Fuel pump causing a rough idle? Superlite SLC Kit car with BMW V10

Hello I'm running a BMW V10 engine with an aftermarket ecu (Delta 880 from SCS) and I'm having a really strange problem!

If I pressurize the fuel system and start the engine, it runs perfectly smooth. When the fuel pump turns on the engine starts to stumble and run very rough.

Any idea why adding fuel pressure would disrupt the engine? I even tried running the fuel pump off a separate battery and is still caused a stumble, so I ruled out an electrical issue.

What should I try next?

Perhaps you have cavitation from your pump/pickup that isn't present when you just pressurize the system. A clear piece of tubing as a temporary fuel line will let you see this. Or just pump into a container and see if there is air in the fuel stream.

Perhaps your fuel pressure regulator isn't plumbed correctly (check inlet & returns). Does a fuel pressure gauge remain steady when the pump is running?

Perhaps you have a lot of fuel pulsations (data logging a fuel pressure sensor at 100hz or higher might be insightful), and fuel damper should be added.

Now onto the wierd stuff --

Does your ECU attempt to control the fuel pressure with PWM control to the fuel pump?

Does your fuel cell have an breather (usually with a check valve -- make sure that's oriented the correct way)

Did this ever run correctly, or is this a new installation you are troubleshooting.

David, thanks for the thoughtful reply. Here's a quick update on some of my tests so far.

Pump Cavitation -

Test Completed with success - I connected a temporary fuel line and put the open end into a bucket and found no air bubbles in the fuel stream.

Fuel Pressure Regulator - Plumbing is good verified with the install manual - no flutter on the pressure gauge. Still need to check the ECU for OEM fuel pressure readings.

Perhaps you have a lot of fuel pulsations (data logging a fuel pressure sensor at 100hz or higher might be insightful), and fuel damper should be added. I will see if I can Data Log this in SX Tune software for my ECU.

Does your ECU attempt to control the fuel pressure with PWM control to the fuel pump? I will look into this

Does your fuel cell have a breather (usually with a check valve -- make sure that's oriented the correct way)- Yes the cell has a breather , but no check valve installed as of yet, but I was testing the engine out of a plastic fuel can with the fuel hoses shoved into it and experienced these stumbling issues.

Did this ever run correctly, or is this a new installation you are troubleshooting. - This is a new installation that I'm troubleshooting

Rather than relying on logging, fit a mechanical pressure guage to the fuel rail so you can monitor it and KNOW what you have. If there are fluctuations, you have reason to look at the fuelling side, if it is rock steady, you should probably be looking elsewhere.

This may be a bit weird, but did you make up any hoses, or have them made up for you?

It is possible to have a 'flapper' that is a result of a badly fitted barb that's cut a bit of the liner - under low fluid flow, there isn't enough drag on the flap to pull it very far, but when there is more flow there can be enough drag to pull the flap across the hose opening and the pressure to hold it there - if there is some clearance to the sides, it may be enough to 'just' keep the engine running.

To check the actual integrity of the fuel system, you really need to test the entire line by disconnecting the return line at the fuel tank and checking the flow there, after the pump(s), reg's, rail, etc. If you can, also have someone monitoring the rail pressure while you do that, so you can confirm it is correct throughout the flow range.

Are you using the same fuel pressure in both the case where it works fine, and the case where it stumbles?

Hello David, I'm using the same fuel pressure of 5 bar when it is working and when it is stumbling. I can try turning down the pressure and see if that affects the idle. The ECU support staff thinks that I may be running rich and that's why it runs better when the pump is off.

I'm now wiring up the main harness, so I'm going to take a break from this issue and come back to it when I have proper wiring and grounds to help eliminate any electrical issues as a possible contributing factor.

Well, a wideband lambda controller that was active at startup and logged data with the ECU would answer the "is it running too rich" question pretty quickly.

I wired a SL-C (LS7, Porsche GT3 Sequential gearbox, MoTeC PDM, Keypad, M150, C185). It was a pretty neat car when complete.

Hello David,

Awesome to hear that you have experience setting up a wicked SLC, sounds like it was built for the track with that setup. I am not running a wideband o2 lambda sensor at the moment, but it sounds like a good investment!

I've now roughed in the main wiring harness, keyless start module, digital dash and ecu. The engine is still idling rough and I'm thinking that I may need to adjust the CPS offset but not sure how to do it with precision since there are no timing marks on the crank pulley for using a timing light.

Here is a video so you can hear the engine idle "searching". I did try turning the fuel pressure down from 72 psi to 50 psi but that had no effect.

I will share some oscilloscope readings tomorrow when I have a chance to do more testing.

Here’s a video of the engine idling.

[url=https://www.instagram.com/p/CBAH0TAFblQ/?igshid=1biq8plhycxb3]https://www.instagram.com/p/CBAH0TAFblQ/?igshid=1biq8plhycxb3

I added some EGR block off plates and will try out a new tuning profile provided by SCS makers of the Delta 880 ECU. It will add some additional idle air to the idle.

seems like it should be easy enough to make a timing mark on that flywheel. Here is a webinar on how to find TDC by dead-stopping the engine. I recently did this with an engine that had the wrong pulley marks (timing marks were 180 deg out).

On one engine - I made a dead-stop by breaking the top off a spark plug, and drilling/tapping the center for an allen-head screw that I could adjust. I have also made one from a some 1/2" hex stock on the lathe with a long thread, and a rounded bottom to touch the piston.

Hello David,

Quick update, I took a break on troubleshooting my engine idle issue since my "test setup" had too many loose ends with air leaks and temporary wiring. I'm running my real wire harness and power and ground cables and installing my driveline. With the driveline final setup the crankshaft position sensor, starter, tone wheel are different items.

I also discovered that I had attached my wideband lambda sensors to the harness, but I had not yet welded the bung into the exhaust and installed them. So, the computer was setting the fuel very rich to compensate. This is why the engine would run better when I shut off the fuel pump so it would lean out and idle cleanly before it shut down.

Once I get the drivetrain, wire harness and o2 sensors correctly installed I will come back to continue the tuning and troubleshooting :)

Funnily enough, had a similar issue a long time ago. Bought a preconfigured ECU/intake package that ran really rich. Turned out that the map required an O2 sensor...even though there was no option to include one in the package. Without one it just defaulted to something ridiculous.

Live and learn.