running rich one day, then spot-on the next!!
 

On sunday, after fitting my AFR meter, I noticed my car (94 WRX) was running rich (>0.88v) which wasn't surprising considering it has a walboro fuel pump.

It sat parked for 24 hours and then I drove it again and now it is showing that it is 'spot -on' (0.84v-0.85v).

How could it just change like this. At first I thought it was 'cos my fuel tank was low but fille dit up and no difference.

It felt faster on sunday which is weird 'cos it should have better power at the leaner mix, but that could be in my head.

air temp, humidity ?????

What live / earth connection did you use? What gauge is it?

Bear in mind that AFR's are a bit of a rough guide and you are looking at differences in 0.02 of a volt. Not much!

Ambient press / temp may als cause slight differences.

MB

I've got it grounded well, straight to battery. It's a dawes AFR meter and I checked the voltages on sunday.

The only other thing I could think of, apart from humidity/pressure (thx scoob+bike=fun) is that I had used octane boost in previous tank, and perhaps the concentration drops towards bottom of tank

It can do.

Remember the Dawes only has a few lights so its pretty innacurate IMHO, and only useful as a guide.

Wideband lambda is the only accurate way of looking at it.

MB

Did you dis-connect the battery when fitting the AFR. If so it's the ECU in start-up mode i.e before it's learn't all it's fuelling parameters.

If you do an ECU reset it will probably do it again.

My old MY94 used to take a couple of WOT runs to settle the AFR after a battery di-sonnect or an ECU reset.

Back to how it was on sunday this morning. Running very rich, but felt quick so WTH. Might get it leaned out a bit on remap next month.

Very weird about the fluctuation yesterday.

DBM, the dawes AFR meter just shows the top end of the spectrum, but should be accurate within that range. I think all these AFR meters use the same LM3914 chip so accuracy should be similar. Infact as the dawes focuses on a smaller range it should be more accurate. I checked the voltage ranges and they are right.