6.7 Cummins Fuel Rail Pressure dropping at 170 degrees coolant temp

I am experiencing a drop in actual fuel pressure vs commanded fuel pressure after my ECT reaches 170 degrees (give or take a few degrees) I have several datalogs that confirm this. I can start the vehicle in the morning, even cold mornings in the mid teens without the engine being plugged in and FP_D is always even with FP_A on my EFILive V2. Both read 59 MPA. I can leave the truck idle for 30 minutes or more and both stay consistent. The problem starts when I start driving and within 5 miles of driving the actual fuel pressure begins to fall. At a steady speed of 55 mph the commanded pressure will be in the 70-80mpa range and the actual pressure will be in the 35-40 mpa range. I have verified my Air dog is delivering constant pressure to the CP3 via a fuel gauge. I actually just upgraded the CP3 to a new 10mm. The truck has compound turbos and 45% over injectors. The ECM is a CM2200 with a custom tune. I have run this tune for several years without this issue. After driving and arriving at my destination, I can let the truck idle for 5-8 minutes and the commanded and actual pressures will even out again. I can then take the truck back out for a drive and pressures will match again for a few miles until the temperature rises to the 170 degree range. I put a pressure relief plug in for testing purposes and it made no difference.

Scott,

A datalog of the event might be useful. I would be particularly curious to see if the FRPCOM (mA) is trending below 1000mA when you're having these issues. If so, that would tell us the regulator is being commanded open. Fuel pressure issues can be tricky to track down, the first few things that come to mind:

1- If it only happens when the fuel is warm, it could be an injector return rate issue. Warm fuel flows better and can get past the internal parts of worn injectors easier. There are return fuel flow tests to measure how much fuel comes out of the return system under certain params.

2. - Check oil for signs of fuel getting past the push tubes etc.

3. - Cascade valve in the CP3 - look it up.

Hope this is helpful,

Nick

Nick,

Thanks for your suggestions. Since I initially posted my problem I have discovered another tendency. My initial data logs were showing that the actual pressure was indeed dropping as ECT would rise. Last week I drove the truck home from my shop (about a 14 mile trip) As usual when I would be a few miles into the trip the pressure would start dropping. However I since have noticed that the pressure would start to rise again getting close to commanded fuel pressure the longer I drove. On the same day I made a return trip to my shop with a dump trailer and my skid loader. Expecting it to be a struggle to tow the heavy load, to my surprise it ran perfect. I have attached data logs for the two above mentioned trips and two other trips. Looking at the data log towing the trailer I noticed that my ECT started out at 185 degrees. For the whole trip Commanded and actual pressures were equal (within reason) This started to make me question the heat being the cause of my problem. Before your reply, I had starting looking at the FRPCOM readings. It is indeed hovering around 1000 and below when the actual fuel pressure is dropping. Since my trip hauling the trailer was fine I decided to try to raise the ECT by blocking off the radiator. (I am in Ohio and it has been in the low to mid teens during all of these attached data logs.) With the radiator blocked, fuel pressure would still drop as normal but it would then equal out sooner. Two of these data logs confirm that. Now I am seeing on the data logs that when the pressure is dropping that the FRPCOM is basically flat lined at 1000 m/a dipping below and rising slightly above. I am assuming my cascade valve is fine since the CP3 is a brand new pump. (I wasn’t sure if brand names were allowed…it is a highly recognized brand, fully tested before shipped) The new CP3 has helped the problem but it hasn’t completely fixed it. All of the attached data logs are with the new CP3. I had also mentioned originally that I took out the fuel rail pressure relief valve and installed a plug to eliminate that as a possible cause. I have not tested the injector return rates. Hopefully this additional information will be useful.

Data log descriptions: (all trips are the same distance)

Day one: CMD_0049 is the 14 mile trip from shop to home empty, starting the trip after a slight warm up. CMD_0051 is the data log from home back to the shop hauling the skid loader and starting out with an ECT of 185 degrees. (radiator not blocked for day one logs)

Day Two and three: CMD_0053 was starting out after a slight warm up towing the skid loader back home. CMD_0054 was driving the truck back to the shop the next morning after a slight warm up. Both of these trips show that the pressure would initially drop for a short period but recover quicker and then maintain for the rest of the trip. (The radiator was blocked for both day tow and day three examples.)

Scott

Nick,

I have some additional information to add. This morning our outside temperatures were in the low to mid 20’s. The night before, I did not plug the truck in. When I started it this morning, it started right up but I noticed a slight miss, or lope in the idle. I have noticed this before a couple of times recently but just attributed it to being cold. This morning I climbed in the cab and increased the rpm’s to see if it would smooth out. It did not smooth out and there was noticeable amounts of white smoke coming from the exhaust at the increased rpms. It reminded me of my old 5.9 12 valve with the mechanical VE pump when it was cold. Once it warmed up the smoking stopped and the lope or miss was barely noticeable. I then began driving and logging data again. The first 5-8 miles produced the same data as previously recorded. Pressures are fine at first and then begin to drop. As with past data logs once the engine was completely warmed up pressures equaled out. The data log attached named “burgcafe_to_sunbury.efi” shows this at frame 22682. This time I decided to data log from the time the pressures equaled out. As I continued driving the pressures stayed equal. I even tried increased acceleration to see if the actual fuel pressure would stay with the commanded. The data logs named “longshore_rd.efi” and “hartford_rd.efi” verify actual pressure staying with the commanded pressure. The biggest variance between commanded and actual would still happen under increased acceleration. Is that due to the latency of the commanded fuel pressure to actually be achieved? The biggest variance would happen at 99-100% throttle. At that throttle position there is around a 4000 psi difference, is that normal or can that be fixed with tuning? Since I eventually get good fuel pressure I am thinking that I could be having a leak between an injector and crossover tube when the engine is cold. Once I was done testing today and returned to my shop, the engine idled smooth, with no lope or miss. Based on your knowledge and experience would you agree with my thoughts about the crossover tube/tubes as a possible cause? The injectors were refurbished and honed to 45% over by a local diesel shop when I built the engine and they now have around 90,000 miles on them. I did torque the crossover tubes to 44 ft pounds back when I installed them and did actually check the torque on the tube nuts a couple of weeks ago and they were still tight. I don’t think I used new crossover tubes when the injectors were installed 90,000 miles ago. I want to do some tune modifications to be able to use the additional fuel capacity that I now have with the new 10mm CP3 but I need to get this problem fixed first.

Thanks for any input you may have.

Scott

Scott,

You've definitely got a few issues here:

1. Regarding your fuel pressure, all control logic looks normal-ish. I don't think the issue is in the tuning. Maybe I missed it in the text, but have you been watching low side fuel pressure while towing? A 10mm pump should easily keep up with this fuel demand. I suspect it's being starved. If not, you'll need to do a return test to narrow the issue down to your relief or the injectors.

2. Push tubes - if they're leaking you'll see fuel in the crankcase. Push tube leaks would get worse with heat and load. You should see the oil in the crankcase visibly diluted with diesel fuel in this case. The oil might smell like fuel and the napkin test would show two distinct fluids in the crankcase.

3. You haven't mentioned which turbo you're running, but the dramatic boost drop under WOT makes me think that either your airfilter is plugged or collapsing under power. This absolutely needs attention before adding any power via tuning.

Nick

Nick,

Since your last reply, I have resolved the intermittent fuel pressure issue. Following your suggestion for testing the injector returns I decided to pull the injectors and have them tested by the diesel shop that built the injectors. They tested each injector at 4 different pressures with various pulse durations. All 6 injectors passed those pressure / flow tests and all 6 passed the return rate flow test. This was good news but that still left me with the question of what is causing the problem? The shop said they thought the injector feed tubes looked OK to be reused. I took a closer look at the tube ends and could see the point of contact the feed tubes had been making with the injectors and I didn't want to take the chance of reusing the feed tubes again. I called my local Cummins dealer and gave them my engine number and installed new OEM feed tubes. I also replaced the fuel rail and fuel lines. None of the external lines had been leaking but I figured that a nice new and clean fuel rail and lines would be nice as the old ones had some external corrosion and flaking on them. It's an Ohio truck exposed to winter road salt. I think the old feed tubes were my problem. They were used when I originally installed the injectors around 90,000 miles ago. Now the fuel pressure consistently stays at or exceeds the FP_D the ecm is calling for. Now that I have a properly functioning fuel system, I have three items I'm hoping you can guide me on.

1. I now have a new fuel pressure issue when accelerating while towing heavy or under 100% throttle. I have data logged those conditions and confirmed that my fuel pressure relief valve is being opened. Fuel pressure will quickly climb over 26,100 psi to 28-29,000 plus psi. The data log will confirm the stuttering I feel with the intermittent quick drop in pressure and then the regaining of power that I feel when accelerating hard. The injectors are 45% over and based on a couple of the bench test results they are indeed outflowing stock injectors at high demand. The tests showed that at 26,100 psi for 1700 us the injectors deliver on average 200 mm3. The maximum mm3 in the tune is 150 mm3. I went into the tune and lowered the maximum called for fuel pressure to 24,000 psi and shortened the injector pulse times at higher throttle positions/torque requested tables. I road tested and under the same load conditions it did not set off the relief valve. However I still logged a pressure slightly over 25,000 psi. Am I on the right track by lowering the called out fuel pressure demand and decreasing pulse widths to match what the new 10 mm CP3 and properly functioning 45% over injectors are actually producing to dial in maximum power without setting off the relief valve?

2. My second question is about my sudden drop in boost pressure you pointed out on the data log I previously sent you. I have compound turbos, and is the higher boost above 36 psi / 50 psia confusing the ecm? (I listed below my setup so you'll have a better idea of what I am working on.) After lowering the maximum fuel pressure and pulse widths desired tables, boost pressure now remains steady under heavy load or WOT. However, at the lower fuel pressure and pulse widths I'm only creating 36 psi boost, way under the potential of my setup. Do I need a 10 bar map sensor capable of reading the higher boost pressure that I am capable of making or is the factory map sensor and ecm able to function and properly fuel once the factory map sensor reaches its maximum reading of 50 psia?

3. My air filter intake is not a factory unit. It is a larger than stock aftermarket round filter connected to 4"

piping and then the piping is enlarged to 5" to go into the S475. Would this setup cause inaccurate MAF readings? I've wondered if the whole air intake system should be 5" piping? The MAF sensor is mounted in the 4" pipe about 6-8" upstream of the 4"- 5" transition.

Thanks in advance for any clarity on these topics you can provide.

Scott

64/71/14 .080 A/R

S475 with billet wheel and race cover

Hamilton 178/208 Cam

Steed Speed exhaust manifold Gen2 T3

45% injectors

Air Dog 150 4th Gen

ARP head studs

AEM AFR/Lambda guage, EGT pre turbo, Banks idash guage

4" mandrel bent exhaust

4" custom Air intake. Filter is exposed under the hood.

CM2200 ECM

2006 Allison 6 Speed with T40 TCM

Installed into a 92 F350 Crew cab 4x4 DRW