in my face. I might look for the possibility to intercept the electrical wiring to one injector somewhere else, may be right at the PCM plug....
...look at the diagram here ....
www.avigex.ca/xport/97enginewiring3of3_19933.pdf
WHen left connected, all injectors are fed all the time with +12V on one side, and the computer grounds the other end, so you would have to ground the wire, not apply 12V....

...but if one could reach one of the injectors plug and be able to disconnect it, I'd try to find an injector connector from the scrap yard with a decent length of wiring ...connect it , and work from there....

[midwestinavan]

Thank you for that. I'll see what I can figure out. I was also looking at some other things that I don't know much about like the fuel tank pressure sensor? It seems to me that it acts somewhat like vapor lock, but again gives me a bank 1 condition. I"m going to see what I can do to try your idea with testing the injectors first. I went out and looked at my 96 olds 88 with the 3.8 and I see what you mean as far as accessability. I wish these vans would have had the 3.8. What do ya do though? lol

[dewaynep]

Bank 1 is the only bank on this engine. There is only 1 upstream oxygen sensor, so the bank 1 sensor 1 really doesn't tell you anything other than the entire engine is running poorly. The best way to test the injectors is to use a scan tool and a fuel pressure gauge. I would be very careful about injecting battery voltage into a wiring harness or grounding a wiring harness. You can very easily fry the ecm. P0171 is for a lean condition, that is caused by either lack of fuel or too much air. If it runs fine going down the highway but not idling you probaby have sufficient fuel delivery since it is going to take more fuel at highway speeds. I would start looking for an obvious vacuum leak (a vacuum leak will be more noticeable at idle) or a slow IAC valve after it warms up. You can listen to the injectors using a mechanics stethescope and hear them ticking. If one doesn't tick or sounds different than the others that will clue you in to which injector. We don't want you to be chasing your tail here.

[midwestinavan]

What makes it so difficult is that now that I'm home and it sat overnight it's running perfect again. Back up to 21ish in town as always, if I were to go drive it 2 or 3 hundrad miles it'll do it again till it sits overnight again Wouldn't a vaccum leak be present at all times? This is where I'm stuck. However, I did not realize that I only had one upstream o2 sensor, I thought it had one for each bank. I'd definatly like to figure out how to trigger it, or just have something go out. That would make it alot easier. lol But nothing is ever easy right? Thanks again for the help and advice you guys. If you think of anything else please let me know, meanwhile I'm going to continue to research and figure out ways to test different components.