Hard to start - hunting & stumbling - only when cold.

Disclaimer: Links on this page pointing to Amazon, eBay and other sites may include affiliate code. If you click them and make a purchase, we may earn a small commission.

DSinOR

Newbie
Joined
Mar 2, 2021
Messages
16
Reaction score
33
Location
Oregon
1996 5.7 Vortec Yukon 4WD, 245K original miles.

Doesn't want to start when cold (if it sits out overnight, 50° or colder temps).

I'm the original owner. We weren't using it much, then we totalled the other car.
So I bought parts and did some preventive maintenance to make it reliable:

New water pump, coolant flush, all new coolant and heater hoses.
Removed fan clutch, installed Windstar e-Fan, controlled by temp sensor in lower radiator hose tee, works great.
Tranny flush, drop pan, change filter.
Oil and filter, chassis lube.
New wix fuel filter (only 8K miles on old filter and delphi fuel pump).
New sensors - IAT, MAP, TPS, IAC.
New PCV valve and tube and grommet and vacuum fitting at manifold.
Cleaned MAF sensor.
Cleaned throttle body.
New spark plugs and cables and new Ign coil.
New CKP sensor.
New aluminum body distributor with new CPS. I used DashCommand to set cam retard value with rpm at 1100 on warm engine. Then I got misfire codes at high rpm, so I did the [warm engine-neutral gear-parking brake on-rev over 4K for 30 seconds] thing for relearning crank. It seemed to work, no more codes.
New headlamps and running lights fixtures with two relays switched by the green and tan wires for direct power feed to headlamps, plus a diode between the green and tan for 4-High mod.
New tires.

The engine is strong, gets 17mpg on the highway at 70mph, sounds good, feels good, starts with one crank when cool or warm, and there are no codes.

But if you let it sit out all night in cold weather, it doesn't want to start or idle.
It hunts and stumbles and idles rough and shakes the car.
After several attempts, it eventually find itself and then it runs fine.

I can't say whether it was doing this before the PM work, because it was garage kept in a climate-controlled garage.
We didn't discover it until we started using it and sometimes leaving it out in the cold.

I tested fuel pressure on warm engine: 60psi with engine-off key-on. 52psi while idling. If you turn off the engine, the pressure actually climbs to 56psi and holds there for 2 minutes before it starts dropping very slowly and i got bored and quit watching after 5 minutes, it was still over 50psi.

I changed the intake manifold gasket at 100K, but i've never changed the oem spider.

I have checked for vacuum leaks with spray water, spray starter fluid, propane, and just listening. I can't find any vac leaks.

Got any ideas?

Thx.
 

Schurkey

Supporting Member
Supporting Member
Joined
Jan 14, 2018
Messages
10,963
Reaction score
13,741
Location
The Seasonally Frozen Wastelands
Does the fuel pump prime for 2 seconds when the key is turned from "Off" to "Run"?

What is the fuel pressure, cold?

What is the battery voltage, cold?

Connect a scan tool, verify ALL sensors, especially coolant temperature.
 

DSinOR

Newbie
Joined
Mar 2, 2021
Messages
16
Reaction score
33
Location
Oregon
Thanks for the reply.

Years ago I modified the wiring under the underhood fuse box: I connected the fuel pump relay to pink. If the key is on, the fuel pump is running.

The low last night was 40°.
Cold battery voltage - 12.55.
Cold fuel pressure this AM: 60 key on, 53 running, 49 off, holds steady for at least 2 minutes.

It stumbled and hunted for 3 or 4 seconds when I started it. Then it smoothed out.
I could see the fuel pressure gauge, it bounced around a bit but it stayed above 50psi the whole time.

For the last few weeks, I've messed with DashCommand and Torque while someone else was driving.
Every number I can find looks normal.
I don't have any other scan tool.
The only odd thing I have seen is that fuel trim seems to be about 5ish on one side and about 12ish on the other.
 

Schurkey

Supporting Member
Supporting Member
Joined
Jan 14, 2018
Messages
10,963
Reaction score
13,741
Location
The Seasonally Frozen Wastelands
1. Isn't 60 psi marginally low for a Vortec 5.7L?

2. I don't know DashCommand or Torque, but what sort of readings are you seeing when the vehicle first starts, cold?

3. Having the fuel pump running if the ignition is on is DANGEROUS, you need to return to the stock wiring. Imagine: You're in a collision that ruptures the fuel plumbing. You're knocked unconsious--or worse, totally awake but paralyzed/stunned or otherwise unable to shut off the ignition. The fuel pump will pump the tank dry, loading the area around the vehicle with gasoline. You and your family burn alive. With stock wiring, the fuel pump stops when the oil pressure drops AND the computer is no longer getting an RPM signal from the crank sensor.
 
Last edited:

arrg

I'm Awesome
Joined
Apr 29, 2015
Messages
401
Reaction score
312
Location
North Las Vegas, NV
Have you tried cleaning your battery cable ends and your grounds? Especially the fuel pump ground near the tank? Also, even though you think your battery is good, you could try hooking up the ole jumper cables to another vehicle the next time you have a cold start. Might help rule out electrical problems as the culprit.
 

DSinOR

Newbie
Joined
Mar 2, 2021
Messages
16
Reaction score
33
Location
Oregon
Spoke to the local dealer, twice.
Pleasantly surprised by the willingness to help and also the on-tap knowledge of 5.7 vortec trucks.

For posterity:

A - my prob is a rough cold start, only on cold mornings. It fires right away, but it hunts and stumbles until it finds itself, then no problems. No codes. If the low is 60°, the problem does not occur. it has to be 45° or colder.

B - Good ideas posted here already.
For Schurkey:
1 - IMO, no. Many internet threads over many years claim that any number less than 62psi key-on, or less than 55psi idle, or less than 58psi throttle-blip means that the fuel pump or regulator is bad. But many other threads suggest that when people see slightly lower numbers and therefore replace the fuel pump and regulator, they get the exact same numbers after the replacement. The dealer guys I talked to said my fuel pressure was fine and moved straight to other possible causes.
2 - Torque Pro and DashCommand both provide realtime ecm sensor measurements with histogram. On cold start, my IAT reads actual ambient temp accurately, the coolant temp sensor reads ambient coolant temp accurately and rises on a predictable scale, the coolant temp sender causes the dash temp gauge to read accurately, the oil pressure sensor and gauge behave normally, the MAF sensor delivers air flow volume measurements that fit within specs and fluctuate as expected with changes in rpm, the TPS is accurate, the IAC valve (new) apparently opens and closes when commanded just like the old one was doing before I replaced it because it was 25 years old, the MAP sensor delivers readings that are in spec and fluctuate as expected with changes in rpm, etc. Everything seems to be nominal.
3 - nope. I much prefer to have a fuel pump running when the ley is on over having a fuel pump that can be interrupted by a defective sensor or wiring short or defective ecm. To each his own. Also, I'm pretty sure your AND should be an OR.

For arrg:
1 - Yes thanks. Battery cables had dielectric grease film applied in 1996. Still clean. There are many helpful posts on how to clean and service all the ground connections. Over the last two months, I checked and cleaned and lightly lubed all of them. I'm in the high desert. They were all clean. I tested several near the battery that showed no ohms between the ground connection and the negative terminal.
2 - Great idea! Can't believe I didn't think of it. If cold weather drops battery voltage down to a marginal level, jump it with a 14v supply and see if the cold start problem disappears. Great idea. I'll report back on this one.

From the dealer service guys:
1 - Carbon buildup under the intake manifold can result in small vac leaks that seal up after a short spell of running due to vacuum sucking things tight or temperature related expansion of fit tolerances or dirty valves heating and beating themselves to a better state. These problems can be hard to find. Run several tanks of gas with quality Fuel Injector cleaner. Or, try this crazy quickie trick: find a way to fog some fuel injector cleaner straight into the air intake. See Post#9 below.
remove the passenger side hose from valve cover to the nipple on the air intake just above the throttle body. Fab up a clear bottle with a drilled cap with a hose stuffed into the hole in the cap. Fit the hose to the nipple on the air intake. Fill the bottle with fuel injector cleaner. Start the engine. Attach the bottle to the intake and invert it. Basically, let the engine suck the FI cleaner into the manifold. Crazy. I've never heard of such a thing. I did it. It took about 7 minutes for the engine to suck a small bottle of FI cleaner. For me, it was about 5 minutes of idle plus 2 minutes of revving to about 2000 rpms. I had to fiddle with the bottle a couple of times, burping it so the fluid would keep flowing.
The whole time, I kept listening for any kind of different sound. I never heard anything at all. The engine sounded totally normal the whole time. Weird. I also dumped a bottle in the fuel tank and sloshed it pretty good with abrupt forward-brake-reverse-brake motion of the vehicle. We'll see...

2 - We talked about fuel trim. My ST and LT Bank 1 trim is a little different than my ST and LT Bank 2 trim, according to both apps. Basically, the trim graph says that the ecm is deciding to fuel the left cylinders differently than the right cylinders. They should be about the same. Possible causes:
a - vac leak. If the vac leak is bad when cold and not so bad when warm, you'll have the hunting problem when cold and then slightly different trims when warm. Maybe carbon deposit cleanup will fix it.
b - get a log of the voltage signals from the upstream O2 sensors. Both apps will do this. If one sensor is much quieter than the other, then you may have a bad O2 sensor which could cause the ecm to have problems in determining how to properly fuel that bank, which could cause the cold stumble and a permanent difference in warm trims.
c - the ECM's on these vehicles can drift bad over time. The dealer guys didn't think my prob was an ECM prob, but they said it was possible.

I'll mess with it for a few days and report back.

Thx.
 
Last edited:

TreeGeared

I'm Awesome
Joined
Apr 18, 2022
Messages
112
Reaction score
194
Location
Ohio
You have the right idea using torque or another program. When I built my new engine I had a similar problem with unbalanced fuel trim. That turned out to be caused by an intake gasket seal that I was unable diagnose otherwise. Could also be on the injection side especially if the vehicle sat. You may have a clogged popit valve on the spider.
 

Schurkey

Supporting Member
Supporting Member
Joined
Jan 14, 2018
Messages
10,963
Reaction score
13,741
Location
The Seasonally Frozen Wastelands
For Schurkey:
1 - IMO, no. Many internet threads over many years claim that any number less than 62psi key-on, or less than 55psi idle, or less than 58psi throttle-blip means that the fuel pump or regulator is bad. But many other threads suggest that when people see slightly lower numbers and therefore replace the fuel pump and regulator, they get the exact same numbers after the replacement. The dealer guys I talked to said my fuel pressure was fine and moved straight to other possible causes.
Pump pressure varies with available voltage through the pump motor windings. That is, the quality of the wire harness before the pump, as well as the ground side.

Low pressure can be a failed pump. It can also be failing wiring. Replacing the pump when it's the harness that's causing the problem would result in the new pump producing the same pressure as the old one.

It's vitally important to verify the wire harness (voltage and amperage delivered to the pump, with the pump running; along with voltage on the ground side) BEFORE replacing the fuel pump. And I replace the in-tank harness along with the pump.

2 - Torque Pro and DashCommand both provide realtime ecm sensor measurements with histogram. On cold start, my IAT reads actual ambient temp accurately, the coolant temp sensor reads ambient coolant temp accurately and rises on a predictable scale, the coolant temp sender causes the dash temp gauge to read accurately, the oil pressure sensor and gauge behave normally, the MAF sensor delivers air flow volume measurements that fit within specs and fluctuate as expected with changes in rpm, the TPS is accurate, the IAC valve (new) apparently opens and closes when commanded just like the old one was doing before I replaced it because it was 25 years old, the MAP sensor delivers readings that are in spec and fluctuate as expected with changes in rpm, etc. Everything seems to be nominal.
What are the misfire counts when the engine is cold?


3 - nope. I much prefer to have a fuel pump running when the ley is on over having a fuel pump that can be interrupted by a defective sensor or wiring short or defective ecm. To each his own. Also, I'm pretty sure your AND should be an OR.
My "AND" was an "OR" before I edited the post to correct it.

Your vehicle would not lose fuel pressure unless BOTH the oil pressure AND the RPM signal failed. If the RPM signal fails, the pump still runs on the oil pressure sensor bypass wiring. If the oil pressure sensor fails, the pump still runs on the fuel pump relay side of the harness. They'd have to both fail to stop the pump.

The chances of both failing are quite remote. The chances of a collision or other accident are much higher.


For arrg:
1 - Yes thanks. Battery cables had dielectric grease film applied in 1996. Still clean. There are many helpful posts on how to clean and service all the ground connections. Over the last two months, I checked and cleaned and lightly lubed all of them. I'm in the high desert. They were all clean. I tested several near the battery that showed no ohms between the ground connection and the negative terminal.
An ohmmeter is not an accurate way of testing those cables. You would need to do voltage-drop testing to accurately assess the condition of the cables and connections.

From the dealer service guys:
1 - Carbon buildup under the intake manifold can result in small vac leaks that seal up after a short spell of running due to vacuum sucking things tight or temperature related expansion of fit tolerances or dirty valves heating and beating themselves to a better state. These problems can be hard to find. Run several tanks of gas with quality Fuel Injector cleaner. Or, try this crazy quickie trick: remove the passenger side hose from valve cover to the nipple on the air intake just above the throttle body. Fab up a clear bottle with a drilled cap with a hose stuffed into the hole in the cap. Fit the hose to the nipple on the air intake. Fill the bottle with fuel injector cleaner. Start the engine. Attach the bottle to the intake and invert it. Basically, let the engine suck the FI cleaner into the manifold. Crazy. I've never heard of such a thing. I did it. It took about 7 minutes for the engine to suck a whole bottle of FI cleaner. For me, it was about 5 minutes of idle plus 2 minutes of revving to about 2000 rpms. I had to fiddle with the seafoam bottle a couple of times, burping it so the fluid would keep flowing. The whole time, I kept listening for any kind of different sound. I never heard anything at all. The engine sounded totally normal the whole time. Weird. I also dumped a 2nd bottle in the fuel tank and sloshed it pretty good with abrupt forward-brake-reverse-brake motion of the vehicle. We'll see...
I have no faith in Seafoam for cleaning hardened carbon. Seems to be OK for stabilizing gasoline for storage.

GM used to--and maybe still does--sell Techron Complete Fuel System Cleaner (in a bottle with an AC-Delco label) for cleaning the fuel system including injectors.

2 - We talked about fuel trim. My ST and LT Bank 1 trim is a little different than my ST and LT Bank 2 trim, according to both apps. Basically, the trim graph says that the ecm is deciding to fuel the left cylinders differently than the right cylinders. They should be about the same. Possible causes:
a - vac leak. If the vac leak is bad when cold and not so bad when warm, you'll have the hunting problem when cold and then slightly different trims when warm. Maybe carbon deposit cleanup will fix it.
b - get a log of the voltage signals from the upstream O2 sensors. Both apps will do this. If one sensor is much quieter than the other, then you may have a bad O2 sensor which could cause the ecm to have problems in determining how to properly fuel that bank, which could cause the cold stumble and a permanent difference in warm trims.
c - the ECM's on these vehicles can drift bad over time. The dealer guys didn't think my prob was an ECM prob, but they said it was possible.
Can't be the O2 sensors. At cold startup, they're not active because they have to heat up before they start working. After the computer is running in closed-loop, the O2 sensors might cause fueling problems.

Have you ever checked the distributor cap for condensation? How about the ignition coil for dampness? Either one could cause misfire, especially when cold before the moisture evaporates.

Run the engine in total darkness. Look under the hood for glowing spark plug wires. Corona discharge means the plug wires have faulty insulation, which may or may not lead to actual sparking to ground.
 
Last edited:

DSinOR

Newbie
Joined
Mar 2, 2021
Messages
16
Reaction score
33
Location
Oregon
Two days later:

1 - don't add FI cleaner into the previously described port on the throttle-body cap that receives the hose from the passenger side valve cover. Here's why: I put 4oz of seafoam into that port on Wednesday, as described above (bottle with hose). Thursday I found about 1oz dribbled backwards towards the air filter. So I had to clean the intake hose assembly and MAF sensor. However, I also pulled the PCV vacuum fitting and the MAP sensor and they looked like they had been fogged with seafoam. They wiped clean easily, leaving a brown film on the rag. I suspect that 3oz of seafoam were fogged into the manifold.

2 - While the intake hose was off, I discovered that you can drip FI Cleaner into the drivers side of the rectangle slot at the rear of the throttle body while the engine is idling and you can hear it get sucked in. I also dripped a teaspoon onto the throttle body plate.

3 - The engine started instantly and smoothly Thursday and Friday morning. It was about 40°.

The dealer guys I talked to said that full strength FI Cleaner fogged directly into the air intake will soften carbon buildup on the IAC valve and the intake valves and seats (and everything else under the manifold), sometimes curing a vacuum problem caused by excessive carbon buildup. About 12 years ago, at about 125K miles, I did the intake manifold gasket job and replaced the regulator and cleaned everything. It hasn't been opened up since then. If you've ever done that job, you know that an amazing amount of carbon builds up on everything under the manifold.
So maybe this idea of Redline or Seafoam or Lucas or Techron FI Cleaner fogged into the manifold is a good idea. Time will tell. Seems like it can't hurt because seafoam is safe for adding to gasoline and engine oil. They even sell a manifold fogging kit now for spraying into the intake air stream.

Before the end of May, we'll have several nights in the 30's. I'm leaving the vehicle outside every night to see if the problem returns. I'll probably fog it again once per week for a few weeks with 2oz dripped into the rectangular slot of the throttle body.

I think it is worth noting here that my local dealer was helpful, and that they felt like my fuel pump pressure was fine, and that they knew I was running the factory spider and weren't worried about that at all. If I did have blocked or partially restricted poppets, a few tanks of gas with FI Cleaner would solve that, which is something I've always done twice a year anyway.

I'll post an update in a few weeks.

I'll get a log of the trims again Monday.

I replaced two wires from the engine bay to the fuel pump, back when I changed the fuel pump wiring. About 7 years ago?

I don't have any misfire codes. I only had those for one day after replacing the distributor, CPS sensor, and CKP sensor. They were resolved by the 4000rpm relearn sequence.

Before this year, I changed the cap and rotor three times. I never had any indication of moisture under the cap. The rotor just sort of gets burned away. Again, I live in the high desert. Relative humidity here is low. As described above, I just replaced the entire distributor, plugs, wires, and coil. If the hard cold start problem returns, I'll do the dark test.

This vehicle never sat. It's always been used between 3 and 30 days per month.

Treegeared - thanks for the info on your issue and trims. I'll keep working towards having balanced trims.

Thanks.
 
Last edited:

alpinecrick

I'm Awesome
Joined
Jun 12, 2015
Messages
1,614
Reaction score
1,695
Location
Western Slope of Colorado
I'm assuming the truck spends the night outside.
You live in Oregon. High humidity unless you're on the Dry Side of the state. Classic symptoms of moisture in the distributor cap. I'm sure somewhere in the Galaxy there are aftermarket, aluminum body distributors that work, but......

AC Delco or Delphi distributor, cap, and rotor, for reliability. Yes, there are the NAPA Echlin/Blue Streak caps with brass contacts that are good, but have a shorter lifespan. They also can come with a vent hole that allegedly vents ozone out, but of course lets moisture in (I'll catch flak for this paragraph).

Also the weather strip that goes over the cowling over time can become compressed and leak--almost invariably right on top of the distributor. Murphy's Law.

I might be wrong, but on a wet morning pull the distributor cap and take a gander.
 
Top