It's simple actually [edit: simple I said when I started this post... ai] in the end you hardly cut down on fuel. Just start with stock mappings of P28 is the best bet. I eventually ended up running a bit leaner in the lower revs but I needed more fuel in the higher revs. The equation is rather simple... it comes down to the performance, not the capacity. Due to the cam retarding and other semi-serious work I did I got in the region of 130-145 horses depending on which stage of my setup. That simply means you need the same amount of fuel approximately of the stock D16Z6 engine or even a bit more since it only gives 125 bhp stock. Fuel maps greatly resemble the actual torque curve generally.
So stage 0 I would like to call the first guess... go for P28. If it is even then stinking rich (you got 1350 cc approximately?) , my bet is the cam is not on it's best position. And therefore max power is really limited. Ok, leaner mappings will improve on that, but neh... it should not be the endpoint having maps 20% leaner or something. That would mean you only have 125*(100%-20%) = 100 bhp...
On a sidenote, in my experience even with greatly changing cam timing, the ignition timing hardly changes. I often put idle timing on 18 deg BTDC (which is max advance OEM for D16Z6 engine) and that gives me always max power ignition wise. Have detailed many cars out and often the original curves with max idle spec are optimum in performance.
Okay, so that is still the starting point, stage 0, get your P28.
Now there are two serious options to get detailing the setup out.
Either ...
...get a wideband lambda sensor and street-tune the fuel maps out. Leave ignition and cam timing... trying to 'feel' improvement gains will often result in a loss in performance since pychologic effects will tend to have you tuning the car out of the perfect zone. Like: " ow it goes fast now all of a sudden at 5000 rpm" while in reality you tuned a loss before 5000 rpm. Fast engines are often very smooth and tend therefore to be a little bit less aggressive than badly tuned engines... but unawareness of this may result in loss of horsepower. I want a car to be fast, if it feels fast as a result that is nice... but not required for me. So well... listen to the AFR signal coming from the wideband and DO NOT LISTEN to the thing between your ears
and put it for full throttle flat at an AFR of 13. Don't do 13.7 or these kind of things... most cars have max horsepower as low as 12 and only high rev Honda's can get max with 13. Doing 13.7 (heard very often in Honda community) is in my opinion a loss of safety factor to prevent detonation and such. Yes Having AFR of 12 will get you black tailpiping.... but both my cars do that at full load sometimes and I love ehm for that. I will not be the best environmentalist around here, but my engines go for reliability which one of them has proven for 355.000 km already. This is my stage 0.5 solution.
Or ...
... and this is my stage 1, the first decent ECU tune you can perform. Is bring it out to a dynocenter and iterate all your maps. Start with either the stage 0 map or if you did some streettune with stage 0.5 maps. But then you start looping. My recipe... we start to look for max horsepower everywhere in steps 1-5 (duh?):
- 1 start getting the fuel maps right for the current setup, so full load max power with current cam and ignition
- 2 then see if ignition is ok, lower a few degrees over the entire range and see were to change for the better. If horsepower worse a location, increase ignition there. Rule of thumb with NORMAL valves and combustion chambers, detonation will occur 4 deg after reaching max horsepower. Lakai his car was not 'normal' and did detonate before this point. Luckily enough I was able to hear it and abruptly stop the test.
- 1R only in case of big changes in step 2, redo fuel maps now (1), if not proceed to 3
- 3 to the first cam change, my suggestion is the smaller engines (D16 is base case, D15 is smaller, D14 then smaller, yours smallest) I notice a trend of cam retardation up to one tooth for my personal EJ9!. And for reasons of valve hitting, cam retardating after a succesful run is safer since VTEC valve hitting pistons is not possible. Only goes for SOHC VTEC cars this advice!!! I always believe valve-piston clearance is hardly an issue and I boldly change cam gears even on DOHC engines... but truth is I just might have been very lucky all these years not to have wasted an engine since extreme cams in SOHC-VTEC are measured by a friend of mine and he did reach hits pretty easy... but well, you have stock ESi cam not? No aftermarket? So... change by a few degrees cam... perhaps half a tooth max for the first attempt.
- 4 I make this another bullet, decision time. Look at the maximum performance for both cam positions attained. If one of the curves is much lower than the other your probably far away from the ideal zone. You probably figured out you should go to the other side and try a third position, redo 3 on another position and get here with another pair of curves. But when you are already close to the optimum it gets difficult. Textbook says this: generally cam retarding gives more horsepower at high rpm but with a loss of hp at lower rpm. Well: my experience tells me it's a very tiny moment this actually happens. With all the SOHC VTECs I've changed cam timing over the years and even the non-VTEC I saw that changing cam timing either increased or decreased hp over the entire range up to 7000 rpm. Right at the peak performance and then retarding it further, I often squeezed out a few hp more at 7000 rpm or such high numbers but lost 10 - 20 all of a sudden over the entire range from 2000 up to 6500 for example or something similar. So I always say there is a very definite un-debatable optimum. Despite loosing 1 or 2 tiny hps somewere in the sky I always opt for this beefier curve... I'd say the surface under the curve must be optimum and often cam timing is pretty strict in the best position for this.
- 5 continue all the previous loops again untill you get stuck at the optimum...
- 6 yes there is another step... detail out closed loop and low load areas for optimum fuel efficiency! Many tuners seem to forget this, but here is the reason my B16B can get 14 km to 1 liter and my D14 once had a setup capable of doing 16 km to 1 liter... which I think I destroyed by putting the S2 IM and TB... now 15.3 best summer average. A dyno with low load options is really convenient. You just put car at 50 km/h for a while and ask 10 hp or 5 hp or so and then see what is the leanest to get it. Do this for a whole lot of points and interpolate a bit in between and presto... part load fuel efficiency the factory cars cannot beat! Well... I sound like a salesman, often you can get, sometimes the setup will not allow for it.
So this is my recipe as how I built up experience in ECU tuning a car by streettuning and/or dyno tuning it with a strong favour for the latter IMHO. My recipe is based on speed and reliability. I often hire a dyno centre and tune 6-8 cars with CRX2. I am the guy shouting and the owner of the dyno and CRX2 do the actual work and taught me this recipe. But with us as a team we need less than 1 hour on average to completely tune a SOHC-VTEC full out to it's max and perhaps up to half an our more with DOHC VTEC including cam gears.
Tip: look at various dynographs to learn how the optimum should look like. I often instantly recognise a D-serie or B-serie Honda engine by the shape of the graph and can even tell what's been done to the engine in basic lines or when it's a fraude... here at DU we once had the non-VTEC 153 bhp case which happened to actually be (still ass-whooping 129). Your car should have lower torque numbers due to it's limited capacity, but it will have these torque numbers typically at higher revs making up in horsepower a lot. You might have power all the way up to 8000 rpm in theory... please don't over rev it and stay below 7600, but still. Max power I expect just there.
Now if you read all this and think... dyno? Wideband lambda? I can't get my hands on that for whatever reason. Especially dyno can be scarse... or sea it as a reason to start your own enterprise locally
. Well in that case I can't give more info then my stage 0... and my experience about ignition (keep stock), cam timing (retard half a tooth to one full) and expect fuel maps not to be as lean in high revs as you r indication. Tuning with exhaust smell, sparkplug checking and these sort of things... out of my experence, I try to avoid relying on such basic methods.
Darn I should write books or something about this stuff...
