Hi All,

Today I did several things to my 2006 S2000 with 112,000 miles. I performed the following, in this order:

• Valve adjustment - exhaust valves were all tight, intake side was in spec. I adjusted the exhaust valves to .012, as per Billman
• Battery change - was getting 12.56v on the battery, so changed it today for new.
• Compression test - done cold, fuel pump fuse pulled (yes, I know it was supposed to be warm), got 210, 212, 214, 236
• Changed plugs with new NGK (7772) PFR7G-11S platinum plugs. Gaps were all spec.
• Changed serpentine belt with new Continental (Goodyear) belt. I can't believe I paid someone to do this in the past.

After all this, I started the car and it idled higher than normal - about 2000 RPMs, but started better and ran smoother than before the work. I did not do a formal idle relearn because I wasn't aware I needed to (but in hindsight, realizing I changed the battery, I think I should have). At the first few stop lights, the idle dropped low - 600 RPMs and stuttered, but didn't stall. After another 10 mins, when at a stop light, idle was fine. I assume it did some form of relearn. However, it appeared my mileage tanked badly. I know the fuel gauge is no way to measure, but I was down to about 1/3 tank when I left, and the trip was about 30 miles. On the way back, I could see the gauge dropping as I drove - I went from 7 bars down to 3 in about 10 miles. Now, I know I can't measure mileage by "bars on the tank meter", but it was pretty severe. There's no odd smell and it runs smooth - even better than before. I hit VTEC a few times and it was fine - pulled strong. So, I stopped for gas and fueled up (premium only, naturally), and it took 10.5 gallons, and I ran 306 miles overall on that tank. So that's the confusing part - overall, I got 29+ MPG. That's great - but I swear is seemed to be guzzling after the work.

So my questions are these:

• Is there anything I did that could cause my mileage to tank, and could it be the ECU is relearning and will become normal again over time?
• I've only gone through three tanks since getting my S2K - is the tank gauge just really heavily weighted at the top and starts dropping fast at 1/4 tank?
• Do I need to do a proper idle relearn to make sure the ECU learns correctly?

Thanks - doing the work was really rewarding today. Got a lot done and saved a ton in labor fees, but it's the sense of accomplishment I like most. I have to say, this motor is just an amazing piece of engineering. It's like it has two personalities - one below 5K and one after. Yes, I've owned several VTEC Acuras before, but man, this engine is something special.

Thanks,

- Mike

 

Hi

Answers to your questions:

1. It was the fact that the ECU had to relearn so it began over compensating by asking for more fuel than usual. If you did the relearn procedure then it should have been back to normal faster than taking it for a drive after the work was performed.
2. I do not think it is weighted at the top.
3. It would have relearned by now after all that driving you did but I say do it and compare the results to what you experienced post the DIY work you did without the idle relearn procedure.

Yes, this engine is a work of art and no other Honda or any other brand for that matter makes an engine like it

 

Probably just the fuel sending unit / float sticking a bit, seen it happen to some S2000's including mine.

 

Did a proper idle relearn this morning, and so far so good. It definitely runs smoother since doing the plugs and valve adjustment. Filled the tank and things seem pretty normal fuel-wise. I'll check the calculated mileage once I fill up, but thanks for the suggestions.

 

My '06 can get 200-220 miles on the first half tank of gas and 100 to125 on the second half when we are sightseeing. We have put 130k miles on our S and the gas gauge always drops like a rock the last 1/4 tank no matter how we drive.

ROD

 

Can confirm, mine does drop faster at the end as well

 

It's called geometry. Drag out those sophomore plane geometry books.

The fuel level sender in our (and most) cars is just a variable resistor connected to an arm that travels a 90° arc from full to empty. Toward the bottom fuel level the angle changes more rapidly than at the top. This is why the tank reads full for so long and near empty so fast at the end. I'm sure someone can explain this geometry more easily than I can. Note also that the arm can't move down until the float moves down and submerging in a very full tank keeps it there. Same at the where the gauge may read empty as the float is resting on the bottom or the tank (or arm limit) and there's still fuel under it.



-- Chuck

 

Thanks, Chuck. They say a picture is worth a thousand words and that rings true here.

And that's exactly what I've seen since my last fillup. I think I had almost 60 miles on the tank before the meter moved. Makes total sense.

Car has been running great since the work. Thanks all!

 

It should have been easily possible for the gauge to have been programmed to be non linear to the sending unit, so as to compensate for this non linear change in potentiometer resistance as the float procedes through its arc as prescribed by the lever arm.

For each step on the gauge, as a percentage of full, what is the potentiometer resistance range when the float is that percentage articulated along its arced path? Then whatever resistance is read, lookup which range bracket that value corresponds with, show that many bars. The existing methods for smoothing display as fuel sloshes around and float bobs around works well enough, so keep that in place along with above.

Iy shouldn't have been that hard to make a linear fuel level gauge.

 

More PII. (Preoccupation with inconsequential increments). "Getting low" is all the information most of us need.

-- Chuck