91 ST185 stalls out when hot, then won't start

Lucid

New member
Hi guys, I'll try and keep this brief.

So I'm having problems with my 1991 ST185 Alltrac again (Gen 2 3SGT). Before I stored the car for the brutal Canada winter last year, it was having some issues. I would drive the car on an expressway for a bit, after which the throttle response would die, the AFM would not read anything, and the only solution to keep the car from stalling out would be to keep it limping between 2nd and 3rd gear and pray I could get it to go. AFM would read extremely pig rich (<9) under boost (max 14psi). After I got to my destination, the car became a crank no start.

Over the winter, I replaced the fuel pressure regulator, as I was 90% certain that it was causing low fuel pressure. Afterwards, car started up no problem at all, and I drove if fine for a few weeks up until recently. Fast forwards to yesterday, after coming back from a drive out of town, it did the exact same thing as I described before.

I know there is a really nasty exhaust leak between the turbo and the manifold. I already have a metal gasket on order from PrimeMR2, just waiting for it to arrive so I can replace it, along with a new aluminum radiator at the same time. However, I don't *think* that should be related, should it? Maybe it is affecting the O2 sensor readings and tripping out the ECU into not sending fuel/spark. Or the heat from the exhaust leak is making something else (ignition component?) act up. At this point, I'm pretty sure the issue is ignition related, but I'm gonna check for ECU codes and continue to test stuff today/tomorrow. In the meantime, would you guys have any ideas as to what it might be based on these symptoms?
 

grip-addict

Active member
Can you check for spark? A coil can overheat and cause this symptom.
Could also be fuel related if the fuel is cooking before it hits the rail. Feel the feed line and see if it's cool to the touch when the issue occurs.
Also try jumping fp and b+ in the diagnostic box and turn the key to on for a minute or two. That'll cycle the fuel through the system and get some fresh cold fuel from the tank in the lines. See if that clears the condition.
 

underscore

Well-known member
That sounds like you're losing your fuel flow, either from the pump acting up or the power feed cutting out. Mine did something similar, but much more quickly, when I had a defective EFI relay in the engine bay.
 

Lucid

New member
Can you check for spark? A coil can overheat and cause this symptom.
Could also be fuel related if the fuel is cooking before it hits the rail. Feel the feed line and see if it's cool to the touch when the issue occurs.
Also try jumping fp and b+ in the diagnostic box and turn the key to on for a minute or two. That'll cycle the fuel through the system and get some fresh cold fuel from the tank in the lines. See if that clears the condition.
I think this theory is probably #1 for validity. Fuel rail line was just as hot as the head/block when I felt it after coming back from out of the city. Then, after the car has sat for a while to cool off, its starts fine. Wouldn't be surprised if the exhaust leak was making the under hood temps wayyy too hot and just vapourizing the fuel in the line. Checking the coil pack is also a good idea too.
That sounds like you're losing your fuel flow, either from the pump acting up or the power feed cutting out. Mine did something similar, but much more quickly, when I had a defective EFI relay in the engine bay.
Thought it was my fuel pump at first too. It's hard to diagnose stuff just because of how random it is and the time it requires to get the car to act up, if it even decides to in the first place. I'd like to avoid firing the parts cannon as much as possible with this thing (I'm still a poor university student 🥲). Are bad fuel pumps and EFI relays known to just stall the engine out immediately or let it limp for a while, like mine did?

I think my next step is to replace that turbo to manifold gasket and then maybe just monitor it to see if it will keep doing the same thing. If so, some more testing of individual stuff is needed. I'm a lot more confident that its causing fuel/ignition components to overheat (maybe also alternator just cause its so close to the leak itself). Plus, I think it will also fix the rich under boost/insane lag issues I'm having also hehe.
 

Lucid

New member
Update: Now, the car doesn't want to start even when cold. I managed to drive it into my garage after sitting for 2 days, then it became a crank no start again. This has me convinced that, whatever the problem is, it isn't temperature related/exhaust leak related.

I did some diagnosis today, just by myself based on the tips given. Tested EFI and fuel pump relays, both work fine. I was trying to find the COR relay and the fuel control relay from the ECU however, looking at wiring diagrams, I was unable to locate it. Tested the coil and it sparks up no problem. Visually checked the fuel lines for any leaks and I couldn't find any. Jumped B+ and FP and still nothing, I couldn't even hear the fuel pump spin up. At that point, I was about to undo the main fuel line and check if the pump is even sending any fuel pressure, but I had an even better idea. I looked underneath the fuel pump relay box, found a blue-white and blue-black wire and traced them to the fuel pump power connector in the trunk. I hooked up my multimeter to this connector and, sure enough, it reads 10.39 volts when I try and crank the car over. After I stopped cranking it, it would read 11.5 volts for 1-2 seconds after. I know there usually is some voltage drop to the fuel pump, but is 10.39 volts normal? Should there always be around 12 volts straight to the fuel pump? I was reading and know the ECU puts the signal to the pump through a resistor occasionally to drop the voltage. There is also another connector that goes to the fuel pump housing too, I assume that's for the fuel level sensor/misc?

Unless anyone can suggest any other checks to do, is there a good option for an OEM+ fuel pump that will fit right into the pump housing? This will be my last parts cannon firing, I'm really trying my best to make sure it's 99.9% the fuel pump, so if this doesn't fix it, I'm just gonna take it in to a local shop I like for them to figure it out. Thanks guys.
 

grip-addict

Active member
When I picked up my st185, someone had already bypassed the fuel pump resistor and had directly spliced both of the fuel pump power wires together to make a single circuit to the fuel pump. I was troubleshooting fuel problems and also observed low voltage, values very similar to what you have.
I think low voltage is a weakness in the Toyota design but I don't think that's the cause of your problem.

mk4 supra fuel pump is a direct drop in. Otherwise, walbro 255 lph pump also fits but you will just have to cut/splice the ground and 12v wire as the connector is different.
 

alltracman78

Active member
It seems like you're flailing around here. Let's start from the beginning and straighten some things out first.

I'm guessing this is on a stock ECU? How are you reading inputs? AFM signal (as well as most others) is an analog signal. You're best reading it with an analog volt meter. Second best would be an analog ohm meter to test the sensor itself. But the best way is reading the live signal (analog voltage) at the ECU connector (it's easy to insert a small needle into the back of the connector, just make sure it's a steady contact). You don't want a regular DMM (digital mulitmeter). 90% Of the time the DMM will work fine, the rest it will make worse because you won't be able to see the problem. And analog meters are cheap.

AFM doesn't read rich or lean. It just gives you how much air is entering the engine. How much fuel is mixed with the air inside the engine gives you rich or lean.
(Since I'm on my soap box, regarding the O2 sensor and rich/lean; O2 only reads oxygen. It can't tell if the mix is actually rich or lean. For example if you have a misfire your O2 sensor will read lean because there's too much O2 in the exhaust. Just theory to keep in mind.)

AFM signal can't be 9V. The AFM gets it's voltage from the ECU. The ECU sends out a 5 volt (not 12 volt like you'll see some places, they're wrong) reference voltage to certain sensors. This is so the ECU knows EXACTLY how much voltage the sensors are supplied with because the return voltage is important. For example if the AFM is supplied 8 volts and all of a sudden it is supplied with 12 volts the return signal to the ECU will go up (which is how the ECU "reads" how much air is coming in), even if the air supply doesn't change.
5V Is used because you want considerably less than battery voltage so it's easy to keep constant. Connections and length of circuit absorb some voltage, and if the alternator goes out you lose more. This way those systems won't be effected.

5V Into your AFM. VS Is the air intake signal. IIRC Voltage should increase as air flow increases (as the flap door opens). You should be seeing just above 0 volts at idle and just below 5 at wide open (not necessarily the same as WOT). More important than exactly how much voltage is a smooth signal. There shouldn't be any blips or jumps in the needle on the analog volt meter as it sweeps up and down (can't see this on a DMM).
You can test with the key on and engine off by having someone open and close the flap and you can drive around and keep an eye on the needle.

Your COR (Circuit Opening Relay) is attached to your ECU by 2 screws on a bracket. On the drivers side. Only one on there so easy to see.

2 Speed fuel pump. There's a relay (yellow relay called the fuel pump {FP} relay in the relay block by the alternator) that switches between high and low speed by bypassing the fuel pump resistor for high speed or sending power through the FP resistor for low speed. Normal driving uses low speed (less fuel), at roughly WOT the ECU closes the relay which bypasses to high speed.
You can bypass the relay by removing it and installing a jumper. There's actually a factory jumper that came in the 5SFE engined Celicas. That's the best option IMO. You won't chance a connection problem with the wrong thickness pins this way.

Your fuel pump should still start and run and power the engine just fine on 10 volts. Remember the low speed setting from the FP relay?

You won't necessarily hear the fuel pump running. And it only works when the engine is trying to start or running. It won't run when the key is just turned to on, no matter what some other people will say. This seems to be a very common misconception.
One caveat. It will run with the key on if you jump FP and B+. That's not normal condition however.

I'm not saying your fuel pump isn't the problem. Toyota pumps are VERY reliable. But these cars are getting old at this point. Just make sure you understand how the system works before you throw a pump at it.
If you do replace the pump a factory or Denso pump is your best bet. You don't need an upgraded one unless you're looking for more power.

You said you had no AFM signal at one point.
The engine will start without an AFM signal, but it won't run without one.

There is no fuel pump switch in the AFM on these cars (most cars that use an AFM will have one).

Engine getting hot and then not running/starting is a classic symptom of a bad coil. Unfortunately at this point the other electronics are so old it could be them as well. 10/15 Years ago the first thing I would look at would be the coil. Still a good place to start but not as likely of a culprit.
Also would tend to be less intermittent with time.
When it won't start do you have good spark at the plugs?
You can also check for spark with a timing light (which you should own anyways if you have one of these).

When it won't start does it even try to start? If it tries but won't start more likely some fuel but not enough. If it won't even try to start either no spark at all or no fuel at all.
You can also spray a healthy dose of starting fluid into the engine. If it still won't start you have a spark problem. If it does start you have a fuel problem.

Unlikely your exhaust leak is causing this. Fix it ASAP though, weirder things have happened, and it's just good business to fix problems instead of leaving them.

Just to start the ECU needs fuel and spark (and compression, but we know that's not your problem here). To keep running is more complicated.
Since your engine is dying and won't start at all see if you have no fuel or no spark. Once you find out which you can trace it back and see why it's not getting it.

One more thing, just for the record. The parts cannon isn't necessarily a bad thing. Sometimes it's quicker and simpler to just go down the list than go through a whole diagnosis.
The problem is when you don't know what you're doing, or it's a more uncommon issue.
 

Lucid

New member
Update: Took my shot and replaced the fuel pump. Car runs fine now, getting that fuel tank in and out was a pain though. Going to take it for a longer test drive to see if it still holds up after I throw in this replacement radiator and cooling fans.
It seems like you're flailing around here. Let's start from the beginning and straighten some things out first.

I'm guessing this is on a stock ECU? How are you reading inputs? AFM signal (as well as most others) is an analog signal. You're best reading it with an analog volt meter. Second best would be an analog ohm meter to test the sensor itself. But the best way is reading the live signal (analog voltage) at the ECU connector (it's easy to insert a small needle into the back of the connector, just make sure it's a steady contact). You don't want a regular DMM (digital mulitmeter). 90% Of the time the DMM will work fine, the rest it will make worse because you won't be able to see the problem. And analog meters are cheap.

AFM doesn't read rich or lean. It just gives you how much air is entering the engine. How much fuel is mixed with the air inside the engine gives you rich or lean.
(Since I'm on my soap box, regarding the O2 sensor and rich/lean; O2 only reads oxygen. It can't tell if the mix is actually rich or lean. For example if you have a misfire your O2 sensor will read lean because there's too much O2 in the exhaust. Just theory to keep in mind.)

AFM signal can't be 9V. The AFM gets it's voltage from the ECU. The ECU sends out a 5 volt (not 12 volt like you'll see some places, they're wrong) reference voltage to certain sensors. This is so the ECU knows EXACTLY how much voltage the sensors are supplied with because the return voltage is important. For example if the AFM is supplied 8 volts and all of a sudden it is supplied with 12 volts the return signal to the ECU will go up (which is how the ECU "reads" how much air is coming in), even if the air supply doesn't change.
5V Is used because you want considerably less than battery voltage so it's easy to keep constant. Connections and length of circuit absorb some voltage, and if the alternator goes out you lose more. This way those systems won't be effected.

5V Into your AFM. VS Is the air intake signal. IIRC Voltage should increase as air flow increases (as the flap door opens). You should be seeing just above 0 volts at idle and just below 5 at wide open (not necessarily the same as WOT). More important than exactly how much voltage is a smooth signal. There shouldn't be any blips or jumps in the needle on the analog volt meter as it sweeps up and down (can't see this on a DMM).
You can test with the key on and engine off by having someone open and close the flap and you can drive around and keep an eye on the needle.

Your COR (Circuit Opening Relay) is attached to your ECU by 2 screws on a bracket. On the drivers side. Only one on there so easy to see.

2 Speed fuel pump. There's a relay (yellow relay called the fuel pump {FP} relay in the relay block by the alternator) that switches between high and low speed by bypassing the fuel pump resistor for high speed or sending power through the FP resistor for low speed. Normal driving uses low speed (less fuel), at roughly WOT the ECU closes the relay which bypasses to high speed.
You can bypass the relay by removing it and installing a jumper. There's actually a factory jumper that came in the 5SFE engined Celicas. That's the best option IMO. You won't chance a connection problem with the wrong thickness pins this way.

Your fuel pump should still start and run and power the engine just fine on 10 volts. Remember the low speed setting from the FP relay?

You won't necessarily hear the fuel pump running. And it only works when the engine is trying to start or running. It won't run when the key is just turned to on, no matter what some other people will say. This seems to be a very common misconception.
One caveat. It will run with the key on if you jump FP and B+. That's not normal condition however.

I'm not saying your fuel pump isn't the problem. Toyota pumps are VERY reliable. But these cars are getting old at this point. Just make sure you understand how the system works before you throw a pump at it.
If you do replace the pump a factory or Denso pump is your best bet. You don't need an upgraded one unless you're looking for more power.

You said you had no AFM signal at one point.
The engine will start without an AFM signal, but it won't run without one.

There is no fuel pump switch in the AFM on these cars (most cars that use an AFM will have one).

Engine getting hot and then not running/starting is a classic symptom of a bad coil. Unfortunately at this point the other electronics are so old it could be them as well. 10/15 Years ago the first thing I would look at would be the coil. Still a good place to start but not as likely of a culprit.
Also would tend to be less intermittent with time.
When it won't start do you have good spark at the plugs?
You can also check for spark with a timing light (which you should own anyways if you have one of these).

When it won't start does it even try to start? If it tries but won't start more likely some fuel but not enough. If it won't even try to start either no spark at all or no fuel at all.
You can also spray a healthy dose of starting fluid into the engine. If it still won't start you have a spark problem. If it does start you have a fuel problem.

Unlikely your exhaust leak is causing this. Fix it ASAP though, weirder things have happened, and it's just good business to fix problems instead of leaving them.

Just to start the ECU needs fuel and spark (and compression, but we know that's not your problem here). To keep running is more complicated.
Since your engine is dying and won't start at all see if you have no fuel or no spark. Once you find out which you can trace it back and see why it's not getting it.

One more thing, just for the record. The parts cannon isn't necessarily a bad thing. Sometimes it's quicker and simpler to just go down the list than go through a whole diagnosis.
The problem is when you don't know what you're doing, or it's a more uncommon issue.
There's really good info in here, thank you. Even though there wasn't a big problem with my AF ratio, I understand it a bit better now. I have an AEM aftermarket wideband gauge with a seperate O2 sensor further down from the stock one, with the catless downpipe the previous owner all installed.
 
Top