3SGTE torque falling flat on its face after 7000 rpm with kelford 262 high lift cams

Just attempted my first tuning on a dyno with supervision of a local tuner.

Took quite some time to sort out various problems so there wasn't enough time for proper tuning and the map is still very rough and conservative with a very rich AFR and retarded ignition.

However, even with the extremely conservative tuning the number from the dyno is looking abnormally low with torque falling flat on its face after 7K rpm. The tuner suggested the cams may not have been properly degreed and suggests I should call it a day and go back to confirm everything had been setup properly rather than wasting more time and $$$ on renting a dyno to it is likely I will have to start allover again after adjusting the cams.

I am new at this but building up my engine, I have degreed the cams several times finding true TDC and adjusting the cam wheel until the valve opening and value closing time both matches with the cam cards. I am quite confident the cams are setup correctly given I have measured both the opening and closing for confirmation.

But other than the camshaft degreeing, we are all out of ideas on what is leading to the significant drop in torque at high rpms. Could it really be just the result of a rough tune?

My setup is as follows:

Gen 3 3S-GTE engine

Kelford 262 camshaft with 10.5/10.0mm lift

BC valve springs

Cometic 1.52mm head gasket (~0.3mm thicker than stock)

Wiseco 86.5mm forged pistons (9.0 CR when used with stock head gasket vs 8.5 stock)

Manly H beams

Precision 5558 turbo with .48 AR

Was there a boost spike and fall off or wheel slip through the middle of that dyno pull? If not I think you need to improve ignition first, looks like weak spark. Gap plugs first then replace coils if the shape of the torque curve doesn't smooth out. Assuming ignition timing is sensible but conservative.

I don't think there is a problem with the underlying torque curve vs rpm ignoring the fluctuating bit. I have two friends with sr20s with similar lift 272? Cams and they don't have that much carry into the top end, although I think they are gt28rs turbo setups so a bit smaller. In saying that your turbine housing is pretty tight. Have the heads been ported or intake changes to suit higher rpm flow? You can probably shift it a little bit retiming the cams if top end is most important to you.

I say gap your plugs a bit and clean up fuel a bit more if you can datalog/drive it on the road and confirm it pulls cleanly on the road thn go back to the dyno. If the ignition is weak adding boost won't increase power much at all.

Oh yes, forgot to mention that I have modified the engine to run on a coil on plug setup with crank and cam sensors using 1zz coils.

The coils could very likely be an issue as I have experienced heavy miss-firing previously at high boost when first converted to COP. Had to gap down the plugs to 0.65mm which appeares to have solved the problem. Maybe I need to gap them down a little more (down to 0.6mm).

It is a hub mounted dyno so there isn't any wheel slip. Boost is smooth. I'll take a look into the ignition problem.

Sorry one more question though. If ignition is weak and the mixture did not ignite properly, shouldn't it be reflected in the AFR where it would appear to run lean due to the unburnt air?

Problem is, I am not seeing this in the logs.

You're running 1ZZ coils? I had a 1ZZ Corolla. 1ZZ is a nonturbo 120ish horsepower engine. What ECU are you using? Why not switch to stronger coils?

Misfire shows up in high HC emission (due to unburned fuel) or low cylinder pressure or low exhaust runner temperature on the misfiring cylinder. Most vehicles/chassis dyno facilities don't have that kind of instrumentation. Wideband might show something, it might not. It only detects oxygen, it doesn't detect HC. I wouldn't go off only that.

Yes I am on 1ZZ coils. They are used because they are cheap, have the proper length and adaptors are available that makes them direct bolt on.

People are hitting 500+ HP with these coils so I thought they should be sufficient. But it seems like I may be wrong. Anway, I'll try gapping down the plugs further to see whether it would make any difference first. If yes, it is definately the coils.

Are you running same dwell as people using them in high boost applications?

My dwell times are based on the following sources:

http://www.mr2oc.com/12-after-market-engine-management/582073-1zz-coil-dwell-times.html

http://forums.linkecu.com/topic/6432-1zz-smart-coils-on-2jzgte-g4storm/

In fear of frying my coils with the timing from the first link my dwell is being set somewhere between the two.

I guess start with gapping the plugs then try to clean the fuel up. Make sure power supply and earth for the coils is good.

I just check your rom, and is gonna be difficult to make power high in the rev range cause you have very little timing gain, in some parts you even have less timing at 7500rpm that at 7000rpm, and higher in the load you have the same ignition at 7000 or 8000rpm... That could explain some things.

How is EGT at high rpm? what boost level are you running? is difficult to say without a log file but you may have lower boost at high rpm, that and that you are probably running a little bit richer at 7500rpm that a 7000rpm could explain the power drop over 7.000rpm.

I will work on the tune to be honest.

Can you post a log file?

do you have a log of a run and also a AFR/Lambda trace? also, how much boost are you running? how long are your pulls on the dyno?

You've set up ignition retard after 110 degrees C and your coolant temp max in the statistics is 118 degrees, if you have a log of a pull we'd be able to see your coolant temp during a pull.

Your torque is tailing of after 4750 RPM with that size of turbo it should still easily be climbing I'm guessing you've been overly cautious with the ignition timing and fuel delivery leaving a very rich mixture that is getting ignited late, this can generate extra heat which could be causing the ECT ignition retard to kick in but without a log it's a total guess

I have foolishly closed PC link without saving the logs but did manager to salvage a shot session from the ECU logs after I got home. Boost was stable at between 1.1 and 1.2 bar but the AFR was quite a bit richer than I anticipated.

Being my first attempt at tuning on the dyno ever, the tuner/dyno owner took control of things and actually made up a very conservative map before I start my tuning on the dyno. Rather than starting the tune with the approach I learned from HPA, I was told to enable boost control and go straight to WOT first. While following his instructions, I only concentrated on getting the boost control right and making sure the AFR does not go lean over the target on the time plot without checking what the actual AFR is. I did not touch the AFR as it did not go lean and only after looking at the logs now I noted it was quit a bit richer than the target and should have been adjusted.

So after reading the earlier comments here and no another thread where I had concern over the high knock level detected (https://www.hpacademy.com/forum/practical-dyno-tuning/show/high-knock-level-reading-on-certain-cylinders), I expressed my concern with the tuner but he still thinks there is something wrong mechanically which had caused the significant drop off in torque. His logic was that the rich AFR and retarded timing was over the whole RPM range, meaning although it would result in an overall low power output, it should not effect the shape of the plot. My plummet in torque at high RPM only suggests there is something mechanically that is not setup correctly.

I am going to take up everyone's comment here to check my ignition hardware, do some tuning on the road to correct the rich AFR and advance the timing a little before heading back to they dyno to see whether it made any difference and resume my dyno tuning from there. Hopefully I can take more control over the dyno tuning process next time.

Attached are the logs I salvaged to hopefully provide more information for more inputs. Thank you for everyone's analysis of the problem and suggestions so far.

Just took a look at the catch can and noted something quite worrying. There was quite a bit oil in the catch can. Probably around 40 to 50 ml. The engine was newly built and did under go semi aggressive procedure run-in procedure where load was put into the engine gradually through accel and decel cycles and going up and down slop etc to build some cylinder pressure or vacuum. Would this semi-agressive run-in procedure produce enough blow by to accumulate that much oil in a baffled catch can before the rings properly seal or is that an indication of some more severe problem with the engine causing the low power output?

The odo meter was broken but I have gone through almost a tank of fuel before heading to the dyno and around one third of a tank while at the dyno so I did not run too many miles to accumulate that much oil in the catch can.

I will most definitely carryout a compression test tomorrow after work.

Usually you run more ign adv after peak torque, so you usually see let say 14° at 6000 and 17° at 8000rpm (is just a example), thats common on most turbo engines. I will say that 11° at 8000rpm on 14psi of boost is very very conservative, check knock and see what happen when you add some timing, and yes is a little bit richer on the top to, that will produce less power over 7000rpm that is where the engine lost quite a bit. You can tune the afr on the road no probem, just be careful. About the timing, you can add 1-2 degrees at top and those 10-15whp of gain you will probably feel it. Then again watch knock, and you will probably improve the tune quite a bit.

What is you ambient temp? cause the iat goes up quite easily, 51°C is a little bit on the high side, specially cause is going up with the rpm, maybe a not so efficient intercooler?

Ambient is around 20C. Being a mid engine car with a side mounted intercooler, cooling will always be a problem especially when on a dyno without a dedicated fan blowing at the two side vents for the intake and intercooler (the intercooler itself does have a fan though). I am already using the largest side mounted intercooler one can fit in an MR2. The cooler itself should be quite efficient too as it uses a quality core from Bell. The engine bay is cramped in an MR2 and there is really not much else I cold do to cool the intake charge short of water injection or hacking into the trunk. Still there are many making good power on an MR2 with a side mounted intercooler.

probably on the road you have a much better condition... I had those problems in formula cars with high ECT.

Agree with other posts that assuming that it's not knock limited, that the engine could benefit from more timing advance. Typically ignition timing is reduced in the area of peak torque and increases as RPM goes up. Again this is all under the premise that the engine isn't knock limited.

I attached a screenshot of one of your pulls that I marked up to show you the area where you'd like see more power after adding timing.