I would think either the sensor is good, or its not. Are you sure the computer is throwing an O2 sensor code? I'd probably look into a sticking EGR valve, or some other mechanical part. I know I had a similar problem on my Buick years ago and the vapor cannister purge valve was sticking. I don't know if Honda used vapor canisters or purge valves on that year/model car though.
Are you sure it is actually a mpg issue. If you fill up at different gas stations with unlevel parking pads and pump cuttoff switches with different sensitivities you might just not be getting the tank completly filled every time.
Probably the gas pump. I filled today from the uphil side of the pump and I was about 1 gal short of what my trip computer said it should take and it wouldn't take another drop. I'm usually +/- .5% if I use the same pump as compared to the readout
My VX stated doing the same. Then it eventually just got bad MPG. The cel would turn on every once and a while like yours. It showed a code 48 for the O2 or laf sensor. I ordered a new one from honda parts on the web.
Hope it works, $363 is a lot of money for a sensor. Local Honda dealer wanted 430.
My O2 sensor has about 130,000 miles on it and sometimes the check engine light comes on and I turn the car off and on again to get the light off.
I guess one question would be can an O2 sensor ?come and go? like that?
I mean, run great (deliver +50 mpg) and then ?foul out?.
Thoughts on other causes?
BTW just had my worst tank ever - 32mpg 11-19-08
imho, your o2 sensor is toast and your ecu is operating in safeguard mode, no lean burn allowed. 130k is high side for the finicky VX sensor. The o2 sensor on my VX was first replaced by the original owner at 117k where it began throwing the CEL 48. The second one took the VX to 217k, right around when I bought the car. I could not get much over 41mpg, the 'lazy' o2 sensor issue.
I ordered a genuine 5-wire Bosch VX specific o2 sensor and it discovered it threw the CEL 48 and I got my worst mpg ever with that CEL issue. btw, the cel went away every time I turned the engine off, only to return within a couple of minutes. Also, on the only full tank I ran with that sensor before returning it, my gas mileage fell off a cliff to ~35mpg, see gas log.
I just put in a new l1h1 this week and immediately got to ~44.8. I still think there is upside MPG potential to sort out.
Below is a link to a place that is selling made in Japan apparent l1h1 oem replacement sensors for $99 shipped. (Having just returned the above Bosch, I was not real confident when I placed the order as it was not from Honda).
I doubt these are going to be around very long at these prices: Got it here, scroll to the very bottom of page. It is truly plug and play, internal resistor, visually an identical to the oem. http://stores.channeladvisor.com/Aut...t=5&scat=20363
I have a strange fluctuation in my Civic VX's tank to tank mpg. (40 to 56 MPG)
Usually, when I get large variations tank to tank, I have not been using my "favorite" pump. So I chalk it up to fill errors. It all averages out in the long run. I always fill up with the pump on the slowest setting, and stop at the first click-off. I also try to use the same pump every time.
The theory about different gas pump shut off points accounting for a 30% variance in my gaslog is a little far fetched. Besides don't we use the amount used to compute our MPG?
I wouldn't say this is far fetched at all. Your gas log shows the avg mpg being relatively stable. Nearly every poor mpg tank is followed by a better one. I know when I fill at the same pump my mileage is pretty consistent and nearly every time I use a different pump I get a higher mpg on that tank and then a lower one when I go back to the regular one I use.