DTC P2635-00 on 2007 GMC 5.3 AFTER Fuel Pump Replacement
Hello GM Driveability Gurus!
Off the road from training and helping out on the Delphi hotline. Call came in Thursday 4-17-20 from a very capable shop owner / tech working in the (healthy COVID-19 free) Upper Peninsula of Michigan. This one is stumping me. 2007 GMC Sierra (new body style with CAN bus) 5.3 VIN J (LY5). Truck was just fine with 165K on it when the customer elected for a "preemptive" fuel pump replacement along with numerous other preventive maintenance related services to get another 100k mile lease on life.
1st replacement fuel pump (Airtex) and 2 subsequent replacement pumps (OEM - Delphi) all setting DTC P2635 (Fuel Pump Low Flow / Performance) codes. NO driveability problems and there were no DTCs prior to the 1st pump's replacement. Fuel trims are low single digit (normal). An OEM dealer supplied fuel pump control module has been tried and the ECM has been flashed to the most recent calibration. No changes. Unfortunately I believe the original fuel pump has been scrapped.
Three questions;
1. Has anyone encountered this DTC for "Indicated" to "Desired" fuel pressure error on a GM SFI?
2. How far off can the "Indicated" and "Desired" fuel pressure PIDs be before this DTC sets? GM's been writing their own ECM CAL software for at least 2 decades now or I would be able to pull that number up through internal records as in the 'good ole days' at Delco Electronics. Mode $06 MIN/MAX #s are no help - this DTC sets fast so it must be a continuous monitor. This is a conventional 2-wire DC brushed pump - not the new fancy 3-phase brush-less style. Pro-Demand and ALLDATA not showing much in their flow charts beyond checking pump to control module circuits for opens. (pathetic) COVID-19 corporate austerity program has me unable to spend the $ to get a short term subscription to GM's SI2000 as I normally would. (understandable but equally pathetic) No TSBs or PI's (Published to aftermarket ESI anyways)
3. I see some cataloging differences - lots of p/n's for 2007 GM V8 fuel pumps based on bed lengths, trim levels, flex fuel, etc. Specs for different p/n's of pumps for this application range from 53 PSI / 20 GPH up to 58 PSI / 42 GPH. Does anyone feel this much variation (5 more PSI / 22 more GPH) could result in this DTC? My gut tells me yes - 'wrong pumps' putting out too much pressure are to blame. Having a hard time believing the tech has obtained 3 defective fuel pumps from 2 different brands . . . all running great but setting pressure spec related DTCs. I have seen this craziness with C.O.P.s fire spark plugs perfectly and still setting feedback DTCs so nothing is surprising me anymore!
FACTORY SCAN TOOL READINGS
Indicated fuel pressure (PID) and fuel pressure gauge read the same - always. Desired fuel pressure PID is consistently lower than indicated PID / gauge measured fuel pressure by about 6.5 PSI. Another curiosity; periodically (and momentarily) indicated pressure PID spikes to 80 PSI at same time the Pump Duty Cycle PID spikes to 44 %. This is very interesting . . . a self check?
Fuel Pressure (on Mechanical Gauge)
Idle - 50 PSI
Under light load – 43 PSI
When commanded lower by scan tool – 43 PSI
Periodically (and momentarily) spikes to 80 PSI
Indicated Fuel Pressure (Scan Tool PID)
Idle – 50 PSI
Under light load – 43 PSI
When commanded lower by scan tool – 43 PSI
Periodically (and momentarily) spikes to 80 PSI
Desired Fuel Pressure (Scan Tool PID)
Idle - 43.5 PSI
Fuel Pump Duty Cycle
Idle 35 %
When commanded lower by scan tool – 32 %
Periodically (and momentarily) spikes to 44 %
Also; when tech uses scan tool to command the in tank pump to run at a lower pressure, it does. Duty cycle goes down (as you would expect) along with desired pressure PID which is always lower than actual pressure by a good 6 or 7 PSI.
Thanks in Advance and STAY Healthy!
Dave … / … / … Mobile
members.iatn.net/forums/search…
I found this on a similar year with a very similar situation in another thread. silveradosierra.com/engine/p2635-a…. Hope it helps. Everything is wired as just enough. So if you do have a loose wire the resistance might be just enough to run it but not pop a fuse. Then as the wire makes contact again and the resistance is gone, you see a fuel pressure spike. Harness by spare…
Thanks Michael - that's an interesting theory. If the new GM dealer source fuel pump the shop owner just ordered doesn't fix it, I'll have him look closer at the wiring. Being up in Michigan's UP - I'll but the truck is used like a truck making this theory even more plausible. Hoping the factory pump's "indicated" fuel pressure PID and "desired" fuel pressure PID are going to be exactly the same…
Thanks Geoff - I checked that out. I think you probably nailed it. That extra 5 PSI didn't effect anything other than failing the DTC software parameters. Most likely the pressure regulators my company Delphi and Airtex are both using comes from the same source. Hopefully GM hasn't changed their source yet. Or they've be coming out with a new calibration. I suspect a lot of calibration changes…
Hi Dave, Just a heads up - did you write the part number down incorrectly? … doesn't show to be the right part for this vehicle. I actually can't tell you what is fits as it's not coming up. They need …MU10206, W/O/Level Sensor, or …MU1847, W/Level Sensor
Thanks Geoff - I checked that out. I think you probably nailed it. That extra 5 PSI didn't effect anything other than failing the DTC software parameters. Most likely the pressure regulators my company Delphi and Airtex are both using comes from the same source. Hopefully GM hasn't changed their source yet. Or they've be coming out with a new calibration. I suspect a lot of calibration changes…
This is a perfect example of why I dont replace things for "maintenance " like this. I'd rather wait for a fuel pump to die on it's own than fight a problem the car never had before. I'm sorry I'm not any help on this one, but almost everytime I try to be preemptively active, it bites me in the ass every time. It's actually nice to see it's not just me. Only thing I could think would be helpful…
Thanks and I fully agree! Outside of timing belts, tires and other items on a standardized PM list. . . couldn't agree more. Especially today with some many trickle down system interactions. Even software calibrations . . . they're not like a new pair of shoes for our wives. Just because there's a newer one out there don't mean you have to have it. As they say; If it's not broke . . . Don't…
Hi Dave: You may wish to rethink your "safe list". import-car.com/honda-tech-tip… HTH, Guido
Hi Guido, I'm surprised that Import Car would publish such a dumbed down article. Three description of how the system works is incomplete and will mislead techs. I suppose if the idea was to inform techs about the need to replace all the tires instead of one or two then it might be acceptable, but it still leaves out a lot of important information. Maybe they just needed something to fill…
Hi E: I'm with you as far as being less than impressed with the quality of that and a lot of the other rags. The only take away for me was the differences in the tire depending on where it comes from. It is quite possible that 1 tire could be from a different location even when replacing all 4. So, there is some value there. Guido
Thanks Guido. Interesting indeed. More reason to replace tires in complete sets but road side hazards can make that an impossibility. On an out of state trip a few years back I sliced a sidewall on a metal curb protector (barely rubbed against it) so I spent TONS of extra time making sure I got the EXACT Goodyear tire to replace the destroyed tire. I've encountered the same thing with stability…
Turns out I was misled or misunderstood the shop owner. Taking hotline calls is similar to service adviser / customer experiences. Memories sometimes change as the story unfolds. As stated on a 2nd call with the shop owner there was a minor problem with the pump - fuel lines coming out of the OEM were rusting away. Wanted to nip it in the bud before leak started. Customer's choice. Truck fixed…
My uncle has an 07 Chevy pickup and he replaced his pump due to rot. After it was setting a low flow code, found he had incorrect fuel pump. Replaced with correct fuel pump and problem was gone. The one he put in was not the correct one for a fuel pump control module vehicle.
Complete voltage drop checks of fuel rail pressure sensor. Could be a slight voltage drop on any of the 3 wires that the computer does not like. Yes, you would probably think that the computer would set a fuel rail pressure code, but I have seen some weird things. I know you said they changed 3 pumps. Is there something there something from the old pump that is being connected to the new pumps?
Thanks Saifullah. That is a good suggestion. Lots of us forget that voltage drops can occur on low current sensor circuits as well as higher current ones like blower motors and fuel pumps. In this case, since the indicated PID for fuel pressure and the mechanical gauge lined up perfectly, the sensor is probably accurate. If a different (and exactly correct) fuel pump doesn't repair the vehicle…
Thanks Shane - was leaning that way but needed some reassurance!
Hi Dave, I believe the answer to the fault is the OEM pump. If a VIN would have been posted I could put up the correct part#. As you mentioned there are multiple choices. This would be interesting for you to get the Delphi pumps back to see where the difference(s) is between OEM and the Delphi and then make the correction(s) necessary. As far as GM SI they're as vague as they sometimes are…
Thanks Scott - I'll going to bug the guy working on this to get me exact PIDs of the new GM pump once he gets it. The IATN case that Geoff linked on another vehicle with the same problem said the new GM pump did have "indicated" and "desired" pressure PIDs at the exact same numbers like you would expect.
Need a VIN and part numbers of pumps used in order to help. I'm pretty sure this truck has the wrong fuel pump installed, I've seen it a few times in the past. The correct, VIN sourced at the dealer, fuel pump fixed them all.
Thanks Eric! I totally agree and so does the shop owner / tech. New GM dealer pump is on order. 2GTEK19J… - the Delphi pump p/n FG1058 is correct for that VIN. Hopefully GM's dealer p/n still has the same pressure regulator in it that gives the correct pressure for given duty cycle.
I have had trouble with Delphi cataloging on these before, and even OEConnection catalog doesn't give certainty of application is some cases. Are you certain that Delphi is selling the OEM module for this? I have doubts. Update- I checked, and the Delphi part is not the same as the GM/ACDelco part. …M10206 These are frustrating applications, because in some cases Delphi sells an OE…
Thanks for your comments and part look up effort Marlin. I apologize if you've had problems with our catalog and our fuel pumps. The shop owner installed a dealer sourced … and fixed the truck. I don't know if the Delphi unit is still the original OEM part we sold GM for their truck assembly plant in Canada or not. The rest of this post is not intended to be excuses - just insider insight…
Many things that said Delphi on pre-bankruptcy GM parts now say ACDelco (along with "GM") when bought straight from the dealer. I was under the impression (from web research), that the company you work for shared nothing but the name, with the company from 20 years ago.
I get what you are saying about part production and sales forecasting. But here is the deal; if some factory can build a module, a pump, a harness, a regulator. etc. AFTER the OEM has ceased production, then the OEM (which did it in the first place) obviously could as well. So, it just comes down to an excuse, with boosted profit margins as the driver. I also agree with what you wrote about…
Hi Dave, "Now you have to depend on someone else's workmanship and quality standards despite your insistence to have them "build it your way". " I'd go along with that, but I think the real reasons come down to lack of testing (quality control) and building a product to a price point. Years ago when GM came out with HEI they had problems which were thought to be the ICM's. Go to parts and…
The "built to a price point" is a reality, but the insanity is that a high-quality part often costs less (at our level) than a garbage part!
Hi Marlin, I've seen that as well. How much did that cost you? You know you could have bought the OEM for less than that? Evidently they didn't know.
diag.net/file/f6hkxs9pn… One example of what I'm talking about. It wouldn't have been so bad, but this part didn't last too long. Sigh
Well, now that Wells is NGK, maybe quality will improve. I sure hope so, because at this point aftermarket engine management parts are garbage, with very few exceptions.
Hey Dave. I know you have had a lot of answers here and the correct pump probably is the answer. Going with what everyone is saying I might add this. Agree with V drop tests to sensor. Also I would think it mandantory too do loaded V drop from the FPCM to the Battery. Also loaded V drop from the FPCM to the fuel pump itself. Does he have Mityvac 5455 he can do an actual flow test with?? We…
I have a MityVac 5455 if it comes down too that. I would be very strict with borrowing/rental terms lol. The tool is hard too come by and one of the best on the market!
What are the fuel pump trim numbers in the FPCM data??
Hi Caleb - thanks VERY much for the MityVac 5545 offer. We actually have an entire section in our fuel systems training manual dedicated to that very helpful tool. I've used one down at the AVI studio. Wouldn't run a shop without one. Most shop owners argue that great pressure rarely is accomanied by low volume hence their excuse to not own a volume tester. Obviously it's served you well! Great…
Thanks for all of the great suggestions. Shop owner installed a GM dealer sourced fuel pump p/n … No more DTCs. PIDs for desired and indicated are identical now. FP duty cycle PID is now at 42 % instead of 35 %. Engineering looking at this. STFT 0.8 and LTFT 0.7-0.98 from the FPCM (Fuel Pump Control Module) did not change with the new pump.
Awesome good deal. So it looks like the wromg pump was actually too much volume??
He said in his original post that at 35% on time the bad pump put out 6-7 psi more then desired. With the new pump the on time is now 42% and the actual psi has come down too meet the desired. Seems like the bad one flowed too much causing the code.
The pressure was too high. Volume (as relates to the pump module) is controlled by injector on time.
Caleb, Could you please explain to me how you came to your conclusion as well as what you meant by your original statement, "So it looks like the wromg pump was actually too much volume?" Thank-you. Eric