First season with new engine (and car)

[falcor]

So this summer was the first season with my new 2256 cc Engine, and the first season with the 914 aswell. I didnt get to drive as much as I wanted due to a whining gearbox but I still managed 3000 km or 1900 miles. I pulled the plugs the other day and they do look a bit different between the banks.

Cylinder 1 and 2 seems to have run hotter/leaner than cylinder 3/4, this is what the CHT temps on cyl 1 and 3 has been telling me all season aswell.
With the engine warmed up and at idle both sides show equal temps, but as soon as you start driving for a bit the 1-2 side goes hotter than the 3-4 side.
Cruising at 70-90 km/h (45-55 mph) the temps settle around 320F for the 1-2 side and 280F for the 3-4 side.

I'm trying to figure out what could be causing this and my first suspicion is the fuel supply.
This is the order things are connected fuelwise: Tank-pump-filter-cyl4-cyl3-cyl1-cyl2-FPR(3 bar)-tank.
Do I need to split the fuel supply so I supply the banks in parallell instead of in series? I didnt think it would be needed since I'm nowhere near the max capacity of the injectors or pump. The engine puts out 160 hp and the injectors are Bosch 350cc/min (Brown ones from a Saab 9-5 Aero)

I'd love to get the temps more even between the banks aswell.

sp_alla.jpg

sp1.jpg

Cyl 1

sp2.jpg

Cyl 2

sp3.jpg

Cyl 3

sp4.jpg

Cyl 4

[UKLuke72]

What ignition setup do you have? Any chances that's where the discrepancy is?
Every production car I have ever seen has had the injectors in series... Have you verified you're actually getting 3 bar? Also it might be worth switching the injectors (unless you've had them flowed)

[_thesatanicmechanic]

Install a fuel pressure gauge to verify you are maintaining 3 Bars of pressure.

I use a GoPro to record the gauge while driving and then look at the video afterwards. It's much safer than running an extension hose into the passenger compartment.

Might also be a loose or ill fitting cylinder head on the hot bank. Also could be an air leak at either the intake or exhaust too. Use a leakdown test rig to verify the internal engine sealing situation. I have a mineral oil smoke generator for finding intake and exhaust leaks, but it's relatively easy to rig up a regulated low pressure air supply to pressurize the exhaust system, ditto with the intake if you are running a plenum.

[raygreenwood]

Check your cooling flap position. The 1/2 side actually has better airflow than the 3/4 side......and.....in series injectors on a ring main loop...actually can have flow issues with the last in line injectors.

1. Check your fuel pressure stability with a GOOD gauge.....not a store bought, no rating gauge kit or a water pressure gauge. Check for +/- 0.5 psi stability. Factory when new was just adequate at +/- 1.0 psi.

What pump are you using? What injection system are you using?

2. Have you flow/volume tested your pump? Pressure while key.... is secondary....IF the pump cannot maintain volume at pressure. Having the last in line injectors starve is indicative of poor supply volume even if pressure is normal.

A fluctuating gauge pressure at idle, a big momentary dip of 3-5 psi......are big indicators of flow volume issue.
Ray

[falcor]

I've got a rail mounted pressure gauge at the cyl2 injector (last in the series before the FPR) and its stable at 3 bar, but then the resolution is very low on it.

Pump is a Bosch 044 so that shouldnt be an issue. The injectors are good for a 250 turbo Engine and they were flowed and checked before being used.

Injection system is aftermarket EFI and 40 mm itb's, LS2 coils, these are not my primary suspects....yet. Engine is newly built by a reputable builder so that should be ok. No exhaust leaks, I've pressureized the heat exchangers and made sure there are no leaks between the exchanger and head.

The engine is out of the car at the moment due to a gearbox rebuild but as soon as its back together I can do some testing of the fuel pressure with a better gauge. I will also have the injectors rechecked before it goes back together.

[raygreenwood]

I've got a rail mounted pressure gauge at the cyl2 injector (last in the series before the FPR) and its stable at 3 bar, but then the resolution is very low on it.

Pump is a Bosch 044 so that shouldnt be an issue. The injectors are good for a 250 turbo Engine and they were flowed and checked before being used.

Injection system is aftermarket EFI and 40 mm itb's, LS2 coils, these are not my primary suspects....yet. Engine is newly built by a reputable builder so that should be ok. No exhaust leaks, I've pressureized the heat exchangers and made sure there are no leaks between the exchanger and head.

The engine is out of the car at the moment due to a gearbox rebuild but as soon as its back together I can do some testing of the fuel pressure with a better gauge. I will also have the injectors rechecked before it goes back together.

Actually the 044 lump can be a problem with regard to fuel volume.....if it's cavitating. In some applications it requires a feeder pump. This is why a good resolution gauge is critical.

Also do you have the cold start injector plumbed inline to the system like normal? It can occasionally become a restrictor to the high flow system. Its worth taking it out of the loop temporarily to check.

The closer to the regulator an injector is the more the harmonic from the regulator can affect it. It is worth measuring the fuel pressure alternately on each bank.

Also....you can have injection accuracy issues when the injectors are too large as well.

Those are some things to think about.

Moving on from fuel......how are your CHT gauge probes installed....and on which cylinders of each bank. The cylinder in the back of each bank...1 and 3.....typically run the most different in temp.
Also have all of your thermocouples been matched to each other? Are you using two seperate gauges or switching back and forth?
If switching.....did you map both TCs before install? Are these CHT gauges compensated?
Ray

[falcor]

Thanks for all the suggestions Ray, I'll make a list and work thru it during the winter.

No cold start injector.

The distance betwwen the #2 cylinder injector and FPR is around 15 inches of fuel pressure line so its quite a bit inbetween.

CHT probes are underneath the #1 and #3 sparkplugs, so the rearmost cylinder or each bank as I figured those two would be the ones running hottest.
The CHT's are connected directly to the ECU as it has 8 type K ports to be used for CHT or EGT probes so I can read them both at the same time, the ecu compensates for the temp differens between the sensor and plug.

[raygreenwood]

Thanks for all the suggestions Ray, I'll make a list and work thru it during the winter.

No cold start injector.

The distance betwwen the #2 cylinder injector and FPR is around 15 inches of fuel pressure line so its quite a bit inbetween.

CHT probes are underneath the #1 and #3 sparkplugs, so the rearmost cylinder or each bank as I figured those two would be the ones running hottest.
The CHT's are connected directly to the ECU as it has 8 type K ports to be used for CHT or EGT probes so I can read them both at the same time, the ecu compensates for the temp differens between the sensor and plug.

That last part can be the problem. What exactly do you mean the ECU compensates for the temp difference between the sensor and plug?

And...what does it do with this information?

Meaning...does it compensate for the hotter or cooler difference it reads in each cylinder by changing fueling (which I doubt)?.....or does it compensate for the ring terminal probe joint difference under the plug?

This is not a good method...the second one.....because as I was getting at...if there is a difference between the thermocouple reading from temp sensor 1 and 3....that is there in the TC wiring (very common) the ECU should be doing no translation. It would have no way of knowing which TC was actually correct.

The optimum method is to...before installation...rig up a contact strip to plug either a gauge or an accurate volt meter into. Then run each TC lead to a heated block (hot water is fine)...with the ring terminal screwed down under the exact same spark plug to the same torque. Read both TC at the same temp. Make sure their output is either the same or the difference noted. You cannot allow an ECU to "normalize" or rationalize differences between two different TC's. It does not have enough data to do so.

Reading at the spark plug...you are just as much likely to be reading spark plug temp as actual head temp. You can have higher reading from a plug that is leaking gas through the threads....or one that is gapped poorly.

It would be worthwhile to backstop the data from your ECU by connecting the TC's in turn to a standard CHT gauge to check what is coming down the line data wise.

You can also have an injector seal leak or an intake runner leak and run hot on one cylinder.

Ray

[falcor]

The ecu does nothing with the data, it just displays it in realtime on my tablet.

It does not compensate between the cylinders, it compensates for the difference between there the sensor tip is and where the sensor connector is at.

The injectors were flowchecked and serviced before the engine was started.

[raygreenwood]

Ah!...sorry...spaced out. I know exactly what you mean. Its TEMPERATURE COMPENSATED.
Excellent.

But the question is still the same.....did you verify that both thermocouple wire sets are showing the same throughput/output in a static test before you hooked them to the ECU?

If one TC is different than the other...that right there could be your temp difference.

Oh...also....is this system bank fired, paired or sequential?
There is always some residual issue with bank fire depending on actual timing. Above about 1500 rpm....its not the open closed valve issue...but there is an atomization issue...always.
That may be worth only a percent or two or three....but if you have a few other items in the stack up...vacuum leaks or not, spray pattern etc....the stack up can make the difference you see.

By the way...what intake set up are you using? Plenum or ITB's ? Ray