Fuel pressure & tolerances

[-Scott-]

My recently EFI'd daily driver is running pretty well - but the mixture display on the dash shows it's idling rich. Not cycling rich-lean-rich like I was expecting from a closed-loop system, just constant at the rich end of the spectrum.

Had mechanic check fuel pressure - spec is 196kPa, mechanic reported 220kPa. Would this be enough of a difference to upset the injector performance at idle? Do I need to get this pressure down, or is this a red herring?

Injectors were cleaned & serviced before installation.

Self diagnosis throws up no codes.

AFAIK (checked according to the manual I have) the coolant temp & air temp sensors are functioning properly.

Any ideas?

Thanks,

Scott

[Ruffy]

My recently EFI'd daily driver is running pretty well - but the mixture display on the dash shows it's idling rich. Not cycling rich-lean-rich like I was expecting from a closed-loop system, just constant at the rich end of the spectrum.

Had mechanic check fuel pressure - spec is 196kPa, mechanic reported 220kPa. Would this be enough of a difference to upset the injector performance at idle? Do I need to get this pressure down, or is this a red herring?

Injectors were cleaned & serviced before installation.

Self diagnosis throws up no codes.

AFAIK (checked according to the manual I have) the coolant temp & air temp sensors are functioning properly.

Any ideas?

Thanks,

Scott

Scott, what sort of engine management are you running??
by it's nature, an internal combustion engine will like to run slightly richer than stoichometric at slower RPM. It will run smoother with a leaner mixture at higher RPM. (wont get into this here. just making a point)
For this reason most engine management systems will not run in closed loop at idle. There is also quite often not enough heat generated in the exhaust for non heat EGO sensors to operate at idle. They will run at a pre-determined injector pulse with at idle depending on coolant temp, air mass metre or MAP sensor readings.
Because of the engine speed and Low intake air flow at idle you could end up with a hunting idle if you try to control idle mixtures to keep them Stoichometric.
Cheers Dan


EDIT . Didn't answer your other question. I wouldn't be concerned with your fuel pressure at 220 kpa as long as it is being regulated to 220. I would actually say 220 would give you a better spray pattern and a nicer idle. I wouldn't be concerned with a 10% deviation from spec unless it was 10% less. During the fuel pressure check did you crimp the return line to check what the pump was cpable of?

[MightyMouse]

Ruffy's spot on - the system shouldn't be in closed loop at idle otherwise it will hunt driven by the response curve of the sensor and the exhaust flow.

And just for curiosity - what are you using as a Lambda meter ?

[-Scott-]

Thanks for the responses.

Scott, what sort of engine management are you running??

I've taken the efi system from a later model of the same engine, complete with the stock ECU. The car had twin DCOEs when I moved down from Queensland, which SA won't allow. So I've gone this route to try to get through the registration process, hoping they'll accept that the emissions equipment from the later model means my emissions will be as good or better than the original carby version. I spoke to the engineer who did my Paj, and he thinks I've got a chance.

Once it's registered I'll look at larger throttle bodies and programmable ECUs.

And just for curiosity - what are you using as a Lambda meter ?

A Silicon Chip kit which I bought from Jaycar - 10 LEDs. I'll have to check the adjustments on it again, confirm that it's not biassed to rich - but it will run off-scale lean, so I'm reasonably happy it's right.

I've not read about idle not being closed loop, but it does make sense. The bit about the non-heated ego makes sense too. I've got a 4 wire, but the donor vehicle used a single wire, so it's all adding up.

Thanks for the help,

Scott

[MightyMouse]

Scott, the only reliable signal from a narrowband sensor is the transition - the performance either side is not very usable. Some sensors are accidentally better than others but its pot luck. Take it from an OE manufacturer - no effort is put into standardizing the non transition area.

If you calibrate YOUR sensor against a commercial LAMBDA system, you could have a little more confidence in its readings but be warned.

If your going programmable and are going to do some tuning yourself, buy a proper wideband unit - where you can rely on its readings.

Its a interesting topic so do some reading.

[-Scott-]

Thanks MightyMouse. I've heard widebands are much better, but last time I saw prices they were scary.

The programmable ECU is probably 12 months away, and I don't know which way I'll go. Part of me wants to go MegaSquirt, but I've heard the maps are a little coarse - not enough steps. But commercial units seem horribly expensive, and it's not like I'm racing it...

One step at a time. Next step, SA plates.

[MightyMouse]

Have a look at TechEdge units and kits.

Not bad value for what you pay and some have narrow band simulation out.. i.e you run a wideband sensor but the TechEdge unit also produces a narrowband output to feed to the existing ECU.

IF you go spark and fuel with the Magesquirt you get bigger tables.