85 DeVille 4100 won't start when hot

[misterc9]

My friend has a nice 85  DeVille front wheel drive 4100.It has 66,000 miles, nice shape no rust ruins great etc ,used only in summer and kept in a nice dry attached garage. Has regular maintence The problem is on a warm day after driving around with the AC on if they stop in the store for ten or twenty minutes it wont start when they come out. The starter cranks good but no fire. Wait twenty to thirty minutes and zoom. Opening the hood to cool makes it start sooner. After this  problem  started  it has had a rebuilt distributor with a module and pick up coil , a new rotor and an ignition coil. The battery is a year old. We thought it was losing the spark when hot but no one has been actually able to test it when it wasn't starting . By the time someone gets  there to look it would start. The AC doesn't  have to be on just need a good warm day. They do most all their driving in the afternoon in the sun. This started the middle of last summer . The problem went away in the fall but is back now. Any ideas any one?

[joe_padavano]

10 hours ago, misterc9 said:

 Any ideas any one?

 

Get a scan tool and monitor what is going on when this happens. Obviously there is a thermal issue with the electronics.  The question is, is the problem ignition or fuel?  Your 4100 uses the GM throttle body fuel injection, so there isn't a crank sensor to worry about.  The distributor module can fail intermittently due to heat, so that's one place to start.  Also look into the coolant temp sensor.

[JACK M]

Sounds like a vapor lock to me.

[Digger914]

Deja vu

This sounds like the problem with the new formula of summer fuel blend and 1st generation FI from way back in the early 90's. Back then the symptoms were; you turned it off warm and started it again within a few minutes no problem, go back a few minutes after that it was hard to start, a few more minutes and you had to put your foot to the floor to start, if it would start and at 30 to 40 minutes after you turned it off, you turned the key and it started right up.

No one I know ever found a cure for the problem and a couple years later the summer starting problem went away. We did find a few different things that helped, like cleaning the heck out of the throttle body including the mas airflow sensor and changing gas brands with every tank full until you found one that gave you the least trouble.

[pfloro]

Here is a very easy way to isolate your no start problem to either ignition or fuel:

 

Borrow or buy an inexpensive timing light. An inductive type clamps over the #1 cylinder's spark plug wire (at any place along the wire).

When you experience the no start, connect the timing light and crank the engine.  No flash = no spark...

 

Inductive Timing Light - SEARS

 

I used this same technique to diagnose an intermittent no hot start on my 1984 Toronado (307 V8 with Quadrajet and CCC). The problem turned out

to be an defective pickup coil in the distributor. 

 

Keep us posted.

 

Paul

 

PS:  I had recently replaced the distributor with a rebuilt unit.  I assumed the pickup coil had been replaced.  Perhaps it was replaced but it failed.  (?)

Edited June 6, 2016 by pfloro (see edit history)

[misterc9]

Progress!   We finally got the car to not start while it was in the shop. It has spark. We let it run till really warm then turned it off for twenty minutes. It will start for a second then cut out but having spark while cranking. A little gas down the throttle and zoom for a second.  We put the fuel pressure tester on but then it started. It has 12 lbs  normal for this car. There is not much time to test because once the hood is open it cools off and starts. Tomorrow  will try again. The fuel pressure tester is hooked up. We want to see if it has pressure during no start cranking so we can further isolate the problem.  Thanks to all those offering suggestions. Will keep you posted of any progress.

[joe_padavano]

On 6/6/2016 at 0:39 PM, JACK M said:

Sounds like a vapor lock to me.

 

I'm still waiting for someone to explain how an EFI motor with an electric pump in the tank and a fuel return circuit can vapor lock.

[misterc9]

Progress ?  Got it to not start again today. Has 12 lbs fuel pressure during a no start crank. Put the noid light on the injector plugs. Both flashed good one at a time during a no start crank. This car starts good when cold or semi warm. Starts right up after sitting for a week or two. Could the injectors not wok when hot?  Has anybody had any luck with injector cleaners?   This is puzzling.

[Digger914]

Injector cleaner does work, but most good fuels have cleaning additives, so the results might not be noticeable.

Two questions, one electrical the other fuel.

When warm and not starting, does unplugging the MAS get the car to start?

When warm and not starting, will it start by holding the throttle to the floor while cranking?

 

 

[pfloro]

Even though you have 12PSI fuel pressure in the fuel supply line, we don't know if any fuel is actually spraying from the injectors.  

Seeing the test light flash at least indicates that the injector solenoids are receiving pulses from the ECM.

I've never done this, but I understand you can use a timing light to illuminate (and freeze the motion of) the actual fuel spray from

the injectors.  When you start the cool engine, direct the light as close as possible to each injector tip. Then observe the spray pattern.

It should be an even pattern like a wide shower head.  Then do the same thing during the no start crank. 

 

It may be time to overhaul the TBI unit.  Evidently, they need attention when they age (time & mileage).  Don't we all...!

 

Not having owned a TBI vehicle, I'm not an expert.  However, IMO, even with bad sensors, if you have any fuel spraying during a warm crank,

the engine should at least attempt to start.  I think your no start condition presents an easier problem to troubleshoot.

 

Continued good luck,

 

Paul 

 

Edited June 8, 2016 by pfloro (see edit history)

[misterc9]

This caddy doesn't use a mass airflow sensor.  Stepping on the gas doesn't seem to have any effect on the no start. Next I 'm going  to try to watch the injector spray on a normal start.  If I can see it then I'll try to watch it during a no start.  This is slow going because it takes quite a while to get a no start. When it does get a no start there is not much time to work on it before it starts again.

[Digger914]

The car didn't start holding your foot to the floor and as you didn't mention smelling raw gas, I'm going to guess that you weren't getting any. Unplugging a MAP isn't as educational as the MAS, it will change fuel air and if do get gas you won't know why.

[cahartley]

The next time you experience a no start dump some gas down the throttle body and try it.

If it starts you'll be a lot closer to getting to an actual fix.

[joe_padavano]

3 hours ago, cahartley said:

The next time you experience a no start dump some gas down the throttle body and try it.

If it starts you'll be a lot closer to getting to an actual fix.

He did and it does. See post #6.

[JACK M]

23 hours ago, joe_padavano said:

 

I'm still waiting for someone to explain how an EFI motor with an electric pump in the tank and a fuel return circuit can vapor lock.

 

My bad. I am old and don't do modern injection.

I was only making my observation. (if it were carbureted)

[joe_padavano]

27 minutes ago, JACK M said:

 

My bad. I am old and don't do modern injection.

 

You do realize that the design of the TBI system on that car is nearly 40 years old.

[pfloro]

16 hours ago, misterc9 said:

This caddy doesn't use a mass airflow sensor.  Stepping on the gas doesn't seem to have any effect on the no start. Next I 'm going  to try to watch the injector spray on a normal start.  If I can see it then I'll try to watch it during a no start.  This is slow going because it takes quite a while to get a no start. When it does get a no start there is not much time to work on it before it starts again.

 

While in the shower this morning, something occurred to me.  Even though the test light indicates that power is being applied to the injectors, there may be enough resistance in that circuit to prevent sufficient current from flowing. 

If this is the case, the test light might flash but the pintle valves in the injectors might not open enough to allow fuel to spray.   The injector solenoids 'pull' more current than the test light.

 

In the period when your car was made, GM wasn't using gold plated contacts in it's connectors (maybe they never switched over). 

 

To rule out this possible problem at the ECM, disconnect the battery, disconnect & reconnect (more than once) the big connectors at the ECM.  The ECM may be located behind the kick panel in the passenger side foot well. 

If there is any oxidation on these connectors (female side) or ECM circuit board (male side), it will be removed by the physical R&R of the connectors.  This is easy to do and will rule out any resistance at the ECM connector point. 

FYI:  there should not be any additional connections in the CCC wire harness (injector circuit) between the ECM and the injector connectors.

 

I've run into this situation before with sensors in the CCC system on my '84 Toronado. Although it didn't have TBI, it did have a computer controlled Quadrajet (mixture control solenoid and TPS).

 

Paul

 

PS:  Whether or not above suggestion corrects your no warm start problem, it would be a good idea to carefully 'break & make' other connectors in the CCC system.

 

PPS:  CCC = Computer Command Control (GM's marketing name, etc.)

 

 

Edited June 8, 2016 by pfloro (see edit history)

[joe_padavano]

Paul makes an exceptional point.  The 1980s were not GM's finest hour, and the electrical systems especially suffered.  I've seen far too many problems caused by inadequate or undersized electrical terminals.  Older models without digital electronics were insensitive to resistance increases as the unsealed Packard 56 terminals got dirty. The early GM digital systems are quite sensitive to these resistance buildups.  Once GM transitioned to Weatherpack and Metripack sealed terminals, the problems decreased dramatically, but the 1980s were a transition period.

[misterc9]

It was cool today 65 F so I couldn't get any no starts. It looks like I can see the spray on a normal start. On the next no start I'll try to watch the spray.  pfloro  Disconnecting the computer connectors sounds like a good idea I'll try it tomorrow. Digger MAS  Are you talking about the manifold air temperature  sensor?  I did  a scan last summer and I believe all the different temperature readings seemed right both hot and cold, but that was last summer. So it might be worth looking at again with a fresh scan. I'm going to try disconnecting it to see if it has any effect. As always Thanks for the help, comments and suggestions.

[Chris Coon]

I can remember seeing an air temp sensor intermittently read very hot (or cold) or possibly a coolant  temp sensor doing the same. If either of these were to read an extremely high temp (inaccurately) the computer would Lean out the fuel, maybe enough to cause a no start. 

 

Just a thought

[Digger914]

When you shut down a warm engine the coolant stops circulating and engine gets hotter before it gets cooler. When I wrote MAS I meant mass airflow sensor, it does a lot more electrically than a MAP does and you need to unplug several sensors to get the same affect as you get by unplugging that one. That's the good thing about the MAS metered systems, bad thing is they cost considerably more than a MAP and any other single sensor if it turns out to be the cause. Unplugging the air and coolant sensors will make your controller think you are going for a cold start and it will richen up your fuel mix. Good thing about the old MAP systems is that if you forget to turn the key to off before unplugging sensors, you're less likely to fry something.

 

Gas is regionally blended and varies around the country, living in a corn growing ethanol producing state we were afflicted with gasohol and the odd little problems that came with a long time ago. And this sounds a lot like our summer fuel, early fuel injection, no start when warm problems from the early 90's that went away after a couple of years It wasn't just GM cars and we never did find a sure cure for the problem.

[JACK M]

21 hours ago, joe_padavano said:

 

You do realize that the design of the TBI system on that car is nearly 40 years old.

 

I am way older than that.

In my trade they didn't make anything with injection until 1997, That is the year that I retired.

Give me an old Evinrude and I am right at home.

Anything I have learned about the old cars, (like vapor lock) came from the school of hard knocks.

To me an 85 Cadillac is just an old car that doesn't really interest me.

Sorry I spoke my suggestion.

[misterc9]

It was cool again today 65F . Couldn't get any no starts. This car doesn't use a Mass airflow sensor,  MAF. Won't be able to make any progress until it gets back up to 75-80F or more.

[misterc9]

I almost believe the gasoline is vaporizing in the throttle body or in the injectors. It turns out the gasoline is mostly from last winter. My plan is to pump it all out and replace it. The tank is nearly full. Any opinions?  Which would be less likely to vapor lock?  Regular or premium, Gasohol or alcohol free?  What do you think about  adding 10% Diesel to the new gas? I've heard the Diesel will raise the vaporization temp of the gasoline a bit.

[Digger914]

Alcohol free if you can get it.

[joe_padavano]

11 hours ago, misterc9 said:

I almost believe the gasoline is vaporizing in the throttle body or in the injectors. It turns out the gasoline is mostly from last winter. My plan is to pump it all out and replace it. The tank is nearly full. Any opinions?  Which would be less likely to vapor lock?  Regular or premium, Gasohol or alcohol free?  What do you think about  adding 10% Diesel to the new gas? I've heard the Diesel will raise the vaporization temp of the gasoline a bit.

 

It is not possible for an EFI system with an electric pump in the tank and a fuel return line to have vapor lock problems.  The fuel is constantly flowing through the return back to the pump. There is no chance for vapor to form, as there is no volume of fuel in the throttle body the way there is in a carb.  This is not your problem.  More likely is that you have an electrical sensor with a thermal sensitivity issue that sends spurious data to the ECU when temps are hot.

 

I think adding diesel is a really bad idea.

[pfloro]

I agree with Joe Padavano. Vapor lock is not causing the hot no-start problem. At this point I would rebuild the TBI unit. Considering the age of the vehicle, it's a good idea. I don't necessarily believe this will solve the problem but let's rule it out.

 

Did you 'break & make' the ECM connections?  Do this first...!

 

If the problem persists, find a shop which can examine the actual 'shape' of the pulses being delivered to the injectors during both good start & no start conditions. This is not rocket science. I'm starting to suspect that the ECM is flaky. I don't think sensors are causing this problem because during cranking, I believe the ECM supplies fuel & ignition values which are fixed. Once the engine starts, open loop values are used until the O2 sensor heats up. Then closed loop (feedback) operation begins.

 

My ace mechanic used simple software & hardware on his laptop to diagnose these types of problems on both TBI and timed sequential multi-port fuel injection. 

 

Paul

 

 

Edited June 22, 2016 by pfloro (see edit history)

[joe_padavano]

52 minutes ago, pfloro said:

Did you 'break & make' the ECM connections?  Do this first...!

 

I'll also suggest a squirt of electrical contact cleaner before remating the connectors.

[misterc9]

It hasn't been really warm the last few days  I did unplug the three connectors from the ECM. They all look nice and I sprayed them with CRC electronic contact cleaner. I was only able to get one brief no start out of it today. It only lasted a few minutes so I didn't get to do much with it. I have the scanner hooked up to it hoping to maybe get a weird reading on something, maybe the MAT or the coolant temp. If it gets hot out  tomorrow maybe I can do something. My common sense tells me it's not vapor locking but if it wasn't impossible it would be a good explanation. Someone suggested cutting a hole in the hood above the throttle body and installing a spiral roof vent. I'm convinced the problem is under the hood because it will go away so quickly when you open it. Leave the hood shut and it won't start for a longer time.

Edited June 28, 2016 by misterc9 (see edit history)

[TerryB]

If it's any consolation my 2013 Dodge minivan doesn't like to run smooth at start up in hot weather.  Acts like it's running too rich.  Dealer can't find the problem as it gets better after a few minutes of running.

[pfloro]

12 hours ago, misterc9 said:

I'm convinced the problem is under the hood because it will go away so quickly when you open it. Leave the hood shut and it won't start for a longer time.

 

Since you've cleaned the ECM connectors and because of the above quoted information, I would now rebuild the TBI unit.  I have read that over time the spring(s) within the pressure regulator (part of the TBI unit)

can weaken due to heat.  Although you have checked the fuel pressure when hot, who knows what is really going on inside the TBI unit.

 

It feels like you are getting very close to solving this problem...

 

Paul

[misterc9]

Thinking the problem was old stale gasoline I hooked up a steel line to the return outlet on the throttle body. I put the other end into an eighteen gallon jug. I let the car run and it had good fuel volume out of the return line. Did this until it ran out of fuel. Then I put in ten gallons of alcohol free super and one gallon of Diesel fuel. They have had their car back for almost a month. We have had lots of hot days and so far so good. That eleven gallons will last them a couple of months. So who knows? Thanks to all those who responded with ideas and suggestions. I'm still keeping my fingers crossed.