Chasing a flat spot

[JulianEdgar]

My turbo Honda Insight (running M400) has a little flat spot at 3800 rpm, full throttle. Second gear is when it is most noticeable.

I noticed boost dropped a little at this spot so put in a special site to address this via a slightly higher boost control duty cycle. Boost no longer falls, but flat spot still there. Nothing obviously different in timing or fuel at that spot.

I tried some variations in fuel and timing at just that site and by leaning it out a lot I was able to improve it (just the one site!) but then I noticed that in different gears (eg third) the car was going too lean as it went past that spot - I imagine because it is there longer. I then returned mixtures and timing numbers back to how they were previously.

I can only guess that there's some sort of odd intake resonance at that combination of revs and boost (and rate of change of engine speed?) - but that doesn't help me fix it!

It's noticeable enough that the car feels poorly tuned - anyone got any ideas?

[enniorancati]

Can you please upload a log file?

Thnaks

[stevieturbo]

Is this were any VVT does anything strange ?

Or where you had very reduced timing levels ?

[JulianEdgar]

I'll upload a log when I am also logging air/fuel ratio (currently reading that off a separate PLM), but am adding it as an ECU input shortly.

No, the problem point is well away from VTEC switching point and timing isn't anything odd.

[stevesingo]

Unfavourable resonance in the intake?

More boost will not improve such a situation as the resonance will be there regardless. It may change frequency, so appear at higher or lower rpm dependant on pressure in the intake tract.

[Scott@FP]

Do you also have the same symptoms at 7600 RPM?

Needs better mechanical filtering of the MAP sensor, you are hitting a resonant point related to the frequency that the MAP is sampled at (100Hz) and the engine type (2 stroke or 4 stroke) and cylinder count. Will be worst if trying to use a TMAP in the plenum.

Log MAP at highest resolution and drive slowly thru the affected RPM, log will show a resonant wave pattern that gets stronger as you approach the resonant point, and reverse as you go above it.

here is an example, this was eliminated by better mechanical filtering, 3000 RPM 4 cyl 4 stroke, the input freq and the intake pulses match freq at 3000 RPM and cause all sorts of problems.

If you do the math a 3 cyl 4 stroke will first order freq match just under 4000 RPM. 100Hz/3=33.3Hz. 33.3Hz = intake pulse rate of 1998 RPM = 3996 engine RPM (4 stroke). 4 cyl 4 stroke 100Hz/4=25Hz=1500 intake pulse RPM = 3000 engine RPM.

I tried a TMAP once, problem was so bad it was untuneable and undriveable at the first and second order freq match points.

[JulianEdgar]

Thanks Scott - very interesting. Cannot test at 7600 rpm - factory redline is 6000 rpm and the rods are tiny so I am reluctant to go over 6300.

By mechanical filtering of the MAP I assume you mean the insertion of a damper of some sort? Any suggestions on how to do that?

[Scott@FP]

A foot or two of hose, a small orifice (.020" or so) and a 10cc or so chamber filled with SS wool.

You don't want to use a lot of input filtering on the main fueling sensors in software because response gets slowed down too much, leave the software at the fastest, least amount of filtering you can get away with and use mechanical filtering.

At any rate if you log MAP at 100-200Hz and drive slowly from 3600-4400RPM and your MAP data looks similar, that is your problem. Also keep in mind the first order will be the worst, but it will also show up at 2x and .5x the first order frequency. In the 4 cyl example above 3K and 6K had severe issues,1.5K, 4.5K and 7.5K less severe but noticeable.

This applies to ALL ECU's that use a fixed sampling frequency.

[stevieturbo]

A 0.6mm or 0.8mm MIG tip shoved up the inside of the line going to the map sensor makes an effective damper.

[JulianEdgar]

Thanks very much for the advice here. I got a map sensor damper from a car at the wreckers (sorry, can't remember the car) and it has solved the problem completely. I honestly would never have considered the map sensor as the culprit so again, thanks very much. Incidentally, other vacuum hosing dampers I also got didn't work - just the one originally fitted to a map sensor.