Strange Power drop on dyno

[vandor]

Hello,

I built a Fiat 2000cc twincam engine for a friend/customer this spring/summer, and this fall
we have been tuning the programmable fuel injection and working out other little problems.

The engine has 10.1:1 compression, ported head, valves ~1.5mm larger than stock, 304 degree
duration camshafts, custom made connecting rods, lightened crank, and IDF style throttle bodies
with programmable fuel injection. (I wish we would have bought forged pistons).

During our first dyno day the engine performed as expected once we tuned the proper air-fuel
ratio, ignition timing, and adjusted cam timing. The only restriction seemed to be the tall
air horns inside the air box (the top is removable, so we could test it open).
Max power was at 6800 rpm, about where we expected it to be, and we were happy that we
only needed shorter air horns. One of the things we learned is that max power needed a
lot less timing than we thought. Max advance was 28 degrees from 3000 rpm on. Anything
more than that lost power.

About a month later we went back to the dyno with shorter air horns in the air box, and after
a few runs realized that possibly it was not only the tall air horns that restricted flow, but
also the hose to the air filter. This session was a mixed bag, as we were not able to pull
consistent runs, it seemed something was off but we could just not figure out what it was.
We would do two identical runs, and the AFR would be a full point off!
At the end of the day we gained 2 hp compared to the first session, so we were not too
worried.

For the third run we made a new tube for the air filter with only a single long smooth bend,
and hoped it would not cost us any power. In the meanwhile I have driven the car ~200 miles
to tune the part-throttle part of the fuel mapping.
Surprisingly the engine was 20 HP down on power, and maximum HP occurred around
6000 rpm instead of 6800. We tried it with the top off the air box to eliminate any of
the intake plumbing resistance, but it made virtually no difference. The engine is smooth,
air-fuel ratio is fine, but the power is not there. We tried adjusting ignition, cams, etc,
but since we already optimized it all in the last runs, we always lost power after an
adjustment.
I ran a leakdown test, and the cylinders were all between 3-5%. I could be better, but I
do not think it explains the power loss. Cranking compression is only ~120 psi, most
likely because of the large overlap of the valves (?).
I plan on checking that the cams have not worn and if there is any blockage in
the exhaust. Any other ideas?

Tip for using fuel injection with individual throttle bodies: we had to put a 0.6mm
restrictor in the vacuum line to the Manifold-Air-Pressure sensor to keep the vacuum
signal stable, without that tuning the part-throttle fuel mapping was impossible.
Thank you,

Csaba

[Guy Croft]

Crikey!

I'd be for pulling the motor and inspecting piston-bore clearances and the rod bearings.

G

[vandor]

Guy,

Do you think the oil got diluted with gas because of incorrect mixture while tuning the FI?
I hope not :-(
I changed oil before the last dyno run and it did not smell of fuel - if that means anything.

However I found that one of the round perforated baffles (tube) in the baffled exhaust tips has
came loose and has fallen down so it is in the path of the exhaust. I am not sure if it
would make enough of a flow difference, but it needs to be fixed.

The baffled exhaust tips look restrictive by themselves, as their outlets (two of them)
are 1.3" (33mm) in diameter. Their combined area is a lot smaller than that of the
2.25" (55mm) diameter exhaust pipe.
Since we need to change it, is it better to have a resonator and a muffler, or a muffler
with baffled tips? A muffler alone is too loud, I have made that mistake several times.
Thank you,

Csaba

[Guy Croft]

Power dropping and dropping but A/F fine? The latter doesn't point to back pressure Csaba.

Pull that motor now is my strong advice! That's what I'd do - or at least get the sump off and the rod caps - it could be a bearing tightening up.

G

[vandor]

Guy,

Thanks, I will discuss with the car owner. Dropping the pan would not take that long.
I really hope you are wrong :-)

Csaba

[Guy Croft]

Check the cam lobes too,FWIW I once had a pair of cams 'wiped out' within 20mins of startup because of excessive back pressure because the ex pimaries on the 4-1 were way too short.

G

[robert kenney]

Chaba,
I know GC has posted his experienced opinion but thought I'd pass on this idea. I have run into similar situations. I feel the 120 is low but of course it depends greatly on where you placed the IN LC. Low cranking compression can be attributed to a retarded intake cam not so much the overlap. Running an advanced intake cam EG 105 CL as opposed to 110 will increase overlap but also increase cranking compression due to the intake valve closing earlier on the compression stroke allowing less air pressure to be lost back out the intake valve. I don't know your cam timing @ say .050 lift but I would expect to be seeing a minimum of 160-175 psi. I am sensing a 42/82 clone cam set. If I recal correctly My sisters Fiat was 180psi with 10.2-1 after being broken in.

If your intake cam has slipped retarded you would see broadening and lowering of the torque curve and a lowering of cranking compression. Did you notice a power band shift of any sort other than rolling off earlier.

Leak down is a bit high but depending on breaking miles could be normal. I would not attempt dyno pulls untill 1000 miles but I am a break in easy kind of guy. If you can listen to the crankcase breather while leaking down you may get a hint to point of leakage is rings or valves.

Good Luck with the bearing check.
Robert

[vandor]

Guy,

I have checked the cam visually and 'manually' (felt the lobes) and compared them to an unused set, and I cannot see or feel a difference. The head comes off on Saturday, so I will be able to remove the cams then and check with a micrometer.

Robert,

Yes, the cams are 42/82s. A gear skipping a tooth would make sense, as the whole power curve has shifted down, although I have only seen this happen on old engines and loose belts. Scary thought....

Today I dropped the pan and removed #1 and #4 conrod caps. The bearings are scored... This is a standard size crank with standard size bearing and custom made connecting rods. Why would this happen? Is the big end too small? I never measured them when received, assumed they were correct :-(
The bearing gap as measured with a plastigage is 2-3 thousands of an inch. Not very precise, I know. When I built the engine all rod bearings were at 3 thousands of an inch. Being a standard size crank I did not think the clearance could be too small...
Thanks,

Csaba

[Guy Croft]

The bearing scuff is likely oil starvation or fuel contamimated oil (overfuelling maybe caused by cam timing slip) and not the cause of the power drop.

If you find the cam timing out, 1" belt might be a good idea.

G

[vandor]

Guy,

A bit of an update. We pulled the engine :-)

Before that happened I did a compression check with a GOOD compression gauge. Compression is 150 psi. A bit lower than what I'd like, but much better than 120... A played with cam timing, but could not get any better than that, only worse.

We found even worse scoring on the main bearing, see pics. On the very first dyno (rolling road) day many months ago the gasket between the block and oil filter block failed, and we lost about 1.5 liters of oil. Did not notice it until after the 'pull', and we did not completely loose oil pressure, so at the time we thought all is well. Now we suspect that oil pressure was inadequate for high rpm/high load, and that is what caused the scoring.

However, I don't think this explains the power loss.

The other interesting thing we found is the light color of the exhaust valves. I know we had some high EGTs during part throttle tuning, is that what caused it?
Thanks,

Csaba

[Guy Croft]

What are the ring gaps?

G

[vandor]

Guy,

I will measure them as soon as I get the parts back from the machine shop, they are measuring the block bore and big-end diameter.
Thanks,

Csaba

[Guy Croft]

I wonder if it's been running over-rich and the gas has contaminated the oil and caused aggravated bearing wear, maybe with ring wear too, hence dropping power. I'm concerned about the modest advance that was put in (28 deg and over that 'lost power'), that can be a sign of an over-rich mixture. Many things incl ex back pressure can make a motor run rich.

I may be wrong.

G

[robert kenney]

Just an observation, cylinder 1 and 4 look to have been fairly lean don't they? White exhaust valves and plugs. Suppose WOT was fat and cruise power settings were lean resulting in the white.

Robert

[vandor]

Hello,

The ring end gaps are all between 0.019" and 0.020" (0.5mm) for the upper and middle ring. The spec calls for .015" for the top ring, and .013 for the middle ring for naturally aspirated engines, or 0.020"/.018" for turbo/nitrous (nitrous is in future plans).

Would you recommend going back with gapless TotalSeal rings, standard rings, or some other type of high performance ring?

The mixture at WOT was between 12-13:1 air/fuel ratio, so it was not too rich there. It was rich at cruise until recently, however the FI system has a knob to instantly change injector pulse width, so we were able to lean it out at cruise - most of the time. Gas diluted oil is a plausible explanation, I am just surprised that it would happened in so few miles.

As far as timing goes, the shop where we did the testing builds Hondas, and they said it is normal for Honda engines to only need that much advance, so we just accepted it.

Robert, I looked for some info on the white valves, and it may not necessarily be an indication of a lean mixture. It colud be from high EGTs. We had the header glowing at one point while tuning because of too little advance at cruise. I added more vacuum advance. The problem was that I originally copied Fiat's vacuum advance curve for the Spider 2000, however this engine operates in a much narrower vacuum band because of the cams. I may have to take Guy's suggestion and add more 'centrifugal' timing aswell.
Thanks,

Csaba