Engine Stalls, Wild Goose Chase, Weird Voodoo "Fix"

[pwuser162557941033]

I have a Suzuki Jimny with an F6A turbocharged EFI engine.

For the longest time, I've been on a goose chase trying to figure out why my engine would stall, after it had warmed up, when pressing the clutch in. It would either stall and die or idle rough at 900rpm, lighting up the oil pressure light. The first time this happened, I had just changed to Valvoline SynPower 5W-30 API SP.

"Synthatec gaarha nahi hota, gari purani hay, 20W-50 dalen."

I'd rather not kill my turbo, but the oil pressure light was an obvious urgent fix. I switched to Delo Gold Ultra, changed spark plugs, but the issue persisted. I even tested my oil pressure, installed an oil pressure gauge, and am going to back to using 5W-30 or 0W-40 full synthetic oil at the next interval. So it likely wasn't a low oil pressure issue.

Eventually, the tiny hoses feeding coolant to the IAC valve in the manifold developed leaks. I bypassed them, and the car idled at 2000-2200rpm, but at least I had oil pressure.

Got sick of the ridiculous fuel consumption, got new IAC valve coolant hoses installed back in.

---

New Problem: engine would surge with clutch in and at idle, but only if I gently raised throttle to about 1200rpm. At that point, it would start surging. 2500rpm, 1000rpm, repeat.

Obviously, this made driving in stop-and-go traffic very difficult. It happened all the time.The only way to make it stop was to turn on the AC, and even that only slightly calmed it. Turning AC off, or when the compressor disengaged due to the evap coil thermostat, would make the surging come back.

Nobody knew what the hell was going on. Until, messing around with vacuum lines, we bypassed this idle control solenoid:

Once again, the idle shot up. No surging. Pinch one line off, engine stalls. Aha.

The old one seemed to not be working. I found an Afghani replacement, and it solved the problem.

Now, I'm trying to figure out what it does and how it's controlled. I notice it doing two things:

• Adjust idle, bringing it down when high and up when low. I'm confident that it's temperature-dependent. BUT IT ONLY does this (I checked) when:

1. Vehicle completely stopped.
2. Transmission in neutral
3. Clutch not pressed in

• Dropping idle back down after AC switch is turned off. Once the compressor clutch disengages, engine RPMs rise until "something" happens and they drop back down to 1000rpm. That "something" doesn't happen if this solenoid is disconnected, and surging starts again once engine rises to 1200rpm.

---

So... what is the control scheme for this thing? How does the ECU determine what to tell it? Is it a normally open solenoid like I suspect? Why does it compensate idle only when those three requirements are met??? I'm puzzled. Is this common across other cars from this era?

It's an inconvenience if the car fully warms up in traffic, because unless I come to a dead stop, put it in neutral and give it a second, the car won't set the idle at 1000rpm and "remember" it. It'll either do something like 1200rpm or 800rpm, or stall out, unless I do this little dance at a dead stop.

The solenoid gets very hot under normal operation, and the Afghan replacement seems to be failing again, behaving inconsistently like the previous one. I want to buy a brand new part, but it's roughly $200. Does anyone have any experience cleaning out and rebuilding these?

P.S. Installed an 88°C thermostat, using Glysantin G48 coolant, removed locked-solid clutched fan, installed electric fan with a shroud and temp switch rated for 88°C. Engine still doesn't warm up all the way (water temp needle slightly below halfway) unless it's a hot sunny day with AC on. When warm, it's more likely to stall unless I do the 3-step dance.

[saboor73]

I have a Suzuki Jimny with an F6A turbocharged EFI engine.

For the longest time, I've been on a goose chase trying to figure out why my engine would stall, after it had warmed up, when pressing the clutch in. It would either stall and die or idle rough at 900rpm, lighting up the oil pressure light. The first time this happened, I had just changed to Valvoline SynPower 5W-30 API SP.

> "Synthatec gaarha nahi hota, gari purani hay, 20W-50 dalen."

I'd rather not kill my turbo, but the oil pressure light was an obvious urgent fix. I switched to Delo Gold Ultra, changed spark plugs, but the issue persisted. I even tested my oil pressure, installed an oil pressure gauge, and am going to back to using 5W-30 or 0W-40 full synthetic oil at the next interval. So it likely wasn't a low oil pressure issue.

Eventually, the tiny hoses feeding coolant to the IAC valve in the manifold developed leaks. I bypassed them, and the car idled at 2000-2200rpm, but at least I had oil pressure.

Got sick of the ridiculous fuel consumption, got new IAC valve coolant hoses installed back in.

----------

New Problem: engine would surge with clutch in and at idle, but only if I gently raised throttle to about 1200rpm. At that point, it would start surging. 2500rpm, 1000rpm, repeat.

Obviously, this made driving in stop-and-go traffic very difficult. It happened all the time.The only way to make it stop was to turn on the AC, and even that only slightly calmed it. Turning AC off, or when the compressor disengaged due to the evap coil thermostat, would make the surging come back.

Nobody knew what the hell was going on. Until, messing around with vacuum lines, we bypassed this idle control solenoid:

Once again, the idle shot up. No surging. Pinch one line off, engine stalls. Aha.

The old one seemed to not be working. I found an Afghani replacement, and it solved the problem.

Now, I'm trying to figure out what it does and how it's controlled. I notice it doing two things:

* Adjust idle, bringing it down when high and up when low. I'm confident that it's temperature-dependent. BUT IT ONLY does this (I checked) when:

> 1. Vehicle completely stopped.
> * Transmission in neutral
> * Clutch not pressed in

* Dropping idle back down after AC switch is turned off. Once the compressor clutch disengages, engine RPMs rise until "something" happens and they drop back down to 1000rpm. That "something" doesn't happen if this solenoid is disconnected, and surging starts again once engine rises to 1200rpm.

----------

So... what is the control scheme for this thing? How does the ECU determine what to tell it? Is it a normally open solenoid like I suspect? Why does it compensate idle only when those three requirements are met??? I'm puzzled. Is this common across other cars from this era?

It's an inconvenience if the car fully warms up in traffic, because unless I come to a dead stop, put it in neutral and give it a second, the car won't set the idle at 1000rpm and "remember" it. It'll either do something like 1200rpm or 800rpm, or stall out, unless I do this little dance at a dead stop.

The solenoid gets very hot under normal operation, and the Afghan replacement seems to be failing again, behaving inconsistently like the previous one. I want to buy a brand new part, but it's roughly $200. Does anyone have any experience cleaning out and rebuilding these?

P.S. Installed an 88°C thermostat, using Glysantin G48 coolant, removed locked-solid clutched fan, installed electric fan with a shroud and temp switch rated for 88°C. Engine still doesn't warm up all the way (water temp needle slightly below halfway) unless it's a hot sunny day with AC on. When warm, it's more likely to stall unless I do the 3-step dance.

perform a vaccum leak test on your engine

[pwuser162557941033]

I will, soon. But even the presence of a vacuum leak doesn't explain why the idle compensation only occurs when I put the transmission in neutral, release the clutch, and come to a dead stop.

If even one of those three conditions is not met, and the car has warmed up to the point where it now idles at 800rpm, it won't compensate and go up to 1000rpm until all three conditions are met.

I really do think this has to do with the solenoid valve not reacting in time, but why it's occurring I have no idea.

[Xulfiqar]

The valve is only for idle adjustment, there is never an idle speed condition while moving.

You should also check if your throttle body bore is worn out, or adjusted out of spec - that can cause it too.

Idle hunting also happens due to the pcv valve causing a vacuum leak.

[pwuser162557941033]

So how does it know to adjust it? I'm assuming that it doesn't know about the clutch, the pedal is cable-driven. I don't think it would know if it's in neutral either. That only leaves the speedometer gear in the transmission. Is that how it knows the car isn't moving? If that's the only input, then how does it know to only adjust when the clutch is released?

And yes, I'll check the throttle body bore. How should I check the adjustment?

From what I can see, hose 19 is used for PCV, it leads straight from the valve cover into the compressor inlet, after the air intake filter. I don't think there is a PCV valve. Please correct me if I'm wrong and I should look for it elsewhere.

The intake path looks like this:
Air Filter Box, then PCV Hose > Turbo Inlet > Turbo Outlet > BOV, then intercooler > Intake Manifold

I checked the air inlet after the filter box for leakage by blocking it, and that causes the car to stall, so I don't think there's any cracks or fractures.

However, the idle hunt doesn't happen unless the solenoid I mentioned is disconnected or working improperly, and you open the throttle up gently to around 1200rpm.

Right now, for example, when the engine is at operating temperature (needle in the middle of the stock coolant temp gauge, and I have a thermocouple attached to the water temp sender) on a hot day, pressing the clutch in and coming to a stop causes the engine to stall. The only way to make it not do this is to stop, put it in neutral, and release the clutch quickly before it stalls. It'll then go from a shaky low idle back up to 1000rpm and stay there every time for the rest of the time I'm driving. It's almost as if it "learns" the new idle point using this procedure.

[pwuser162557941033]

Oh dang. I just noticed there are two temperature sensors (blue) with the same part numbers, along with the coolant temp gauge sender (red.)

I'll try and swap the connectors for these two and see how it behaves.

There is also an idle control valve in the manifold (green) which has two coolant lines plumbed into it. These lines I disconnected and later reconnected as I mentioned in the original post. I'm assuming this lets air in when cold and keeps idle high until the coolant warms up, then closes off when it reaches operating temperature, leaving only the primary throttle to let air in. I can not replace this if this is screwed up and causing the problem.

[pwuser162557941033]

Standard vacuum for 660cc F6A 4v naturally-aspirated engine at sea level is 15.8 - 18.8 inHg, or 53.5 - 63.7 kPa.

My F6A is 2v, turbocharged. Not sure how much of a difference it makes.

At 62°C coolant temp and 1000rpm, my boost gauge shows vacuum as ~~45 kPa~~ 50kPa.

Will double-check with an analog gauge.

[Xulfiqar]

Firstly make sure the cold start idle up thermostatic unit is working,

Remove it and clean it out with vinegar to remove all rust sediment, then test its ports for air passage when hot and when cold.

Once its working, then set idle speed on warm engine and retest your situation.

[pwuser162557941033]

You're talking about the part I marked in green above? I believe that's integrated into the throttle body. It seems to be working fine, because I had the plumbing disconnected for a while and the car would always have a high idle.

Now that it's connected, cold idle behaves like it should. ~2000rpm cold start, gradually dropping to 1000rpm as the coolant warms up. I filmed a time lapse to make sure. As oil pressure decreases and coolant temp rises, engine speed seems to decrease linearly.

However, I will take the time to take it apart and inspect it. It's had coolant in it since I've owned the car, but you're likely right about rust and sediment accumulation.

I'll say again, the car idles at 1000rpm quite stably, AC on or off, when warm. However, it seems to need this learning procedure I've described if it warms up while driving.

Lets say I leave the house with coolant temp 35°C and cold oil pressure 400kPa, driving slowly and keeping revs low as the engine warms up. Once idle oil pressure is ~100kPa and coolant temp reaches 60°C, if I'm still driving and come to a stop, the engine stalls. It seems to need to learn the new idle compensation set-point before it idles stably at 1000rpm.

Yes, it takes a long time to warm up, even with an 88°C thermostat. There is an oil/coolant heat exchanger present.

[pwuser162557941033]

Solution

The AC idle up vacuum solenoid valve has a set screw that I put exactly 2 counterclockwise rotations into. I did this right after triggering the stalling, which happened after a short drive after the car had been sitting in direct sunlight for a few hours, which caused the 850rpm idle, gasping for air behavior. Engine was stalling immediately unless AC was turned on or throttle applied. Additionally, I put the headlights on high beam, AC blower to the highest setting. I turned the screw until idle was 1000rpm.

The side effect is that the idle is higher, anywhere from 1500 to 1100 RPM, when the coolant needle is not exactly in the middle. Engine is basically never at operating temperature with a short trip <5km and without engaging boost, which I don't do unless the oil pressure reaches its operating temp value anyway.

Now the car idles perfectly at 950-1050rpm with the AC on at operating temperature. By that I mean the thermocouple I have sitting around the water temp sensor thread reads 75°C or higher. It isn't liquid-immersed, but it corresponds to the coolant temp gauge sitting right in the middle with an 88°C thermostat. The problem is essentially gone, and my AC blows much cooler.

I will update this thread if the engine stalls for no reason again.

---

Observations

I gather that apart from the ECU water and air temp sensors, there's three ways this engine compensates idle:

• A coolant-immersed, wax pellet-type, thermostatic cold engine fast idle valve, located in the throttle body
• The (idle stop?) VSV (the one pictured)
• The AC idle up VSV with a set screw

There is no neutral switch or clutch switch, so the car does not know it's in neutral or whether the clutch is pressed. There is no crank angle or speed sensor. It seems the only input for signaling idle conditions is the vehicle speed sensor in the gearbox.

My hypothesis is that only when the car is fully warm - which doesn't happen at all unless it's been sitting outside on a hot day and you do some driving on boost and with AC on - does the cold engine fast idle valve close completely, making the real idle set point apparent.

With a properly maintained cooling system, I've gone from having constant overheating problems and a blown head gasket to the car just not getting warm quickly enough in short, slow trips.

Since no mechanic has ever really taken the time to observe this behavior, decipher the vacuum routing, or understand the sensors, I've been running around trying to chase irrelevant crap based on their confidently false diagnoses.

Finally, to summarize: as long as these three items are in working order, and the ECU can see them and the associated sensors, and the idle set screw is set at operating temperature, the engine maintains idle and doesn't stall or run rough when warm. Hot idle oil pressure with 15W-40 is 60 kPa (9 psi) for me, oil pressure warning light is not on.

Please let me know if anyone has any feedback or corrections.

---

Questions / Unsolved
I still don't know whether the VSVs are normally open or normally closed. When the AC clutch disengages at idle, RPMs rise slowly until they quickly come down again, and this seems to be the role of the VSV pictured in my initial post.

If there's no crank speed or angle sensor, how does the ECU know when idle is at the set point? My guess is injector or ignition coil duty cycle. Or some kind of lookup table for the MAP sensor?

I'm missing the one EGR VSV. I don't think it matters, except that EGR is basically always on because the diaphragm always sees engine vacuum and diverts gases towards the intake. I haven't checked if the EGR system is working. If anyone suspects it is somehow involved in AFR calculations, please let me know; there are no O2 sensors. Yes, my fuel consumption is atrocious.

I have a cat temperature warning light permanently on in my dash. I was sent on another goose chase trying to find an O2 sensor for the hollowed-out catalytic converter housing, but there is none. It's a one-wire connector for a "thermal fuse," as Suzuki calls it. This fuse has a solid rod that pokes into the cat housing, with no openings to sample the gases, so it doesn't seem like an O2 sensor. Does the ECU use this to calculate AFR? How can it, if the unit doesn't have openings to sample the gas like every other O2 sensor?

[helper_learner]

when the coolant needle is not exactly in the middle. Engine is basically never at operating temperature with a short trip <5km

T-stat valve is missing or wrong temp, Or stuck open.

Car only took about 5-10 mins of idle to get to operational temp.

[pwuser162557941033]

I've checked and installed two 82°C and two 88°C thermostats. All of them open and close to the specified height, and only the 88°C ones bring the needle up to the exact halfway point. I'm assuming the gauge is calibrated for the 88°C type. JDM-spec car, colder average climates.

There's a difference of 100% between 5 and 10 minutes, not exactly a consistent measurement. On cold start and at idle, I've recorded the car taking an average of 15 minutes to get to 1000rpm, while the coolant temp gauge and thermocouple show that operating temp hasn't been reached.

[Xulfiqar]

I have a cat temperature warning light permanently on in my dash. I was sent on another goose chase trying to find an O2 sensor for the hollowed-out catalytic converter housing, but there is none. It's a one-wire connector for a "thermal fuse," as Suzuki calls it.

This is not a sensor, but a literal "fuse" - its only job is to trip the dash lamp on that is warning a cat converter overheat, (dont park on dry grass or something that can ignite) that part is no longer made and not available from any manufacturer anymore, I have a customer here who imported a toyota century V8 and it had that, - Its unobtanium from Toyota and even Denso warehouses globally, the only source would be a single item sitting on a dusty shelf in some obscure place in the world.

[Xulfiqar]

I still don't know whether the VSVs are normally open or normally closed. When the AC clutch disengages at idle, RPMs rise slowly until they quickly come down again, and this seems to be the role of the VSV pictured in my initial post.

If there's no crank speed or angle sensor, how does the ECU know when idle is at the set point? My guess is injector or ignition coil duty cycle. Or some kind of lookup table for the MAP sensor?

There is always a crank speed sensor (how else would it time the ignition) - even if its just a simple 3 pole reluctor to time the cylinders.

VSVs can be both types, NC or NO - it depends on what their purpose is.

But - good job on resolving it. It has the same setup like an old corolla, the AC has its own idle up VSV with a bleed adjustment. On the older corollas, the problem I usually faced was removed thermostat, completely jammed IAC (because it would default to high idle due to cold temp) - and the additional wax pellet valve also stuck open.

I had to remedy most with a thorough cleanup or replacement - usually the small metal coolant tubes would be clogged full of rust (highest point in system and air loves to trap there causing steam cavitation) - let the engine warm up fully to let the wax pellet open up - then refiddle the idle speed from the "verbotten" stop screw - (the screw that no mechanic in the world would touch, other than a Pakistani ustaad)

Finally start the AC and adjust the VSV to correct idle quality - the way it should work is that the idle up would start a few miliseconds before the compressor starts - this way the engine would never lower its idle speed and would smoothly keep running. (make sure your car has its original AC amp in it too - it enables that little delay), Similarly when the compressor disenages, you will feel the engine rpm raise a bit (car wants to nudge forward if in D and foot on brake) - this is all normal.

[pwuser162557941033]

This is not a sensor, but a literal "fuse" - its only job is to trip the dash lamp on that is warning a cat converter overheat, (dont park on dry grass or something that can ignite) that part is no longer made and not available from any manufacturer anymore, I have a customer here who imported a toyota century V8 and it had that, - Its unobtanium from Toyota and even Denso warehouses globally, the only source would be a single item sitting on a dusty shelf in some obscure place in the world.

Thank you! That's exactly what I suspected, and Suzuki calling it just that is a dead giveaway. Everyone I spoke to insisted it's an O2 sensor (because why else would it be on the cat?) completely oblivious to the fact that it has no ingress point for sampling the gases.

And I priced it out before I realized what it does, it's readily available from Suzuki dealers in Japan, anywhere from $150 - $200 shipped. I believe the fuse breaks at 300C, accounting for the EGT drop after the compressor. I can imagine it being impossible to find for the Century though!

But - good job on resolving it. It has the same setup like an old corolla, the AC has its own idle up VSV with a bleed adjustment. On the older corollas, the problem I usually faced was removed thermostat, completely jammed IAC (because it would default to high idle due to cold temp) - and the additional wax pellet valve also stuck open.

I had to remedy most with a thorough cleanup or replacement - usually the small metal coolant tubes would be clogged full of rust (highest point in system and air loves to trap there causing steam cavitation) - let the engine warm up fully to let the wax pellet open up - then refiddle the idle speed from the "verbotten" stop screw - (the screw that no mechanic in the world would touch, other than a Pakistani ustaad)

Yeah there's no way I'm touching that screw lol. You're absolutely right about that, I've had to scold mechanics who were trying to fusk with it without hesitation.

Yes, when I bought the car, there was the usual litany of cooling system violence that we collectively inflict on our cars. I had every rubber hose replaced, the little coolant flow valve that feeds the heater core, put in a new thermostat, installed an electric fan (fan clutch dry and locked solid, new clutch Rs. 100k+ shipped excl. fan blades, not worth it, eased the load on the crank and water pump anyway) and flushed my radiator and block out with citric acid, filled in G48, because I couldn't find G05.

I haven't had cooling issues since, except the odd overflow reservoir decrease that I have to trace and fix time and again because of our shit quality counterfeit hoses, worm gear clamps, and tendency for mechanics to "silly-cone" hoses when I'm not looking.

I'm dumbfounded that people don't realize the abuse their mineral water-cooled passenger vehicles go through. Steam in the lines, corrosion, and cavitation damage to basically everything sounds horrific, and this is 5th grade Physics we're talking about - doesn't matter how "meetha" your "line ka paani" is.

The tiny turbo is both oil and water cooled, and looks exactly like the Garrett GT12. There's a large steel line that was packed with rust that runs coolant into it, teed off of the lower radiator hose, and I don't look forward to replacing it. I shudder to think of the damage all those years of mineral water have done to the turbo.

Finally start the AC and adjust the VSV to correct idle quality - the way it should work is that the idle up would start a few miliseconds before the compressor starts - this way the engine would never lower its idle speed and would smoothly keep running. (make sure your car has its original AC amp in it too - it enables that little delay), Similarly when the compressor disenages, you will feel the engine rpm raise a bit (car wants to nudge forward if in D and foot on brake) - this is all normal.

Fantastic, thank you, I did not know that delay and control was by the amp. This is super co-incidental: I discovered recently the amp's been thrown out, and replaced with taped together underrated relays and an ancient thermostatic switch with a copper probe. I posted about it here. The compressor clutch engages as soon as the button is pressed, causing engine speed to drop. A new (or even used) AC amp is at least $100, not to mention probing it for the pinout and wiring it correctly with the right harness/"girip" to undo the damage.

[pwuser162557941033]

even if its just a simple 3 pole reluctor to time the cylinders.

Oh holy shit is this it? In the distributor? I'm assuming this is the signal generator that's outputting a pulse to the ECU.

[Xulfiqar]

That signal generator is telling the ECU two things.

engine is spinny spinny at falana falana speed
a cylinder is approaching TDC - it doesnt care about compression or exhaust as the dizzy cap job is to time the ignition events to the correct cylinders.

quite same as older corolla, the "crankshaft position and speed sensor" was inside the distributor.

[Xulfiqar]

And I priced it out before I realized what it does, it's readily available from Suzuki dealers in Japan, anywhere from $150 - $200 shipped.

I doubt that, on the century sensor fiasco, I had 6 different vendors confirm the order only to contact me back for its cancellation due to non-availability. Technically speaking there arent much warehouses in Japan due to its size limitations, thats why JIT manufacturing is such a focus for them.

[pwuser162557941033]

Seems like a fuse and thread adapter would be something you could find at McMaster-Carr if someone was really desperate to get one. This model year Jimny is pretty popular, especially with them being legal to import in the US since a couple years ago. I was able to get in touch with some shops in Japan that have new old stock lying around. Guy in Wisconsin even has a store full of odd parts like the steering box, transmission towers, U-joints, etc that match with the Samurai - but that's the States for you. It's all warehouses.

Anyway, the engine is still stalling, but only at operating temperature and when the AC is off - which suggests to me that it's that vacuum solenoid valve misbehaving.

With AC on, no stall, maintains idle presumably due to the AC idle up VSV. Turning AC off gives idle control back to the idle stop VSV, which is the part I replaced. Seems like it's sticky, and at a dead stop, takes some gentle throttle just before the lower sensing range of the TPS to coerce it to stabilize.

Could I perhaps open mine up carefully and clean it out? I don't think it's meant to be disassembled, or would survive disassembly if there's a plastic valve that's on the edge of disintegrating.

[Xulfiqar]

Wd40 in the ports and a cyclic on/off helps