Exhaust temp imbalance

I have an F355 engine and trransmission I am installing in a kit car (a P4 replica). It is running on MBE management instead if the Motronic. It also has a handmade exhaust - properly done - equal length primaries etc.

I have set up the idle and it runs OK, and definately on all 8 cylinders, but the exhaust manifold on cyl bank 1-4 runs much cooler than on the dirvers side.

At idle the temps are 160 deg C vs about 90 deg C.

The idle has been setup, the throttle bodies synchronised to 5lbs/hr and the lambda is 0.9 on both banks using a wide band lambda sensor on a PC. Mixture and ignition timing is set across all both banks with one setting.

Running the engine at higher RPM - blipping it a few times to 6,000 increases the temps to 180 and 150 respectively, but still a 30 degree difference.

About the only thing that COULD be different is cam timing, but seeing as the cams were set to a +/- 0.005" tolerance at TDC,I doubt this.

So before I look for something far more complex, does anyone know if this situation occurs on the original F355 installation as well?

Thanks

 

Injectors have been cleaned and spray patterns checked. Cooling system is 100% (whihc considering its entirely bespoke is very comforting) and wiring is so simple compared to the massivelt over complex Motronic the diags would quickly pick it up.

The only thing I have found after looking at the manual for a while is to do with the balance of the individual throttle body pairs. On one bank these are not very well balanced +/- 2lbs/hr but overal the sum of each pair is 10lbs/hr. I'm womdering if the butterfly which is open too much is lifting the idle speed, and thus on the other bank their is simply less combustion "work" to do. This would certainly explain why the temp difference reduces once the speed is increased from idle.

I noticed there is an air bleed screw whihc allows the idle throttle balance to be set, and I am wondering how this works? Is this effectively an air bleed which allows bypass of the butterfly? I can see a hole with a steel ball above the butterfly, does this have an air capiliary which exits on the head side of the butterfly?

 

No, the actual combuustion is optimal - lambda 0.9 giving a AFR of around 11.75:1. So not a fuelling issue.

 

Ignition drive means timing is identical on both banks. The engine uses only one TDC sensor, offsets are all done by the ECU. Its using the stock Bosch sensor on the standard 60-2 front pulley.

I set the trigger point for TDC at 105 degrees to whihc 15 degrees advance is in the map at idle. This was measured using a duial gauge to find TDC and then some new marks on the crank pulley. I think our determination of TDC position is good as we are getting the correct lambda levels in the exhaust on both banks. (*the banks are not linked, so no interference side to side.)

I'm going to look at the idle bleed screws today then rebalance the banks. I'm hoping its just the cool bank is simply not breathing enough and is being "pulled" by the hot one.

 

Will do, its supposed to be having a "inaugural fire up" on New years Day, so video/pics are supressed until all my mates have been over....

Looks like that occasion may be cancelled though as it is becoming clear the cam timing is wrong - I stripped the engine when I got it as it had not been run for 7 years.

I found the problem by running one side of the engine at a time - the "warm side" runs fine and is responsive to throttle. The "cool side" hardly runs on its own, and when it does you can see combustion back up through the throttle bodies.

Its not possible for the ignition to be out due to the way the ECU works.

So should be a few hours to fix, but its not very accessible as the exhaust is in the way. Back to the garage for a while.

 

Fail. Got the cam cover off and the timing marks are pretty much lined up - about 3 deg advanced on the exhaust and 2 deg advanced on inlet. That isn't going to cause a problem of this magnitude and would be from when I set the lift @ TDC when rebuilding the engine.

I've had a borescope down the plug holes and all looks well, so running out of ideas. Leak down test is next :-(

 

Checked valve lift with a dial gauge and found the exhaust cam is actually 14 degrees advanced at TDC from its true position. Testimony the marks on the cam caps are less than precise.

Need some gasket sealant to re-seal the cam cover, so will report back tomorrow. Hopefully things will now be OK.

 

I rechecked again this morning and realised the belt tensioner design allows the marks to move a lot when you turn the engine backwards and don't take up the slack. (my previous experience is of race engines whihc do not have self adjusting belt tension)

So I've done it all again and it seems things were not wrong in the first place.

I've now metered the fuel delivery, and swapped the coil pack and leads across sides and things are no better.

When its running it seems removing the injetcro plug on 2 and 3 has a far less significant effect than on 1 and 4. So I'm off to borrow a leak down tester.

Now looks like HGF even though the gasket was new. I've been told this is very unusual if true.

Will also check I have not bent any valves whilst originally timing it up, just in case, but it was OK on leak down when it was all reassembled when the engine was in bits, so I doubt this is going to be the case.

 

OK here's an update.

After the cam timing investigation, the engine ran no better.

So I have now completed a leak down test on the bank in question and it has only 2% loss, so it appears the mechanics are 100%. This would defo highlight any badly seated valves or any HGF.

I'm about to get a compression tester on it as this will give a comparison of the breathing on each side - i.e should highlight any mechanical (cam) timing issue - even though I've checked it twice.

SoCal1, the ECU does not do any error trapping beyond issues with sensors. The engine is only using the crank sensor and a throttle pot to drive ignition and injection, air temp, water temp etc only being used to drive the compensation maps. Its very simple which is why I chose to use this over the unnecessarily complex Motronic.

The ECU does however give real time display of all the parameters it is using such as injector duration, ignition advance, throttle position etc, so I can see what is happening. All this looks good.

I'm, also using a wide band lambda probe to display the mixture - its always in the 12:1 range on both banks - the exhaust system is discrete straight through for each bank - no H-Pipe like the OEM system.

So I am baffled.

Will report the results of the compression test shortly.

 

No the throttle position sensor just gives the ECU a signal to drive the fueling under load. At idle it serves little use other than tell the ECU which fueling value it should use in the map.

The actual throttle bodies have been balanced using an synchrometer like this: http://www.webcon.co.uk/images/products/9800100000/9800100000_1_hr.jpg


Anyway the compression test results are with us and this yields another weird anomoly:

Cyls 1-4 (the cool running ones we are trying to fix the problem with) all around 150-160psi

Cyls 5-8 - these are the side which appear to run fine.... 70-80psi on all 4.

So what do you conclude from this??