Here is the thread for those interested.alcoholisms
How does methanol affect wideband readings?
- Thread starter Steve
- Start date
Easy. You run a mixture that yields as low of an air fuel ratio and the motor make the desired power.
The more alky volume run, the lower the target is.
You run just enough alcohol to not allow knock activity to register.
Then you work on raising the air fuel target in hopes of increased performance.. if it doesnt pick up when you lean it out.. back it down.
What works for you doesnt work for someone else. This is basic tuning and not rocket science. Especially on cars running stock chips and MAF's. You want to start splitting better atoms.. start with an aftermarket engine management and start datalogging everythings.. individual cylinder EGT's, intake temp, exhuast back pressure, IC efficiency, have your heads flowed, etc etc etc.
Summing it up.. start very rich.. work your way leaning it out observing performance gains and knock.
You want an air fuel number.. 10.0:1 to start. Not 12.5 then work your way down.
The more alky volume run, the lower the target is.
You run just enough alcohol to not allow knock activity to register.
Then you work on raising the air fuel target in hopes of increased performance.. if it doesnt pick up when you lean it out.. back it down.
What works for you doesnt work for someone else. This is basic tuning and not rocket science. Especially on cars running stock chips and MAF's. You want to start splitting better atoms.. start with an aftermarket engine management and start datalogging everythings.. individual cylinder EGT's, intake temp, exhuast back pressure, IC efficiency, have your heads flowed, etc etc etc.
Summing it up.. start very rich.. work your way leaning it out observing performance gains and knock.
You want an air fuel number.. 10.0:1 to start. Not 12.5 then work your way down.
Razor said:
This might be irrelevant, but a friend sent me this link:
http://www.boostcreeps.com/phpBB/viewtopic.php?t=614&highlight=
B
spins4 said:I actually wrote those posts. I'm 'racegate' on that board.And its not irrelevent for target numbers on a wideband, but to calculate your ACTUAL AFR for fuel calcs, you need to correct for the mix of fuels to know your real air to fuel ratio.
Perhaps I'm misunderstanding then, but it seems to me that it's a difference between calculated AFR's, which are displayed on the wideband itself, vs. *actual* AFR's. For instance, I'm aiming for a 75/25 ratio on my vehicle at high loads. Would I still follow the typical mid 11's:1 AFR display or do I recalculate according to the chart and am for a display of, say, 10:1?
B
BDC said:
Yes, the difference is exactly what you stated-- displayed AFR on the wideband is NOT your true "AFR". Just target the AFR you normally would running race fuel (mid 11s if thats what works for your car) and go from there.
To understand, its best to think in terms of lambda. The wideband measures lambda, and displays an "AFR" number based on an arbitrary scale. If your car on a particular fuel combo makes best power 20% rich of stoich, thats .8 lambda, period. What "AFR" the wideband displays for that lambda value is dependent on the scale calibration. On gas it displays one value, methanol another, LPG yet another...etc.
Now the more important question becomes, what lambda number will make the best power? Thats something you have to experiment with..... for which Julio's advice is excellent, which is to start rich, and then lean it out until your incremental increases in power are diminshing or you register some knock-- leaving room for saftey. I would not tune on the 'edge' on a methanol/pump gas tune, because although the methanol content is very stable, the pump gas you get from station to station, day to day can vary and cause problems if you leave the tune on the edge.
Strictly speaking, if you run a 25/75 meth/petrol ratio, and your car makes best power at .8 lambda, and you then move to a 40/60 meth/petrol ratio you will find you will make best power at a slightly richer lambda. This is because of the burn properties of methanol compared to petrol.
If you want to know what your actual air to fuel ratio is... then use the chart to figure that out.
spins4 said:
Very interesting info, racegate. This begs another question: I wonder if my wideband is calibrated strictly for the use of gasoline, therefore making its displays based off of a calibration from its lambda reading. Is that correct?
And if so, I ought to stick to just my normal AFR #'s and not do a calculation according to 'The Chart' to mega uber rich:1?
B
BDC said:
Yes that is exactly correct.
Here is a a very simplified explanation of how a wideband works, which may aid in understanding:
There are 3 essential parts to a wideband oxygen sensor which are using for arriving at an lambda reading: the Nernst cell (which is also composed of the reference chamber), and the oxygen-pump cell. The are on opposite sides of a diffusion gap, which is essentially a cavity where exhaust gas enters to be sampled. So think of a cavity for air to enter surrounded by two cells on either side, with a bridge in between. Just like a bridge connecting two landmasses, over a body of water.
The nernst cell is a cell which is only conductive to oxygen ions and has two probes-- one probe is expose to ambient air, and the other to the reference chamber.
The engine is running exhaust is flowing and exhaust gas enters the diffusion gap into the cavity. The reference cell has a voltage Vref and reacts to the exhaust sample by producing a voltage V above or below the Vref voltage depending on if the mixture is richer or leaner than stoich. If the mixture is stoich, the voltage simply remains Vref. Now, how does it know what 'stoich' is? Remember the nernst cell has a probe exposed to ambient air, so it knows the partial pressure of the oxygen content of our air. Normally this is around ~20% O2 at sea-level. Stoichometric means, every oxygen molecule is used up in the reaction, and there is NO excess oxygen nor excess hydrocarbon left over. So by comparing the partial pressure of the O2 of its sample, with whats in the exhaust gas, it knows if there is excess O2 or excess hydrocarbon in the mixture, thus knowing if its 'lean' or 'rich' of stoich. Important to note: it doesn't matter WHAT fuel you are burning, as the sensor makes NO reference to the fuel. Simply it compares the partial pressure of the O2 in the atmo, to the partial pressure of the O2 in the exhaust gas sample, without regard to fuel type which generated the exhaust mix. This is how a sensor produces narrow-band readings.
Now for the wideband portion of it, here is why a controller is necessary. The pump cell can either consume free fuel or free oxygen. It 'pumps' oxygen into and out of the Nernst cell-- this is controller by the wideband controller. In order to pump oxygen into/out of the Nernst cell you need to generate a pump current I. The direction of this current is dependant on which way you are pumping. What the wideband controller does sense the V from the Nernst (which tells the controller of the mix is rich or lean of stoich), and then it turns on the pump to either pump oxygen into/out of the Nernst cell until the Nernst cell voltage is back to Vref. As I said, driving the oxy pump takes a current I, and by measuring this current I once Nernst has reached back to Vref (equilibrium) the wideband controller knows how much oxygen it had to pump into/out of the Nernst cell, and thus knows EXACTLY how far off and in what direction the exhaust gas sample is from stoich. If the mix is already stoich, I = 0 thus the wideband knows Lambda = 1.
Now, once the controller knows the pump current it can calculate lambda as a function of pump current, time, etc. Lambda is a measure of completion reaction. 1 lambda perfect reaction, less than 1 lambda means excess hydrocarbons, more than 1 lambda means excess oxygen. Again, not no where in this is it taking into considering WHAT fuel generated the excess oxygen or hydrocarbons in order to arrive at a lambda.
Ok, so now we have lambda. What the controller now does is take the lambda value and multiply it by the stoich ratio of the fuel you tell it you are running. Most widebands are calibrated for gaslone, which has a stoich ratio of 14.7 to 1. So if the controller calculates 1 lambda, it says: 1 lambda * 14.7 = 14.7 to 1. If you tell teh controller you are running methanol it says: 1 lambda * 6.4 = 6.4 to 1.
So you can see how LAMBDA is mesured independant of fuel used, but "AFR" as displayed by the meter is calcualated with respect to what fuel you are using based on what YOU tell it... because you tell your wideband, how you want that lamdba displayed. The multiplier is arbitrary for the fuel you are using. 10% rich of stoich is 10% rich no matter what. The WBo2 measures correct lambda without any user input.
BDC said:
Yep. Optimal lambda might change with the mix, but you dont want to use the 'chart' to know what AFR you need to target.
