Odd apparent ignition timing requirements

[JulianEdgar]

I am feeling my way with tuning my Honda Insight engine, but the ignition timing seems very odd.

The engine is a 1 litre, 3 cylinder that revs to 6000 rpm. It uses a particular form of VTEC that switches off one inlet valve at less than ~3000 rpm, giving better swirl.

I have turbocharged the engine.

Idle ignition timing is 12 degrees BTDC.

At the moment, when boost starts coming up at around 2000 rpm at full throttle in high gears, I need to run very retarded timing or it will detonate. As in, much retarded over idle timing! For example, 1 and 2 degrees BTDC at 1750 rpm, 110 - 120 kPa absolute.

Intake air temp and fuelling are fine.

I have never heard of any ignition timing requirement like this. Should I go looking for problems (I have checked CRIP only at idle), or is it quite possible that this ignition requirement actually exists?

Edit: not sure if it's relevant, but I am running a very high energy ignition - direct fire MEC-723 Bosch coils with around 120mJ of energy per plug.

[stevieturbo]

It does sound more like actual timing at the plug is not what you're asking on screen.

It would definitely be worth checking with a light at those loads/rpm, probably best not with the CRip feature in case it masks something.

[JulianEdgar]

I can't check the timing at that load and rpm combination - I am doing tuning on road, no dyno.

Timing when set to 12 degrees test advance doesn't change with free reving.

Timing appears to change as designated in ignition table when free revving and working off table.

Edit: 10.8:1 compression, peak torque at 4800 rpm though

[JulianEdgar]

OK, standard car apparently runs up to negative 10 degrees of ignition timing when lugging up hills in high gears.

So apparently that's what it needs....

Perhaps it's because of this?

The IMA gasoline engine's 3-cylinder engine block is a highly compact, aluminum-alloy die casting. An unusual feature of the engine's design is that the crankshaft axis has been shifted, or offset, 14 mm, relative to the cylinder-bore axis. In other words, the crankshaft does not sit directly under the cylinder. This was done in the interest of minimizing friction caused by the side thrust of the pistons against the cylinder walls, just after top-dead-center, as each piston begins its descent on the firing stroke.

In a conventional engine, piston-side thrust and the friction it generates are the result of the crankshaft's resistance to turning at this point. This resistance is the result of the small angle formed by the crank throw relative to the centerline of the piston and cylinder. The IMA engine's cylinder bores are offset to be over the crank throw at this point, so the piston and connecting rod push straight down, thereby minimizing side force on the piston.

Cylinder-bore offset in the IMA engine accounts for as much as a three-percent reduction in internal friction.

[AdamW]

I can't check the timing at that load and rpm combination - I am doing tuning on road, no dyno.
Timing when set to 12 degrees test advance doesn't change with free reving.
Timing appears to change as designated in ignition table when free revving and working off table.
Edit: 10.8:1 compression, peak torque at 4800 rpm though

Your timing sounds about normal for high comp, pump gas, turbo application.
For example, Mitsubishi lancer Evo 5 with 9.3: comp and 1bar boost will only take about 10deg at torque peak.

[JulianEdgar]

Spent the morning tuning.

Did more work on cold start; did a lot of work on acceleration enrichment; lifted boost to 50 kPa (~7 psi); tuned timing and fuel for this boost. Am running with the intercooler water pump switched off to create high intake air temps, so I can tune for less-than-optimum conditions - however, hasn't gone over 50 degrees C (ambient about 10 degrees C). When I do turn on the pump, it pulls the temp down to about 10 degrees C over ambient very quickly.

However, I am not all that happy with today's results. At around 2500 - 3500 rpm, the engine sounds like it is detonating. Take the timing out in this area and it's as flat as a tack. I also have a misfire at 50 kPa boost at these revs, all the world like there's a wrong timing or fuel number in the M400 maps - but there isn't.

At the top end at 50 kPa boost she's running really well.

So at the bottom end, off boost, she's really nice. And at the top end, with lots of revs and boost, she's pretty good... but in-between she's either making a sound that is very much like detonation, or (with timing suitably retarded) is flat and misfiring.

Peak torque is at around 5000 rpm in the standard engine, so I figured it would be at these sort of revs where maybe cylinder pressures were getting too high and I might have problems. But it's at mid rpm where there are issues. VTEC (where in this version the car switches from 1 inlet valve operation to two inlet valves operation) is set at around 2700 rpm; I don't know if that change along with boost means that something can't keep up. (But then it's at higher revs than this point where the major issues are.)

Took out a plug and with a really good magnifying glass you can see some really tiny deposited (aluminium?) specs on the plug - perhaps indicative of the detonation I have heard. Plug electrodes not eroded or anything like that.

Need to get my assistant in the car to watch for error codes (ref / sync?) or something else that I can't see when I am also driving.

But I am a bit stumped - you'd think an ignition breakdown would be worst at high revs, full boost... I don't think I've ever come across a problem of quite this complexion before.

Suggestions welcome!

[stevieturbo]

As a temporary test, assuming ignition timing etc has been 100% confirmed as accurate at all times.

What about chucking in some toluene or similar octane boosting agent for a few runs ? To see if the same noises can be heard.

Not a long term solution for you, but it would offer a quick test to help eliminate any detonation, if that is indeed what is being heard.

[JulianEdgar]

Thank you.

I am deliberating running on fuel that is lower in octane than I can get (again, tuning for worst case scenario) so I can easily put some higher octane fuel in it tomorrow.

[David Ferguson]

Maybe that third valve doesn't have enough valve spring force, and your boosted condition (vs. the normally aspirated case it was designed for), is opening the valve -- what you're hearing might be essentially the opposite of valve float, the valve being slammed closed by the compression stroke. Perhaps reducing the timing is lowering boost or raising the exhaust pressure so the delta pressure across the valve doesn't allow it to open.

Perhaps you could change the vtec to always control that third valve and see if that changes things.

Sounds like quite the challenge to figure out.

[tepid1]

Have you had anyone check over your calibration file to make sure that it's not some sort of correction that is messing you up?