Removing Ethanol from Gasoline

[Coyote X]

Originally Posted by Nerys

dcb for the other guys solution. is it pegging fuel trim because of the vac leak or is the vac leak not letting it peg enough fuel trim? IE we don't WANT to use more fuel :-) (just clarifying)

The ethanol needs a different air fuel ratio to properly work. Even a the amount you get from running E10 changes the desired ratio from 14.7:1 to something like 14.1:1 I am just guessing on the numbers but that seems close to me. The OBD1 computers only had a limited range they could deviate from the fuel table so they seem to be adjusting to the limit and that is where E10 is more or less happy with. It is close enough that there will be no extra room if there is the slightest problem with the car. So a very small vacuum leak would normally make it run slightly richer to compensate and the car would run just fine at 14.7:1 with normal gas. But with the table already at it's limit there is no more room left to compensate and you get terrible mileage from running lean. On light throttle highway cruising you can get away with lean running. On acceleration or medium throttle that will get you knock and that is really bad for mileage and power.

Newer computers have more ability to self tune so should not be affected as much. But a OBD1 system is really not much different than a carb system when it comes down to it other than the 10% or whatever feedback adjustment it can make on the fly from the O2 sensor depending on the car's computer setup.

[autoteach]

Originally Posted by mwebb

..."the fuel doesnt usually mess the pump up, it usually gets to the contacts on the fuel sending unit. this does not necessitate a change of the pump."...

no , it is BOTH the fuel pump and the sending units that are destroyed .

BMW and Mercedes Benz will not warranty a fuel pump IF the fuel in the tank has contaminants like water or fuel with more E than E10 .
those fuel pumps have a very hefty price tag
the warranty claim must be submitted with fuel test results or it is denied .

many OEMs demand or at least strongly recommend top tier fuel because of the contamination problem

I have 30:1 failure of sending units, whose contacts are directly in the fuel, to the fuel pumps. This is far more common, and maybe that is the way my post should have read.

[tjts1]

Originally Posted by Nerys

I do know on E10 my HC is just shy of max for failure of an emissions test. If I can find someone who will do it for me I would like to run that test with E0 and E10 and compare. I could have E0 in the car for this years inspection in a month but comparing that with YEAR old E10 data is not a very valid comparison.

Your O2 sensor is probably toast which would would also explain the poor running on E10. I've said it before and I'll say it again. Get your car fixed.

My smog check numbers before (top) and after (bottom) replacing my O2 sensor. Both tests were done on E10 in a 1992 mercedes with mechanical fuel injection (Kjet designed in the early 70s) and a closed loop O2 sensor. Hardly cutting edge stuff.

Originally Posted by Coyote X

The ethanol needs a different air fuel ratio to properly work. Even a the amount you get from running E10 changes the desired ratio from 14.7:1 to something like 14.1:1 I am just guessing on the numbers but that seems close to me.

The O2 sensor doesn't measure air fuel ratios. A simple narrow band which every car on the road has, measures the O2 content in the exhaust and compares it to open air (20.9%). EFI always strives to maintain stoich where all the O2 in the exhaust is consumed. From the Bosch EFI manual:
87 octane (AKI)= 14.8 afr
93 octane (AKI)= 14.7 afr
E10 = 14.2 afr
E85 = 9.87 afr
Propane (LPG) = 15.67 afr
The same O2 sensor will give a perfectly accurate switching point at stoich with any fuel, whatever the afr.

Originally Posted by Coyote X

The OBD1 computers only had a limited range they could deviate from the fuel table so they seem to be adjusting to the limit and that is where E10 is more or less happy with. It is close enough that there will be no extra room if there is the slightest problem with the car.

This is incorrect. Even the most basic EFI system from the 80s has at least 10% and usually much more room to adjust air density or fuel content based. Running on E10 (~3% extra fuel at lambda 1) is well within the limits of any EFI system.

Originally Posted by Coyote X

But a OBD1 system is really not much different than a carb system when it comes down to it other than the 10% or whatever feedback adjustment it can make on the fly from the O2 sensor depending on the car's computer setup.

This is gross generalization and blatantly wrong. OBD1 or 2 only deals with the diagnostic equipment. There were plenty of very advanced EFI systems during the OBD1 era. The original Civic SI stood for sequential injection in the late 80s. Bosch Motronic came on the scene around 1990 and went through several generations by 1996. Even LH 2.4 could handle sequential fuel fuel injection, coil on plug ignition, and VVT all from the same box in 1992. I would like to see a carburetor do that.

I don't know where you're getting your information.

[Nerys]

tjts - I would have to say your diagnoses is incorrect. the O2 sensor is fine. In fact it was one of the first things I replaced. I installed a new O2 sensor and when nothing changed in 2000 miles I removed it and reinstalled the old sensor Still no change.

Are you seriously trying to tell me I have 10 bad O2 sensors and no codes ? the probability of that approaches zero.

in fact while getting 17-18mpg in the jeep it suddenly started getting 9-10mpg about 2 years ago. It was the O2 sensor. I replaced it mileage returned to 17-18mpg but never to its correct 22mpg unless I got E0

I had a similar problem in the VAN but in this case the "wire" leading to the O2 sensor came out of its stay and came into contact with the exhaust which promptly melted it.

Fuel economy off course tanks and it threw a code. I replaced the sensor and secured the pigtail better (the old twist lock thing that held it had become brittle)

Mileage returned to its E10 lower ratings ie 13mpg.

thats is 3 of my vehicles where I replaced the O2 sensor post ethanol. 2 needed it 1 did not. the 2 that needed it did NOT need it on ethanol. it failed POST ethanol for an unknown reason on the jeep and because of the melted plug on the van.

I DO now have a scope now. once I learn how to use it I will even be able to VERIFY that the O2 sensors are fine.

The minivan is OBDII and was checked both O2 sensors are functioning correctly.

also where do you keep getting this fictional running poorly idea from? who ever said my car's run poorly? not me that for sure.

The only issues I am having is poor fuel economy compared to what I get when I do not use Ethanol.

While I THINK the cars sound better on E0 that is purely subjective and I was very clear about that. there is no concrete difference that I can detect in operation EXCEPT for fuel economy.

NOT ONE of my cars is running poorly.

Anything else?

tjts is correct regarding obd I and II the only real difference directly related to the OBD was the encoding of the information. Previous to ODBII the manufacturers were starting to intentionally obfuscate access to the data to prevent easy mechanic/customer DIY repair.

Laws were past to force them to open and standardize to OBDII this did not change the actual cars or their systems.\

[Coyote X]

I am sure there were some pretty advanced systems around but I am really generalizing with my descriptions. When I talk about OBD1 systems I am talking about the older computers that normally didn't have the cpu power or memory to handle long term fuel trims. Some did but most didn't.

I am also not going to get into the exact inner workings of an O2 sensor. suffice to say it measures lambda and leave it at that. I know how it works and anyone who cares to look can figure it out. When I said it tries to change the ratio from 14.7 to something else that is accurate. It doesn't 'know' what 14.7 is nor does it care. I have no interest in typing so much crap here that I accurately describe every test case and am just generalizing with the gm obd1 $0D system I have sitting in my driveway and also the megasquirt in my Metro. if you want to wander off on some rare exotic injection and test it then by all means test it out and give us some actual numbers. The GM and similar type TBI systems are really not that complicated. Trying to prove I am wrong by picking one random technical point and saying it isn't accurate isn't a good way to disprove the entire idea.

I have more diagnostic equipment than most people and have access to some pretty exotic stuff so I usually have numbers to back up what I say. All the testing I have done on my van was accurately measured and datalogged so I feel confident in my numbers. The only thing I don't have is an exhaust gas analyzer and steady state dyno but I will probably build one eventually to fine tune my metro a bit more than is possible with just a wideband and egt. The van is done being tuned and this weekend I will begin the diesel conversion project on it probably and actually swap it out in the next few months.

If anyone wants to learn how to fiddle with the gm obd1 systems grab tunerpro rt 5 beta. It is free to use and a few bin files and look at what information the computer uses. My exact setup is the 0D 1.75 advanced from moates.net and the corresponding ads file. My van's bin is BJLA but the BJLF is the 8 cyl version of the same bin and is also available on moates.net

[mwebb]

these areas as identified in this dated EPA link
are where the greatest / worst tampering with the fuel are

Where You Live | Reformulated Gas | US EPA

i happen to live inside of an "Opt In" RFG NON Attainment area ....
my fuel economy is 20% off from others with similar cars in other areas
==================

for those that have an interest
to test your 02 sensor s

you will need a scope ,
the 02 sensor must switch from 150mv to 850mv and from 850mv to 150mv in 100ms or less at 2k rpm hot .

if it does , it is "good"
assuming the 02 sensor heater is also functioning as designed

on OBD2 systems -
those of you who have the ability to view
Mode 5 or Mode 6 can look up the TID for the CID for the 02 sensor in question .... however those test results may pass an 02 sensor that would FAIL the above test .
when in doubt
the test above supersedes all other tests

the 02 sensor below passes the Mode 5 test
but
fails - this Bosch 02 sensor is brand new and DEFECTIVE
(new really means Never Ever Worked)
it does switch from 150 to 850mv in less than 100ms
but from 850 mv to 150mv takes much longer .....
KAKA - you can not see this without a scope


scan data showing 2 front 02 sensors , one is bad the one that takes too long to change state
------------------------------------------------
AS you can see -
an emissions test pass or fail is not a good way to identify a bad or failing 02 sensor
--------------------------------------------------
on a speed density system
with a MAP sensor as primary load sensor
a vacuum leak will result in rich condition with retarded ignition timing

NOT lean condition with advanced ignition timing

because
MAP = Baro - Intake manifold vacuum

so when intake manifold vacuum is reduced due to vacuum leak
MAP is increased so Load is increased
-----------------------------------------------
sometimes
this site is like watching children bickering - many of you know a great deal and many of you know just enough to get in trouble ....
remember why we are all here
to improve our fuel economy
try and work together to exploit each others strengths ... and overlook the minor errors and weaknesses

we are all on the same side

[tjts1]

Originally Posted by Coyote X

I am sure there were some pretty advanced systems around but I am really generalizing with my descriptions. When I talk about OBD1 systems I am talking about the older computers that normally didn't have the cpu power or memory to handle long term fuel trims. Some did but most didn't.

Every single Bosch system I've deal with does. Even my mechanical KeJet does in its electronic piggy back. It simply adds or subtracts fuel pressure based on the lamda sensor switch point the last few times times the ignition was turned on. Simple but effective. Maybe this wasn't a popular feature on GM cars. They were very fond of TBI for a long long time so I wouldn't be surprised.

I'm using a bosch LSU and 14point7.com wide band controller/data logger. Megasquirt should be ready to install in a couple of weeks.

Originally Posted by mwebb

I drive between Seattle and southern California (usually with a week long stop in the Bay Area) and back about 4 times a year. FE is consistent no matter where I fill up. Its all E10 now no matter where you fill up on the west coast. I used to see a 1-2 mpg difference between Oregon and Northern California gas before E10 invaded the PNW. Not anymore.

[dcb]

Originally Posted by mwebb

[COLOR="Blue"] ...on a speed density system
with a MAP sensor as primary load sensor
a vacuum leak will result in rich condition with retarded ignition timing

NOT lean condition with advanced ignition timing

Correction, it will result in a stoichiometric condition as long as the system can compensate for the additional air via the o2 sensor (and updates the trim values, etc). Coyote and others have noted that their cars are basically "pegged" on the "try to add more fuel" side on e10 as the ecu tries to bring the mixture back to stoic (14.2 ish for e10). Once at the limit of the system to compensate, any additional bypass air means leanness.

Though their spark plugs should be pretty clean if they are constantly lean.

It may also be a function of "driving on the o2 sensor", i.e. the internal maps are targeting 14.7 for the fuel schedule and the o2 compensation algorithm just kinda sucks when you get to 14.2 (large mixture "ripple")


EDIT: Nery's what fuel injections system(s) do your cars employ, all TBI by any chance?

[Nerys]

No - the Geo is TBI and one of the 94 Lumina's is TBI the rest are MPFI.

The only one I am not 100% sure of is the 92 Clubwagon I just assumed it was MPFI. I will have to verify that.

[dcb]

Yah, I'm running out of ideas here, I'd say try more fuel pressure, hotter thermostat, more spark. See if you can interpret the o2 sensor.