2002 Suburban Stalls while driving
June 20th, 2021, 12:57 PM
#22
Yep; I've always lived by that. I have a good friend who years ago was responsible for a large fleet of cop cars, all GMs, and they had horrid problems with fuel pump failures until they started INSISTING that nobody let them get below 1/4 tank, with the target being 1/2 full or better. The failures stopped. Obviously there's a heat-dependent problem there but what he reported wasn't intermittent but rather complete, usually no-warning failures (those are fun when you're 100 miles from nothing and in a truck they're double-fun since there's no hatch and you have to drop the tank -- not exactly something you can fix on the side of the road!)
June 21st, 2021, 8:27 AM
#23
One of the issues with buying used vehicles, don't know if previous owners played E roulette and beat up the fuel pump.
I never had a fuel pump issue with mine or family vehicles purchased new, only used ones.
I've taken multiple vans on my job to 250k on the original fuel pumps with the 1/4 tank rule.
I never had a fuel pump issue with mine or family vehicles purchased new, only used ones.
I've taken multiple vans on my job to 250k on the original fuel pumps with the 1/4 tank rule.
June 21st, 2021, 9:12 AM
#24
True, but I'm the original owner of the '02...... the gas gauge has read 1/4 when actually empty for 10+ years. Makes for an easy "don't get low on fuel" paradigm since you never really know where empty is!
And the fuel pressure at the rail performs as expected. And yeah, I know, on the kid's van she was no way to know.
All I got at this point, however, with neither the tach going to zero or the oil pressure going to zero when it happens (e.g. sensor failures that would legitimately cause the PCM to command the pump off) is the pump itself, since there's no fuel control module in the ones from these years, the relay has been swapped without change and the contacts show no evidence of heat damage (e.g. problem with the electrical distribution box.) So out the tank comes today on the '02 whether I like it or not. At least in my case I should get a properly working fuel gauge out of the deal for the first time in a decade. Unfortunately with no PCM pressure readout -- the regulator is mechanical at the rail on these engines and there is no sensor I can find -- nor any PID I can find for the commanded state of the pump from the PCM I have no realistic means of knowing, and with it not being reproduceable except at altitude and high temperature both my attempts to provoke it here in eastern TN have failed.
And the fuel pressure at the rail performs as expected. And yeah, I know, on the kid's van she was no way to know.
All I got at this point, however, with neither the tach going to zero or the oil pressure going to zero when it happens (e.g. sensor failures that would legitimately cause the PCM to command the pump off) is the pump itself, since there's no fuel control module in the ones from these years, the relay has been swapped without change and the contacts show no evidence of heat damage (e.g. problem with the electrical distribution box.) So out the tank comes today on the '02 whether I like it or not. At least in my case I should get a properly working fuel gauge out of the deal for the first time in a decade. Unfortunately with no PCM pressure readout -- the regulator is mechanical at the rail on these engines and there is no sensor I can find -- nor any PID I can find for the commanded state of the pump from the PCM I have no realistic means of knowing, and with it not being reproduceable except at altitude and high temperature both my attempts to provoke it here in eastern TN have failed.
Last edited by tickerguy; June 21st, 2021 at 9:16 AM.
June 23rd, 2021, 8:31 PM
#25
So.... dropped the tank and found that while I jumpered it and thus should have bled all the fuel into jerry cans out the test port well.... no. there was north of 10 gallons still in the tank!
So two separate problems, which the new pump appears to have fixed.
First, there was obviously an air leak on the suction side somewhere up high enough that once you got down to about 1/4 tank it sucked air and was done, even though there was fuel.
The second is a design issue from what I can tell. The new pump has two chambers in the cup, the original has one. Here's the hypothesis:
At altitude boiling point is lower, and heat under the hood (IATs, etc) are quite a bit higher. Gas starts to boil off at 95F at seal level, ~90F at 5,000'. The return line goes into the cup in the original design. This means at altitude the return fuel at highly elevated temperature can not circulate into the tank and thus be cooled down, plus the motor of the pump is cooled by the same flow, so it's getting even more heat added to it. So if you get fuel that's hot enough you wind up with it boiling under suction, the pump sucks the vapor and fuel pressure collapses. You're done.
It restarts immediately for the same reason a pot you take off the stove stops boiling immediately; it only has to drop a fraction of a degree to not boil in the intake. Vehicle starts, but, if you don't wait at all it will do it again under similar circumstances. If it's cooler or at lower altitudes you either get less-heated fuel return, higher boiling point or both and it runs fine.
The NEW pump has *two* chambers in the cup. The first has the flapper valve at the bottom and a suction line, plus the return coming in at the top. That suction line from the bottom of that chamber then goes to the *second* chamber, which is JUST the motor. I presume there's a bleed on the output so the pump gets liquid (gas) for cooling. This resolves the issue because....
1. If the fuel in the cup boils since the pickup is at the bottom it doesn't matter; if enough vaporizes to reduce the level below that in the tank then new cooler fuel from the tank will be admitted through the flapper. In any event the pump gets liquid, not vapor.
2. The pump itself, assuming there is a bleed, has *positive flow* cooling from the fuel flow, not passive from the what's in the cup. It's heat gets dumped when the cup overflows into the larger volume of the entire tank; ergo, the motor runs cooler and is not subject to boiling in the suction section.
3. This also implies the pump is now both a positive-displacement (suction side) AND turbovane (high-pressure) side device. Why not positive displacement on the output side? Because a fault in the regulator could dead-head it, and it would immediately burn up in that instance. However, being positive-displacement on the suction side, which is *necessary* otherwise the feed would never work if the tank was not full to beyond the height of that hose's loop, means that IF there is an attempted vapor problem there it is immediately cleared because the positive-displacement side will force the vapor through the vane pump anyway and fuel will follow.
The design change is obvious on inspection; I am deducing the reason, but I would not bet against it. IMHO this darn well should have been TSB'd or even recalled.
Next time I'm at altitude I bet it doesn't happen again since the current design appears to be immune to it.
So two separate problems, which the new pump appears to have fixed.
First, there was obviously an air leak on the suction side somewhere up high enough that once you got down to about 1/4 tank it sucked air and was done, even though there was fuel.
The second is a design issue from what I can tell. The new pump has two chambers in the cup, the original has one. Here's the hypothesis:
At altitude boiling point is lower, and heat under the hood (IATs, etc) are quite a bit higher. Gas starts to boil off at 95F at seal level, ~90F at 5,000'. The return line goes into the cup in the original design. This means at altitude the return fuel at highly elevated temperature can not circulate into the tank and thus be cooled down, plus the motor of the pump is cooled by the same flow, so it's getting even more heat added to it. So if you get fuel that's hot enough you wind up with it boiling under suction, the pump sucks the vapor and fuel pressure collapses. You're done.
It restarts immediately for the same reason a pot you take off the stove stops boiling immediately; it only has to drop a fraction of a degree to not boil in the intake. Vehicle starts, but, if you don't wait at all it will do it again under similar circumstances. If it's cooler or at lower altitudes you either get less-heated fuel return, higher boiling point or both and it runs fine.
The NEW pump has *two* chambers in the cup. The first has the flapper valve at the bottom and a suction line, plus the return coming in at the top. That suction line from the bottom of that chamber then goes to the *second* chamber, which is JUST the motor. I presume there's a bleed on the output so the pump gets liquid (gas) for cooling. This resolves the issue because....
1. If the fuel in the cup boils since the pickup is at the bottom it doesn't matter; if enough vaporizes to reduce the level below that in the tank then new cooler fuel from the tank will be admitted through the flapper. In any event the pump gets liquid, not vapor.
2. The pump itself, assuming there is a bleed, has *positive flow* cooling from the fuel flow, not passive from the what's in the cup. It's heat gets dumped when the cup overflows into the larger volume of the entire tank; ergo, the motor runs cooler and is not subject to boiling in the suction section.
3. This also implies the pump is now both a positive-displacement (suction side) AND turbovane (high-pressure) side device. Why not positive displacement on the output side? Because a fault in the regulator could dead-head it, and it would immediately burn up in that instance. However, being positive-displacement on the suction side, which is *necessary* otherwise the feed would never work if the tank was not full to beyond the height of that hose's loop, means that IF there is an attempted vapor problem there it is immediately cleared because the positive-displacement side will force the vapor through the vane pump anyway and fuel will follow.
The design change is obvious on inspection; I am deducing the reason, but I would not bet against it. IMHO this darn well should have been TSB'd or even recalled.
Next time I'm at altitude I bet it doesn't happen again since the current design appears to be immune to it.
September 7th, 2021, 9:42 AM
#26
CF Beginner
Join Date: Sep 2021
Posts: 1
Likes: 0
Received 0 Likes
on
0 Posts
Has anyone Considered Purge Valve or Evap canister?
I saw someone saying crank sensor, but nope -- the tach is normal and when it acts up it does NOT go to zero. I've seen a CPS failure before and the instant "giveaway" is that the tach goes to zero as soon as it happens. On the Express it misfires before it dies and flashes the check engine light, but on my Suburban it does not; it just has zero throttle response, no flashing check engine, and as you pull over (and thus now the engine is forced to produce "some" power against the torque converter) it dies. But in every case thus far cycling the key off and on has produced an *immediate* restart -- a quick stumble but then normal idle and power is back as usual.
I've never seen a cam sensor fail without throwing a code and there are no codes in the ECU. I've got Torque and an OBDLinkMX but have yet to catch it when it actually happens, and both MAF and MAP (vacuum) look normal while driving. There is no PID exposed for fuel pressure that I can find which sucks big ones because that would *definitively* implicate the fuel pump but it appears that the ECU doesn't have a sensor for that on these vehicles.
The Suburban has the OE fuel pump; it has the relatively common fault that the gauge is off by a fair bit (fuel reading 1/4 is actually EMPTY) but I'm loathe to replace it with a Chineesium one that, if it pukes, will strand me. The Express was bought used so what's been changed on it is unknown but its fuel gauge reads properly.
I'm very interested in the actual design of the fuel pump and exactly what drives and regulates it. If there's a rail sensor at the engine that cycles it then that switch could be involved, but I *suspect* there's a circuit board in the pump that likely either regulates output pressure and/or has thermal and/or overcurrent protection and, if it trips, you get no fuel pressure which would produce this EXACT symptom -- and killing power to it would reset it.
I hate playing "throw parts at this"; it's almost always stupid to go down that road, but given the intermittent nature of the issue and that I can't provoke it "on demand" it's a serious problem trying to find this. The Express is my daughter's and she often uses it for longer-term travel; getting stranded 100 miles from nowhere is NOT cool for obvious reasons, and the threat of an intermittent failure turning into a hard failure is out there.
I've never seen a cam sensor fail without throwing a code and there are no codes in the ECU. I've got Torque and an OBDLinkMX but have yet to catch it when it actually happens, and both MAF and MAP (vacuum) look normal while driving. There is no PID exposed for fuel pressure that I can find which sucks big ones because that would *definitively* implicate the fuel pump but it appears that the ECU doesn't have a sensor for that on these vehicles.
The Suburban has the OE fuel pump; it has the relatively common fault that the gauge is off by a fair bit (fuel reading 1/4 is actually EMPTY) but I'm loathe to replace it with a Chineesium one that, if it pukes, will strand me. The Express was bought used so what's been changed on it is unknown but its fuel gauge reads properly.
I'm very interested in the actual design of the fuel pump and exactly what drives and regulates it. If there's a rail sensor at the engine that cycles it then that switch could be involved, but I *suspect* there's a circuit board in the pump that likely either regulates output pressure and/or has thermal and/or overcurrent protection and, if it trips, you get no fuel pressure which would produce this EXACT symptom -- and killing power to it would reset it.
I hate playing "throw parts at this"; it's almost always stupid to go down that road, but given the intermittent nature of the issue and that I can't provoke it "on demand" it's a serious problem trying to find this. The Express is my daughter's and she often uses it for longer-term travel; getting stranded 100 miles from nowhere is NOT cool for obvious reasons, and the threat of an intermittent failure turning into a hard failure is out there.
September 7th, 2021, 9:45 AM
#27
Both were checked and ok. On the Express, which was doing a similar thing, I also changed the pump after, once again, both purge and cannister were checked and ok.
Next time both are out "at altitude" when its hot it'll be interesting, but I have noted on the Suburban that the very, very slight intermittent miss at idle is gone. My bet is that in both cases it was the fuel pump itself.
Next time both are out "at altitude" when its hot it'll be interesting, but I have noted on the Suburban that the very, very slight intermittent miss at idle is gone. My bet is that in both cases it was the fuel pump itself.
July 26th, 2023, 6:49 PM
#28
CF Beginner
Join Date: Jul 2023
Posts: 3
Likes: 0
Received 0 Likes
on
0 Posts
Did the fuel pump replacement solve the dying in heat and altitude?
Both were checked and ok. On the Express, which was doing a similar thing, I also changed the pump after, once again, both purge and cannister were checked and ok.
Next time both are out "at altitude" when its hot it'll be interesting, but I have noted on the Suburban that the very, very slight intermittent miss at idle is gone. My bet is that in both cases it was the fuel pump itself.
Next time both are out "at altitude" when its hot it'll be interesting, but I have noted on the Suburban that the very, very slight intermittent miss at idle is gone. My bet is that in both cases it was the fuel pump itself.
July 26th, 2023, 7:19 PM
#29