.
I also noted a comment you made elsewhere stating that the Honda ECU works in closed loop only at small throttle openings?
A couple of years ago, I had a CTSC supercharger on my car and the power would totally vanish on track after a couple of hot laps.
I suppose it's the MAP sensor that told the ECU to go in a safer mode of operation when the IAT's went sky high?
Many thanks for your great explanations!
Not really 'only at small openings'. Better to think in terms of 'not at large throttle openings'. On my 'other car' I have the ECU set to go open loop at MAP values above 80 kPa or a TPS greater than 70%. Its pretty common for older OEM ECUs to switch out of closed loop at large throttle openings or high MAP values for three reasons:
1. the older OEMs use narrow band sensors which only work at a 14.7 AFR and most OEMs run their engines (even emission controlled engines) at something less than 14.7 under wide open throttle. If they tried to run closed loop at wide open throttle they would be forced to run at 14.7.
2. there is a time delay between the injection of fuel at the injector and when it shows up at the O2 sensor to be measured. At wide open throttle the engine may be accelerating quickly so the exhaust gas being sampled may no longer reflect the current engine operating condition (the engine may have transitioned to a new operating point on the fuel map). There is no point in correcting fuel mix when the operating point has changed. Wideband controllers add a little extra time delay into this equation.
3. O2 sensors and controllers do go 'off' their calibration. Incorrect O2 measurements at part load operation result in poor performance. An O2 measurement that caused an engine to run lean at wide open throttle can lead to melted pistons. O2 sensor problems that result in calibration errors can be harder to detect than outright failure of the sensor. Also, for those fans of wideband O2 sensors the O2 measurement is very sensitive to the temperature of the sensing element. The wideband controller measures (infers) the sensor temperature and attempts to control the sensor heater to maintain accuracy. However, during rapid wide open throttle engine transitions the exhaust temperature changes quickly presenting a control challenge for the O2 controller causing potential errors. I think EGT probes still remain the go-to for wide open throttle monitoring.
I don't think the MAP sensor (pressure) would have caused the ECU to go into fail-safe mode. I suspect it was the MAT sensor (temperature). There are four possible causes for the problem;
1. Honda put a safe upper temperature limit feature in. That would be more typical of a ECU set up for an OEM forced induction car and not a feature that would normally be incorporated into an normally aspirated car like the NSX.
2. The ECU has a sanity check on the sensor value and your manifold intake temperatures hit a value that Honda figured would be impossible on a naturally aspirated car so they triggered a fail safe mode of operation.
3. Temperature sensors are resistive temperature devices and they have a highly non linear response curve (of the general form R = a/T^2 + b/T +c). At higher temperatures it takes a bigger temperature change to cause a slight change in resistance or a slight change in resistance 'looks' like a big change in temperature. If Honda only calibrated the sensor between -40 C and 60 C and you took your MAT up to 100 C the calibration curve may be slightly off which can lead to large errors in the 'measured' temperature. If the ECU 'measured' a very high temperature it will reduce fuel delivery in line with the equation:
pulse width = k'/T
This will result is loss of power and a potentially dangerous operating condition (running lean) if the actual air temperatures are less than the measured air temperatures.
4. Good old physics. n =PV/RT says that as the air temperature goes up the mass of air entering the cylinder is reduced and you will make less power. However, the T in this equation is based upon absolute temperature (deg Kelvin) so a change in MAT from 80 C to 120C would only be around a 10% drop in power and the power loss would tend to be gradual as the temperatures increased rather than all of a sudden.
Its more likely that the ECU was coded to interpret the temperature as non sensible and intervened to put the engine in a fail safe mode. If you got a copy of sr5guy's definition for the NA1 ECU, you might be able to figure out exactly what actions the ECU takes for elevated air temperatures.
Andreas also got close to that with this modifications, but he had a tune done. Andreas kept his Honda valve springs and has been driving on them problem free for over 15 years. What this means for my build is that the stock ECU can adjust enough for these mods even without a tune.
