View Full Version : fires up, but will not stay running

07-27-2014, 05:38 AM
I have a 94 Celica GT with the 2.2L 5sfe.
for a couple of months I have been nursing a bad head-gasket, I'm having a rebuilt motor installed in September.

the head gasket leaks exhaust into the coolant system and blows coolant back out the overflow tube, for a long time this has been the only leak and lately I drive real cool, shifting at 2200 and setting the cruise control at 55 and it does just fine without overheating or loosing much coolant. a couple of months back I flushed the coolant system just to see how dirty it was, and it wasn't very dirty at all yet after I flushed it the headgasket leak got worse from this I decided it may have been stop-leaked before I bought it (just last 0ctober) and therefore it is a better idea to get a whole new motor instead of doing a headgasket on this one. I have another guy doing a full rebuild for me it just isn't going to happen until September.

for the most part it has still run just fine and been quite dependable. except that the idle is often up around 1500-1800, until you goose the throttle a bit and it will come down to 800 or so like it ought to. and lately if I top off the radiator it doesn't want to start on the first try in the morning (most likely coolant leaking into the cylinder). but after it warms up it runs fine, sometimes it might die when its cold but always fires right back up.
then, I started messing with the throttle body to try and fix the idle speed problem. I took the throttle body off, and cleaned the Idle speed control valve, the coil ohms out like it should, and the valve turns freely. and I calibrated the TPS with an ohmmeter and feeler gauge as it shows in the 5sfe manual I have. but I think I set the closed position of the throttle plate too low, I don't have an angle gauge to measure the 6 degrees position from all the way closed, I noticed that the screw was out so far that the throttle opener diaphragm would only move the throttle about 1/8" when it was all the way out. but it ran fine for several weeks it just that when it was idling you could hardly hear it running.

but for weeks that worked fine and I was too busy to take it back apart and mess with it, but once in a while when it was cold first thing in the morning it might die the first couple of times. and the idle was still too high, and it was best to Let it idle high because as soon as I goose the throttle and bring it back down to idle speed it might die. but it would always fire right back up with no problem. but it was slowly getting worse. and it started giving me CEL codes, 12 - distributor, ignition coil, starter circuit. 14 - Distributor coil, igniter. and 21 - O2 sensor.

So this week summer class ended and ive got three weeks off and I started to mess with it... I went to pick-n-pull and got an igniter, O2 sensor and a fairly new looking ignition coil from a Camry 5sfe. at first I had the sense to change things one at a time and tried the igniter and it still wanted to die the first couple of times when cold, but it would start right back up. then I tried the O2 sensor, and mine has had the plug rigged this whole time and was held together by black tape so I took that apart to redo the whole wiring connection( hopefully improving it) installing the plug connectors I got from the salvage yard. But at this point I also sprayed intake cleaner down the throttle body (B-12 chemtool).

after this part it didn't want to start at all, it will crank and fire up, but shuts right back off. I had though at first I just needed to burn off the throttle cleaner but after starting it several times it just won't run anymore. and It fires up almost instantly but then seems to cut out as if something is cutting it off. and if I hit the key again it will not fire, but if I turn the key to OFF and then try again it will at least fire up for a moment and then cut off again. after trying dozens of times and running the battery down and jump starting it I have got it to fire up and run 3 times and I revved it and seemed like it was going to run and as soon as I let off the throttle it dies, and does not want to start back up.

today I took the dizzy off and swapped in the ignition coil I picked up, the ohm readings from the dist connections were within spec( G- G+ 200 ish ohms, Ne- Ne+ 440 ohms) but my ignition coil was the original 20 year old one and had a big split in the plastic and instead of reading .5 or so ohms it fluctuated wildly from 10, to 5 to 15 ohms before finally settling down to something close to 3 or 4 ohms so I replaced the coil thinking for sure I'd found the problem. I set the crank to 0 degrees and lined up the tabs on the dizzy and slipped it in. and it still does the same thing, after trying several times I got it to fire up and as long as I kept it reved up it was ok, and as soon as I let off the throttle it died and will not fire right back up. again, if I hit the key it fires but cuts off immediately, and if I just hit the key again it cranks but wont catch, if I turn the key to OFF and then try again it fires up and cuts off. about every 6 tries it will start and as long as I keep it revved up it will run.

so what do you think? what should I look at next? really the whole thing got worse as soon as I started dicking with the O2 sensor connection. how picky is that thing? I mean the guy I bought it from had it just twisted together and held with black tape this whole 10 months ive had it. or did I just spray too much throttle cleaner down it and need to burn it off?

thanks for your help

07-27-2014, 06:03 PM
Have you run a compression test on your cylinders? I just ask because I find it odd that a leaky head gasket would put exhaust gas into the coolant. I mean it is possible, but a more typical symptom of a leaky head gasket is oil and coolant mixing. And if the compression across the cylinders is low enough to be out of spec, it would keep them from firing properly. Which is my segway to my next question, does it burn any oil?

If the compression is fine and it doesn't burn oil, I'd be looking more towards fuel supply. Pumps, filters, injectors including the cold start injector.

07-27-2014, 07:30 PM
thanks. I have not run a compression test and do not have one that I can use, at least not for down into the spark plug tubes. I have pulled the plugs and cleaned them up, they are not wet as if they have a massive leak problem. its just that it was having a problem occasionally dying when at low idle, and then I messed with it, now it won't run at all , except for once out of 6 tries it will start up and stay running as long as I've got it revved up. I went and pulled a distributor housing from pick-n-pull today and it looks like a denso reman so it should have less than the 180k miles that were on the camry I got it from, and a nice looking Toyota coil in it as well. the other thing the manual has listed under "stalls occasionally" is the map sensor so I snagged a couple of those too.

I have replaced the fuel filter in the last couple of months. the fuel pump is another one of those things I had planned to replace just because it very may well be 20 years old. but I can hear the pump when I hit the key, although I know that doesn't prove anything. but the codes I was getting before I started messing with it were distributor, igniter, coil, O2 sensor. I had a guy at autozone scan it with the OBDI tool and he said it needed a gas cap, which I forgot to get.

when I first got it the coolant seemed fine, a couple of months later I replaced the water pump and had noticed that the coolant was getting cruddy like it was leaking oil into the coolant. and it kept spitting coolant back out the overflow. I changed the cap and it didn't help, other people told me that's a symptom of the headgasket leaking exhaust into the cooling system. and after I flushed it, it got worse and so we think the motors been stopleaked. furthermore, it is a Camry motor with the balancing shafts on it and they just cut away a chunk of the transmission cover to make the oil pan fit. all the more reason why im going to have this guy on 6gc do a full rebuild on the 5sfe he pulled out of his gf's celica to do a turbo swap. ive got a couple of weeks off but ill need to get this thing working dependably for about another month in September. Im considering running some more stop leak through this one. I just want to be careful and not mess up the radiator or heater core.

im going to put this new dizzy together and see what happens.


07-27-2014, 07:34 PM
Autozone or similar will rent you a compression tester for $20 deposit they'll give you back when you return it.

07-27-2014, 09:03 PM
MCcelica: its quite possible to push exhaust into the coolant and not vice versa. That's actually the most common failure on the 7M.

If you are going to attempt a stop leak to temporarily "fix" it to get by until the new motor is built, contrary to a lot of other mechanics, Id say its worth a try. It wont significantly "gum up" your radiator. The safest (but not necessarily the strongest) products to use are the ones that require a system flush before and request you use straight water after. The good thing is they break down in the alcohol based anti-freeze so when you get the new motor in, you can run anti-freeze and essentially clean out any remaining product in your radiator. That might get you by until September. Worth a shot I say.

As for it starting and then immediately shutting off Id suspect fuel pressure as my first instinct and specifically the low voltage side of the circuit. Try jumpering the fuel pump at the diag block with a wire from Fp to B+ and see if that improves your condition. What that does is give the pump 12V all the time the key is on, rather than the 8-9v is gets when not under demand. The test will show either a weak circuit feeding the pump or a failing pump.

Another reason I suspect fuel issues is that when you have an ignition system failure, they usually aren't RPM dependent. Meaning they don't usually just go away at a certain RPM (usually). That would lead me away from coil/dizzy/ignitor. That's not to say some sensor isn't malfunctioning, but usually you can still start the car cold as the ECU uses a different set of parameters.

However, having said that, I would like you to check your TPS sensor readings again using the feeler gauge, only read the signals AT THE ECU. HERE (http://www.celicatech.com/bgbonline/94_6gmanual/J%20pin%20voltage%20charts.pdf) is a link that will show you the pinouts and voltage/resistance charts for testing at the ECU. If the TPS proves to be correct, then move on to testing the other sensors there at the ECU since you are there.

Other notes:
When cold the ECU doesn't use the O2 sensors input, so I don't believe that is any of your problem with it not starting. if it was and AFM based system I would suspect the AFM, as that classic behavior, but since its MAP, we have to look elsewhere. Replacing the MAP is a good start, but they are pretty reliable devices.

07-27-2014, 10:25 PM
Thanks for the expert advice Facime, that's exactly what I was hoping for. The code I was getting was a 14 - Dist, coil, igniter, circuit. so I went back to the picknpull and got a nice looking dizzy with a good Toyota coil. and it was tighter fitting but I got it to slip in and it fired right up, idles nice too. it's obdI and I don't really know how to properly set the timing so I just have the distributor about in the middle of the bolt slots on the flange. other people have told me on OBDI its not a big deal.
but here is something else strange. When I was trying to put the old distributor back in yesterday, you know how the coupler has a notch in it and you're supposed to line it up with the grove on the dist shaft, mine would Not go in that way and I tried for a freaking hour and it would not slip all the way in, then I just flipped it around 180(opposite the notch and groove) and it slipped right in. wtf?? this new one only wanted to line up when the notch and groove were on the right side. freakn weird. I wonder if at some point somebody had removed the coupler and put it back on the wrong way or something? either way it fired right up.

I am however interested in examining this TPS setting a bit more and getting it right. and the stop-leak im going to use is liquid ceramic, I wasn't going to use the shit with little pellets in it O.o.
I have a multi-meter and have a decent understanding of electronics and electrical systems although I'm not a professionally trained auto mechanic. (ive been to school to learn electrical and Hvac, now working on my mechanical engineering degree)
so I understand the basics of how to test volts and ohms, but I'm not sure about testing pins on the ecu, is the plug designed in a way that allows you to probe pins without removing the plug? surely it has to be something like that. I just hate poking at expensive stuff before I know all about it. and that chart says its Up under the dash, under the carpet? ill have to look for it and get back to you. I want to make sure ive got the throttle body and tps adjusted right.


07-28-2014, 03:47 AM
Cool glad the Dizzy worked for you.

Yes, you can backprobe the ECU. That means you leave the plugs connected and just use thin probes on the multimeter to touch the connections down inside the plugs.
the ECU for the 6G is on top of the center hump behind the console. You can get to it from the passenger side by pulling the carpet back..

If you are going to use a ceramic based sealer realize that its more difficult to get back of the system. Its a bit more of a permanent application. I would recommend then either the Blue Devil brand stuff or the Bar's Leaks Professional (the one with carbon fiber). Both are around $60 but I believe they are going to be more effective from what Ive read.

07-29-2014, 05:45 AM
ya, I was afraid maybe the compression was so bad I was going to have to stop-leak it. but now it runs fine 'cept the idle still isn't right, it wont die but it fluctuates a bit sometimes and you can tell the throttle opener is saving it. but I don't think the compression or head leak is so bad that I will even need to use the stop-leak so im not going to do it.
but the front left has been growling a bit and today the wind blew and I felt the whole car rock. So I got home and put it on a jack and grabbed ahold of the front wheel and it wobbled back and forth in my hand quite a ways. So I've got to deal with this wheel bearing first. and it's the same one I just did back in December so I need to figure out what I did wrong. they do have a couple 6th gens up at the pick-n-pull I think I should go up and grab the whole knuckle. I slid on the ice and bit the curb on that side last winter and we replaced the wheel bearing, ball joint, and CV on that side so maybe the spindle is damaged in some way, or we just torqued the axle nut wrong.

08-05-2014, 03:10 AM
bump for new thread continuation