Lean idle with headlights on

I am having an issue where my car idles fine (with an AFR around 14.7), but when I turn the headlights on, it gets very lean. My microsquirt, injectors, ignition coils, crank position sensor, and everything else EFI related is powered directly from the battery through its own fuse box and isn't seeing a voltage drop when the headlights turn on. I'd assume something else in the stock wiring harness is experiencing a voltage drop, but I don't see how that could change my AFRs. Additionally, I have added an additional relay as Cameron outlined a long time ago to not send all the power through the headlight switch. I also have GT grille lights and a set of Hella 700ffs on the bumper, which would produce a larger load than just the stock headlights.

Any ideas what could be causing this? Is there any way to tune the idle to vary with load?

I've attached a datalog of the event. I turn the lights on twice; the first time is happens when the AFR's spike, and the second time is indicated by a mark.

Here's the download link: https://www.dropbox.com/s/rn2rh9y9rglc09m/2016-04-08_08.47.05.msl?dl=0

Is the ecu powered from the same source as the injectors?

Yes. My battery is in the trunk, and I am using a BMW E36 positive cable. This has a large 1/0 gauge wire, and a smaller ~10 gauge wire. The 10 gauge wire goes to both my megasquirt fuel pump relay and my main relay. The injectors are powered from the fuel pump relay, the ecu is powered by the main relay. So same source, but two different relays.

So not the same source. How do you know the voltage the injectors are seeing is exactly the same as what the ecu sees?

I guess I don't exactly. I assumed because both relays have the same voltage source and are triggered by the same source that they would output the same voltage. I'll spend some time tonight with my multimeter and check if they are indeed getting the same voltage.

Is this simply a case of needing to tune for higher load when lights are on at idle, causing the alternator to load down the engine more? What does auto-tune want to do to your VE table around idle if you turn on the lights and other electrical? You can turn on the A/C also to get the VE table up to still higher loads at idle.

You haven't said anything about what your voltages are doing when you switch on the lights and the MS voltage is not recorded in the logs. When the RPM drops your alternator may not be capable of supplying all the load causing the voltage to drop. At some point, you may be drawing current out of the battery. Batteries have internal resistance (more as they get older) so if you are loading up your battery, your MS supply bus may be dropping lower than you expect. Easy enough to check by including the voltage in the log.

If the MS supply bus voltage is dropping, that should not cause problems for the MS if the injector offsets (opening time) and offset Vs voltage characteristics are correct. If they aren't correct, then the tune that you have created with the system running at 13.8 volts (or whatever your alternator regulator is controlling to) may no longer be accurate at 12 volts (or less), although your change in AFR is a little more than I would have expected.

If your MS supply bus voltage is dropping it will also affect your fuel pump. Normally at idle I would expect that you are by-passing a lot of fuel and if the voltage drops, the fuel pressure regulator should be able to maintain pressure with the lower flow at reduced voltage; however, it is something you can easily check with a fuel pressure test. Perhaps if your pump is approaching the end of its life low voltages may be causing it problems.

Of course, if your log shows that the supply bus voltage for the MS is nice and steady when your AFRs go up, then all of the preceding is not particularly relevant.

To gross polluters point, it would be very easy to confirm that the injectors and the MS are, or are not getting the same voltage. Just measure the voltage on the load side of the main and fuel pump relays with everything running. As a matter of practice, it is probably better to have the main relay power up everything with the fuel pump relay being on the load side of the main relay (and only the fuel pump being supplied by the FP relay). That way you know that the injectors have the same supply point voltage as the MS.

The RPM drops and the MAP increases when the AFR goes up (presumably that is when the lights were switched on?). Could this be something as simple as the operating point has moved to a new cell on the Ve map and the Ve value at that new operating point is not properly set in that cell?

Sorry, I forgot to update this thread with my findings.

The voltage at my injectors is the same as seen by my ECU and does not change with the lights on or off.

I think I probably found the cause of my lean condition. My exhaust manifold had a significant crack. Once I welded that up, the AFRs remain constant whether lights are on or off, at least on my current tune, which has become a bit rich at idle. I need to work on the idle a bit and test again. I will log the voltage and revisit my voltage offsets just to be sure they're correct.

since you relocated the battery to the trunk how is your ground wired? back up to the stock locations on the body and engine? or grounded to the sheetmetal in the trunk?

if you grounded to the sheet metal in the trunk then the added resistance would cause the ground to be at different voltages at each location. ideal would be 0v but in reality you could have a couple of volts.

try putting the battery back where it belongs and watch your weird problems go away.

I welded a stud onto the frame rail.

The lean condition doesn't happen anymore.