Weird non-start issue

[Baxta]

Hi gang -

 

I've been a member for almost a year now. Bought a 93 for my daughter to learn to drive on last November. It's been a 4th vehicle for us, sitting for most of the past year. This summer, in trying to figure out an issue that had it running too rich, I replaced the coolant temperature sensor. (VW guy here - that's one of the first steps in that world...)

 

Anyhow, the vehicle sat for a couple of months after my wife took it on a girls road trip, where it ran well for them. Last weekend, I started it up and changed parking spots. A couple of days ago, I went to start it up, and no go. Lots of cranking, no catching. Here's where it gets weird: when I unhook the coolant temp sensor that I replaced, it starts and runs like a dog. Really really rich, but it starts. So I figure that I had a bad sensor. I bought a new sensor today, and same deal. I didn't install it, but hooked it up hanging loose, and it cranks like crazy, but no start. Unplug the new sensor, and it chugs to a start and again, runs like a dog.

 

I've gone over the service manual, but can't figure out the EFI system. There doesn't seem to be anything in the troubleshooting part of the manual that would cover this.

 

Can anyone out there give me a place to start looking?

 

 

Cheers,

 

 

 

Greg

[SkiMachine]

Take a look at page EC-23 (1994, 93 may be different) in the FSM - to summarize it states the ECM will assume the coolant temp is 20C at start-up if you disconnect the sensor.

 

The ECM aims for a richer fuel mixture the colder the engine is. Thus if the coolant temperature is less than 20C (which I'm sure it is right now, even in Langley) when you are trying to crank it then the ECM will aim for a richer fuel mixture than it would with an assumed (ECT sensor disconnected) temperature of 20C. In other words, when you disconnect the coolant temp sensor the ECM is aiming for a slightly leaner mixture. This matches up with your description of it running really rich. With outside temps lower than 20C, the fuel mixture might be too rich to ignite. At the assumed 20C, it may very well be just lean enough.

 

There are a bunch of possible causes. The FSM's diagnostic procedures include fuel pressure and injector tests pretty high up on the checklist. Both of those items could easily cause a run-rich situation. Keep in mind gasoline goes bad after a while, and moving the truck breifly last weekend may have been just enough to stir sediment up. It could also be timing, or the MAF, amoungst a few other things.

 

All that said, have you taken the time to check the basics? With a truck that's been sitting for a while I would check/replace fuel/air filters, and make sure squirrels aren't nesting in the intake to start with. I'd then check for ECM codes, fuel pressure, test the EGR valve (should be closed at start/warm-up), test the PCVs, and the plugs/wires/coil for good spark.

Edited November 2, 2013 by SkiMachine

[Baxta]

Thanks for the response, SkiMachine.

 

The thing is, I can get it to start with the connector hanging, so the basics are good. At the assumed 20C, I get enough air, fuel, spark all timed together for it to run, albeit roughly. I think I can rule out bad gas on this basis as well.

 

To summarize:

 

Connector hooked up to temp sensor = crank, no start

Connector hanging = start, rich running condition

 

I'm going to start fresh this morning and trace the wires on the connector - they seemed pretty brittle, and could be cracked, and maybe there are bare wires rubbing, but that would likely be the case if the connector was hooked up or not, right?

 

One more tidbit - after it starts with the connector disconnected, I can reconnect it, and it will start ok, but it runs like it does while disconnected.

 

Just bought another vehicle last weekend - all I want to do is get this thing running well enough that I can put it up on Craigslist and sell it!

 

 

Thanks for the help!

 

 

Greg

[Kingman]

Try this. Disconnect the battery to reset the ECU, install the temp sensor and plug it in, then try and start it. If it doesn't start after a lot of cranking, don't do Anything before checking the ECU for codes. By doing that you're giving the ECU a fresh start with everything plugged in to record accurate codes and give a better chance of telling you what's wrong.

[Baxta]

Trying that right now, Kingman.

 

The only code I pull is 34 - knock sensor.

 

This morning, I went to start it, and it fired right up. :weird: sounded, felt like it was running smoothly, so I thought I'd take my chances on passing emissions... Failed on all three criteria - NOx, CO, & HC on the driving test... I can't imagine the KS causing all this.

 

 

Getting ready to throw in the towel, but I do appreciate all help & suggestions!

 

 

Greg

[SkiMachine]

The coolant connector/sensor is probably a red herring. It's a symptom of the problem, but not likely the problem. That you can plug it back in after the truck has been running makes sense given the coolant temp will rise very quickly once started. I also doubt the knock sensor would be causing problems at idle. My understanding of the KS code is that it's triggered once the ECM decides the readings are "unreliable", i.e. open or short circuited.

 

Be sure to check the fuel pressure. At idle it should be at 34psi, and at WOT it should be at 42psi. The insidious thing about the fuel pressure regulator is that it's at rest (engine off) target is 42psi. This is because it takes engine vacuum to pull it's diaphragm down to the 34psi target, and is also why it goes to 34psi at WOT (no throttle restriction = little vacuum). It is quite possible - likely even - that the fuel pressure reg got "stuck" at 42psi when the truck was sitting for so long. A vacuum leak would have the same results.

 

Higher than expected fuel pressure would of course mean more fuel into the engine. It'd also completely blow out any emissions tests.

 

edit: The knock sensor is a common problem with our trucks - and an absolute nightmare to replace.

 

edit 2: My pathy had some serious engine issues last spring. I spent a TON of time tracking down various EC sensors/actuators that I thought might be the problem, the whole while ignoring the basics - which is why I think your coolant sensor is a red herring. At the end of the day it was basic valve timing due to a loose crankshaft bolt and a almost completely destroyed index key on the timing gear. I was probably within a week of it completely freewheeling and destroying the engine by the time I figured it out.

Edited November 2, 2013 by SkiMachine

[Baxta]

*That* makes a lot of sense, SkiMachine!

 

Unfortunately, the location of the FPR makes it a difficult diagnosis, let alone changing it out if it is bad. I figure it's worth $1000 if it's aircared (bad front main seal, also leaking oil out of the pan) $400 if it isn't.

 

I don't have the luxury of time (any more!) or space (thanks, Township!) so I think I'll let it go.

 

 

I do appreciate your help!

 

 

 

Greg