SPFI "cold" weather won't start - now running

[msteel]

I currently can't get my '92 Loyale going. Here's the story:

 

Back before Christmas we had some colder than normal days - here that means it is below freezing for a significant portion of the day. One of those mornings I started up the engine and it started right up, but it only ran for 5-10 seconds, then died. I cranked it some more but it wouldn't catch. Actually, I cranked it enough to run down the battery.

 

I didn't have time to mess with it then so I drove the Sienna for a day or two. When I had a chance to spend a little time on it it was warmer. I added 5 gallons of gas and a bottle of fuel line antifreeze to the tank (obviously I was thinking along the lines of a frozen fuel line). I jumped the battery and cranked. It acted pretty well like it has the 2 or three times I've run it out of gas - a lot of cranking interspersed with turning the key a few times to run the fuel pump - and eventually it caught a couple of times then started. Then it ran fine for a couple of weeks.

 

Well, we were out of town for over a week and the Subie sat. I drove it once after we got back. It was quite cold that morning but it ran fine. A couple of days later it did the same thing as before - started right up, died after a short period and wouldn't hardly even catch after that. So it's been sitting there a couple of days and I need to get it going again soon. This time the gas tank is full because I filled it up before we left town.

 

Anybody have an opinion on where to start looking? I guess I would probably start by jumping it again and cranking it a bunch, making sure there is spark, and then pulling a plug to verify that fuel is reaching the cylinder. If it actually was water in the fuel the first time (not so sure now) then replacing the fuel filter would be a good move.

Edited January 16, 2010 by msteel

[Frank B]

Replacing the fuel filter is a good move anyway. When it doesn't start, do you hear the fuel pump coming on when you turn the key?

[naru]

I`m guessing a worn fuel pump that doesn`t like cold temps.

If so,tapping on it while someone cranks the engine will probably get it going.Could be a bad relay too.Or CTS connection.

 

It ran on residual fuel pressure on the occasions when it only ran for 5-10 sec,the fuel pump never started.

Edited January 8, 2010 by naru

[hooziewhatsit]

I`m guessing a worn fuel pump that doesn`t like cold temps.

If so,tapping on it while someone cranks the engine will probably get it going.Could be a bad relay too.Or CTS connection.

 

It ran on residual fuel pressure on the occasions when it only ran for 5-10 sec,the fuel pump never started.

 

I concur. My BILs car had that same problem. I won't say how long it took me to find

 

When it doesn't start, connect the green test connectors, turn the key to On, then listen for the pump. Then tap it if needed to see if it comes on.

[msteel]

It also seems to me like it was running on residual fuel pressure.

 

But I don't see any evidence that the fuel pump itself wasn't working. I could hear it run for 1-2 seconds every time I turned the key from off to on. I was doing that on purpose to try to move fuel through the line. The pump sounded the same as it always has. If the pump was frozen or locked up I expect it would have sounded different.

 

Now the coolant temperature sensor is an interesting idea. Is is possible for the CTS to be bad in such a way that the ECU doesn't give it the right amount of fuel to start in cold temps, but won't throw a code? I've wondered about the CTS because of another intermittent problem but any time I've measured it it seemed in spec.

[naru]

Yes,but,probably not your problem as you have recently inspected the connector.

 

I`m wondering if the pump runs while cranking?

A fuel pressure test would be good.

Failing that,at least direct the filter output into a container to check flow while cranking.

Note how plugged the old filter is.

 

Check for spark first!

 

Could be a cold weather vacuum leak too.

[Cougar]

Try spraying some starter fluid into the intake the next time the trouble happens. If the trouble is with the fuel system it will start and run briefly. If you get no action from the engine then suspect the ignition system for the trouble.

[msteel]

This morning:

1) Fuel pump runs briefly when key is put to "on"

2) Fuel pump runs while cranking

3) Fuel pump doesn't pulse on and off with green test connector connected (I don't think I care because of #2 above) but the purge solenoid does pulse.

4) NO SPARK at plug or input to distributor.

5) Distributor turns.

 

Not totally making sense but I'll work on the spark before going back to fuel...

Once I find the screw that holds the distributor cap on

[Gloyale]

No ignition pulse will make the pump not run when cranking.

 

You may have a distributor issue.

 

Check it's connector. And really it would be a good idea to open it up, remove the lower cover, and spray some brake cleaner on the wheel in there. Sometimes grime will "clog" the slits in the wheel and give funky Disty operation.

[msteel]

The fuel pump DOES run while cranking (well, OK, I can't hear it while cranking, but I can hear it continue for a split second after I quit cranking. To me this means it has been running while cranking).

 

I DO have pulse across the two-wire connector that feeds the power transistor (F43 on the wiring diagrams). This should clear the Distributor and ECU.

 

I DON'T have pulse across the coil input.

I Can't find the appropriate diagram that shows all of the ignition conections. But think I am supposed to have power on the round connector to the coil and I don't.

[Gloyale]

The fuel pump DOES run while cranking (well, OK, I can't hear it while cranking, but I can hear it continue for a split second after I quit cranking. To me this means it has been running while cranking).

 

I DO have pulse across the two-wire connector that feeds the power transistor (F43 on the wiring diagrams). This should clear the Distributor and ECU.

 

I DON'T have pulse across the coil input.

I Can't find the appropriate diagram that shows all of the ignition conections. But think I am supposed to have power on the round connector to the coil and I don't.

 

Yes, you should have power to that round connector. IIRC on the Black/White stripe wire.

[naru]

This morning:

1) Fuel pump runs briefly when key is put to "on"

2) Fuel pump runs while cranking

3) Fuel pump doesn't pulse on and off with green test connector connected (I don't think I care because of #2 above) but the purge solenoid does pulse.

4) NO SPARK at plug or input to distributor.

5) Distributor turns.

 

Not totally making sense but I'll work on the spark before going back to fuel...

Once I find the screw that holds the distributor cap on

 

Starting to sound like a wonky ignition amp on the coil bracket.

 

If you have 12V at coil - w/key on,no trouble codes and a test light that does not flash when connected to coil - while cranking the amp is highly suspect.Check connections and grounding.

[naru]

The fuel pump DOES run while cranking (well, OK, I can't hear it while cranking, but I can hear it continue for a split second after I quit cranking. To me this means it has been running while cranking).

 

I DO have pulse across the two-wire connector that feeds the power transistor (F43 on the wiring diagrams). This should clear the Distributor and ECU.

 

I DON'T have pulse across the coil input.

I Can't find the appropriate diagram that shows all of the ignition conections. But think I am supposed to have power on the round connector to the coil and I don't.

 

Sounds even more like the ignition amp.

[msteel]

The primary of the coil reads a dead short.

 

If I understand how things work, then that would not only kill the spark, but also the tach signal and hence the fuel pump.

 

In any case it has lasted long enough and a new coil won't hurt anything.

[Cougar]

I doubt that the coil is bad. You stated you think you have a short from the primary side of the coil. Is that to the minus side of the coil or ground? A better check is to see if you have voltage getting to the minus side of the coil with the ignition ON. If that is ok then check to see if you have pulses getting to the minus side of the coil while cranking the engine using a test light probe with the clip lead tied to ground. You most likely will see a steady light indicating that the CAS in the disty is bad or the ignition amp is bad. Since this trouble seems to be intermittent there may be a bad connection in those areas causing the trouble.

Edited January 9, 2010 by Cougar

[naru]

The primary of the coil reads a dead short.

 

If I understand how things work, then that would not only kill the spark, but also the tach signal and hence the fuel pump.

 

In any case it has lasted long enough and a new coil won't hurt anything.

 

They are close to a dead short,but,not quite.

Dead short across the winding would not kill the tach signal as long as the fuse held.

 

If you have a "diode test" on your meter,use it to check for an open junction on the power transistor/ignition amp.

 

IMO you should get a coil w/bracket and amp from the junkyard instead.

[msteel]

They are close to a dead short,but,not quite.

Dead short across the winding would not kill the tach signal as long as the fuse held.

 

Yes, I realize that now. The thing is, my meter was showing lower resistance across the winding than just connecting the test leads together. That didn't make sense so I tried another meter and read 1.2 across the coil and 0.2 just shorting the leads. That difference of 1ohm seems to be right for the primary of the coil.

 

If you have a "diode test" on your meter,use it to check for an open junction on the power transistor/ignition amp.

Even if the transistor wasn't bad before it is now. I destroyed it trying to rremove it from the bracket. It is not socketed like I thought. Rather it was bolted to the bracket, connected, and potted. Mine now has a loose lead (that is, wiggly inside the TO-3 case) which means it is definitely bad.

 

IMO you should get a coil w/bracket and amp from the junkyard instead.

Would that I could. These cars are rare in junkyards around here. It may be I can find one. But they're all closed now until Monday.

 

If I can find the specs on the transistor I might be able to replace it out of my junk pile. So far I haven't been able to cross reference it. All I can say about it at this point is that it is in a TO-3 case and stamped 1501. From the way it is connected it could be a bipolar NPN. I expect it won't be too picky on specs as long as it can handle the currrent and the back EMF from the coil.

Edited January 9, 2010 by msteel

[naru]

Sorry to hear that.

They are a little specilized.Big back emf from the coil,me thinks.

 

This one might work:http://www.allspectrum.com/store/pro...b7eb4563092414

 

Simplier to get a used one though.

[msteel]

Jacinta is running!!!

 

I got an ignition amp from mdjdc and got it in this morning. Jacinta fired up the first time.

 

Because it is pretty warm (40's) this morning I am not yet convinced whether the amp was the original problem. The fact that it started right up does support that because a fuel issue would have probably required some cranking. Time will tell.

 

Along the way I learned some more about the ignition amplifier module and how it is wired. I'm sure many here know all this already but some this info might be of interest to others.

 

The two-bladed connector on the ignition amp is the input to the coil. It carries the ignition timing signal from the ECU (white/yellow wire) and power (black/white wire).

 

As best as I can tell the amp module itself is mostly an NPN darlington transistor with a protection diode, in a TO-3 case. Electrically, this is a common combination. However, because the TO-3 case is grounded through the coil bracket, it needs to be the emitter terminal on the transistor. This is is unusual since every TO-3 transistor I looked up had the collector connected to the case.

 

The ECU signal goes to the transistor's base, and the power simply passes through the amp module to the + side of the coil, and from there through the round connector to the condenser (capacitor) which is taped into the wiring harness. The collector of the transistor goes to the - side of the coil and becomes the tach signal. It also goes through the round connector to the regions beyond. The emitter of the transistor is grounded through the heat sink on the coil bracket.

[Txakura]

I got lost after 'Jacinta is running'

 

then something something pdq zistor in a case with stuff makes magic thingys happen and it runs

[msteel]

I got lost after 'Jacinta is running'

 

then something something pdq zistor in a case with stuff makes magic thingys happen and it runs

 

Yeah, sorry.

 

But the big thing is that the round connector actually feeds 12v from the coil on to the condenser, not into the coil. So if you unplug it you won't read any +12 on the harness there.

 

The rest of the stuff is only going to be useful if you need the part today, can't get the real thing, and have the resources to build something functionally equivalent.

[naru]

The amp is probably similar to a BU941.

http://www.datasheetcatalog.org/datasheet/SGSThomsonMicroelectronics/mXsxr.pdf

 

I think you are good to go in cold weather.YAY!

[msteel]

The amp is probably similar to a BU941.

 

Functionally, yes. But even that one has the collector on the tab, so it can't be mounted directly to the grounded heat sink.