Honda Insight w/Acura TSX engine - Ecomodding

[Ecky]

I think the Accord version of the engine has EGR, I'll have to look into how it was done. I don't think the block is any different.

[Ecky]

Looks like I'm in luck!

The K series are extremely modular engines, and were used in tons of different applications. Parts from different generations are all cross compatible. Oil pumps or cylinder heads from a 2002 bolt onto 2015 engines, and vise verse. And, it appears the EGR system is *probably* the same way.

Here's the water pump assembly from the Accord, which has EGR:



Here's the TSX, which my engine is out of:



The TSX even has the EGR chamber, stamped "RAA" same as the Accord:




Incidentally, the EGR valve used is identical to the one on the Insight.

So, it looks like I'm going to get a water pump housing from the 2.0L RSX Type S for the oil warmer, the water outlet passage from an Accord 2.4L, and both should bolt onto and work with my TSX.

[serialk11r]

Nice. It's hard to tell from the picture, can you manually control it? Does the TSX have wiring to control it with a hypothetical ECU flash?

With external EGR, the ideal valve timing changes a bit, since you'd want the intake cam retarded and the exhaust cam advanced to reduce pumping losses, and supply EGR flow externally. The literature suggests 15% EGR dilution is usually not an issue, 10% is definitely not an issue, 20% is starting to push it.

In the absence of internal EGR, 15% dilution would heat the air up enough to raise MAP by around 25%. I would guess that the actual improvement would be closer to half of that, which still isn't bad. Might turn out to be something like 2 mpg.

[Ecky]

I'm not using a TSX ECU, I went with an early RSX ECU because it was most compatible with the Insight body wiring, and was drive-by-cable. The RSX ECU has no EGR output, unfortunately. I might be able to repurpose another output, but it depends on how it's controlled. PWM? 0-5v? Probably not just on/off.

Manual control might be fine. Maybe leave the valve open or at a fixed duty cycle below, say, 750mbar - that would be very easy to program.


At 0 degrees advance it appears the valves close slightly early - fueling values are lower than at 10 degrees advance. 10 degrees gives the highest fueling values at low MAP, suggesting it's the crossover point where the valves are opening late enough to get the most air in with minimal exhaust dilution. Hondata suggests to use around 30 degrees advance at low load on these engines to get some hot EGR effect. At 40 degrees the engine starts to stumble at low load operation, and fueling falls dramatically - MAP is definitely higher, but combustion quality seems to suffer.

I put a physical stop in to prevent more than 40 degrees advance, as I believe around 45-50 the high speed cam profile can cause valve/piston collision, and it seems with this intake/cam combo, having more than 40 degrees of advance doesn't provide any more WOT power.




0 degrees: -3.9% fuel
10 degrees: peak fuel
20 degrees: -2.2% fuel
30 degrees: -14.2% fuel
40 degrees: -30% fuel



I haven't had my car on a dyno yet. With my crude record keeping of fuel economy back and forth on my commute, I have not noticed any significant fuel economy difference between 10 degrees and 30 degrees advance, and perceived a slight reduction with 0 cam advance (early valve closing). I do notice that 30 degrees advance has significantly more exhaust noise (or maybe I'm hearing it from the intake?) than 10 degrees, so I've been driving around with 10 degrees intake cam advance.

[serialk11r]

Can the advance go negative? On my FR-S it did. I found best fuel economy at -20 degrees, while stock was -10 or so.

I think unfortunately EGR control is a little harder than what you described, most importantly you need to prevent overdilution at low load where there's a lot of intake vacuum that would draw a lot of exhaust in. It should be completely off at idle, then ramping up to max EGR fraction, then decreased back to 0 at high load.

Easiest is to just have it open to a preset position in your cruising rpm range and shut everywhere else. That way it doesn't work against you most of the time.

[Ecky]

The total range of adjustment is normally 50 degrees, for mine 40 degrees from a mechanical stop I added. It can't be negative, and I don't have an exhaust cam phaser. I'm unsure if 0 degrees on my engine corresponds to 0 on yours.

At 0 degrees cam advance I found my fueling is somewhat erratic, in a way it isn't with 10 degrees advance. It seems to swing +/-4% without much rhyme or reason (in the same cells) whereas +10 is generally within 1%.

[Ecky]

Went for a drive today, one of the first days this year it has been clear and over 50F. I broke 70mpg on several occasions as per the display, and the tires I have on right now are 1.3% larger than stock, so I may have been over 71 at points.




My last fillup was up by around 9mpg when compared with just 2 months ago. Can't wait to see how it does in high summer!

EDIT: Plus gas is 40% cheaper.

[serialk11r]

Oh crap I just remembered, Hondas are all alpha-N right? That means EGR is trickier, since you are changing MAP which affects fueling, instead of relying on MAF which is unaffected.

[Ecky]

Sortof, it's MAP based rather than MAF, it would swing rich until the O2 sensor corrected, unless EGR were consistent.

[serialk11r]

Sounds like a nightmare to correct for, maybe a warm air intake + mildly lean burn via narrowband sensor emulator would be the way to go.