Fooling the Computer for Better MPG

[srimes]

The scanner I bought only shows the last fuel trim values that existed when a DTC code, like a missfire is thrown. So it takes a better, more expensive scaner than what I bought to get continuous fuel trim data, but now that we know what it is and what it does, and how to use it, we know better what specs to look for in a scanner tool. I just wish I had known it 3 months ago.:twak::smsoap:

The scangauge manual talks about xgauge, which can look at non-standard stuff and includes sending codes to the computer to ask for specific information. I didn't read that part too closely 'cause it was over my head and I didn't have any ideas about what to ask for, but I think that the scangauge may be able to display fuel trim IF I knew how to ask for it. Which I don't.

Now, I wonder if this trim can be messed with to adjust the mixture? I don't know how quickly it updates, but if a command could be sent to lean out the mixture by 2%, say every second or three, it would still accept input from the 02 sensor but would bias the result. Not a simple, practicle idea though.

 

[Ecomike]

The scangauge manual talks about xgauge, which can look at non-standard stuff and includes sending codes to the computer to ask for specific information. I didn't read that part too closely 'cause it was over my head and I didn't have any ideas about what to ask for, but I think that the scangauge may be able to display fuel trim IF I knew how to ask for it. Which I don't.

Now, I wonder if this trim can be messed with to adjust the mixture? I don't know how quickly it updates, but if a command could be sent to lean out the mixture by 2%, say every second or three, it would still accept input from the 02 sensor but would bias the result. Not a simple, practicle idea though.

Can't do it directly like that. The trim is the actual computer calculated fuel value change (% increase or decrease of the previous value) that it feeds to the injector as a DC voltage for a specific amount of time that results in opening the injector for that amount of time. The change in the amount of time the injector is left open is the fuel trim. The only way to change the fuel trim is to change the sensor input data to the computer so that the computer calculates a different fuel trim.

What I like about this is, if you have a 96 or newer Jeep, OBD-II, (mine are 85-89, so I am screwed), and the right scanner, it allows you isolate and locate a bad injector(s), with out ever pulling the injectors! I could have used this trick on my son's 96 Ford last month given the right scanner. His was throwing a missfire code, and the fuel trim data that came with the miss fire code was meaningless to me until now!

 

[gone postal]

Want to talk about poor mileage?!? My '97 Country 4x4 w/ 115k on it averages 13mpg in town. I did a full tune-up about 5k miles ago, new properly inflated tires, etc. I bought a new O2, but haven't put it in yet. I think I might have a slight exhaust leak in the manifold though...need to order an APC header.

 

[BBeach]

On a side note of fooling the computer, how can I lower my idle to say 650-700rpm? Sometimes it goes down there, but other times it'll hang around 1000 (always on cold start warm up, but sometimes when its already fine). If there is any way I could modify the IAC through splicing in a resistor or something, let me know. At 750rpm I was getting ~.75gph and at 1000rpm Im getting .9gph. So it'll be a sure help for city driving.

 

[Ecomike]

On a side note of fooling the computer, how can I lower my idle to say 650-700rpm? Sometimes it goes down there, but other times it'll hang around 1000 (always on cold start warm up, but sometimes when its already fine). If there is any way I could modify the IAC through splicing in a resistor or something, let me know. At 750rpm I was getting ~.75gph and at 1000rpm Im getting .9gph. So it'll be a sure help for city driving.

Better to find the cause of the variable, high idle speed and just fix it. It should idle steady at 700-750 rpm. Look for TPS problems, vacuum leaks and ground problems (even if it is not a Renix), and try cleaning the IAC and throttle body area where the IAC piston seats.

 

[BBeach]

Better to find the cause of the variable, high idle speed and just fix it. It should idle steady at 700-750 rpm. Look for TPS problems, vacuum leaks and ground problems (even if it is not a Renix), and try cleaning the IAC and throttle body area where the IAC piston seats.

Last time I checked with a multimeter TPS was fine, scangauges shows it at 0% when closed and 99% when WOT so I think its set. IAC was cleaned up recently. Only possibility of a vacuum leak is one of my 4 bolts for mounting the throttle body to the intake manifold is stripped. The remaining 3 are torqued to spec and there is a gasket. Any ideas?

 

[Ecomike]

Last time I checked with a multimeter TPS was fine, scangauges shows it at 0% when closed and 99% when WOT so I think its set. IAC was cleaned up recently. Only possibility of a vacuum leak is one of my 4 bolts for mounting the throttle body to the intake manifold is stripped. The remaining 3 are torqued to spec and there is a gasket. Any ideas?

I was thinking Renix, but yours is newer, yes? I would suspect a hard to find, lurking vacuum leak on the newer, post 90 jeeps, and not the TPS since it is fixed, no calibration. Could very well be that fourth bolt, or intake manifold gasket. Vacuum leaks are very hard to seal!!!!

I discovered that the slightest seal leak, or changes in the CCV tubing size or layout, (it should have a restrictor in the small line on top of the Valve cover) will cause high idles. It just acts like a vacuum leak that the ECU/PCM can not compensate enough for.

 

[BBeach]

I was thinking Renix, but yours is newer, yes? I would suspect a hard to find, lurking vacuum leak on the newer, post 90 jeeps, and not the TPS since it is fixed, no calibration. Could very well be that fourth bolt, or intake manifold gasket. Vacuum leaks are very hard to seal!!!!

I discovered that the slightest seal leak, or changes in the CCV tubing size or layout, (it should have a restrictor in the small line on top of the Valve cover) will cause high idles. It just acts like a vacuum leak that the ECU/PCM can not compensate enough for.

Yeah its a 98. I did change the CCV tubing a bit using heater hose (i know the gas fumes should eat away but it hasnt happened yet) so that's a possbility, but its mostly on startup. After its all warm it'll go to 750-800rpm.

 

[Ecomike]

OK, I rechecked everthing this time, this data is correct, some of the prior data I posted may be off.

Power and engine off. Sensor disconnected. All readings were taken with the meter ground wire at the battery negative post. Cold O2 Sensor read:

Infinite resistance to ground on all three wires!!!!

O2 harness connection wires read:

#1 (black) 1.3 ohms
#2 (black) 4,000 ohms
#3 (orange) 53.5 ohms!!!!!!

Then with all three wires connected to the O2 sensor they read:

#1 (black) 1.3 ohms
#2 (black) 4,000 ohms
#3 (orange) 8.3 ohms!!!!!!!!!

Power on, engine off, all three wires connected to the O2 sensor read:

#1 (black) 150 MV ( I may need to clean and tighten my gounds again)
#2 (black) 5.00 V
#3 (orange) 11.87 V

Engine on, and hot:

Running 2500 rpm, then rapid decel to idle stop, I got under 1 volt for about 8 seconds!!!!!

So if renix is shutting off the fuel and running it lean the Renix FSM is wrong and the Bosch web site is right about the low voltage being lean. If the renix FSM is wrong (or mine is acting up) and the fuel is staying on during decel to 1200 rpm, then the Renix FSM is right and low volts is rich.

I got 1.5 to 4.0 volts at idle, about 1 second between swings across the 2.45 volt sweet spot at idle.

Got a rapid 2.5 to 3.5 volts at 1600 to 2500 rpm steady engine speed, about 2 swings side to side per second. It finally got to a tight 3 volts at 2,400 rpm steady engine speed once the engine was good and hot (about 10 minutes of run time). Last time I ran this test I was getting 2.45 volts almost dead on, +/- 0.1 volts, this time it was reading closer to 3 volts.

The voltage readings were taken with a high quality high impedence FET analog meter that was zeroed before the ohm tests were started. The ohms readings were taken with a high quality, high impedence digital ohm meter.

For some reason mine seems to be running a little richer than it was in my previous tests. It may be time to check my CTS and MAT sensors to see if one of them is out of calibration. My CTS is the only old sensor left on this beast.

 

[chewysdad]

A very similar discussion is available at http://www.energeticforum.com/renewable-energy/1387-o2-sensor-modification-using-spark-plug-non-foulers.html

I haven't read all of this one yet and I don't know if anything definitive has been reached on the other one. I'm just finishing a 94 XJ that I bought in a basket (somebody else tore it apart and lost most of the bolts) with a blown engine. I put some Singh grooves in the head of the new rebuild There's a can of worms that will create some heated discussion. Here's a link if you're unfamiliar with that: http://www.allpar.com/fix/holler/head-porting2.html or put "singh grooves" into Google. It's not V.J. Singh the golfer, but a man in India. Interesting read. If correct, the Jeep could tolerate a very lean mixture. I'll know soon, I had to order a new A/C filter dryer from NAPA, Probably several weeks from startup because of that and other committments. At one time in my life I was an ASE diesel engine tech so like my retarded cousin used to say, "I may be dumb, but I ain't stupid." :lecture:

 

[Ecomike]

Renix HEGO update:

I have been debugging my 87 Renix engine overheating problem. Found the following today:

OK, we have some REAL interesting data here for many reasons!

The HEGO (Heated Exhaust Gas Oxygen Sensor) on mine is running between 0.2 and 0.9 volts, mostly at 0.8 volts at idle, 0.6 volts at 2000 RPM steady, and 0.2 V during decel back to idle for about 10 to 15 seconds! This is a Renix system and it should be running near 2.45 Volts!

So the ECU thinks it is running rich and is trying to compensate by running it lean out to the max setting for lean on a Renix!

I will test this by disconnecting the HEGO tomorrow in another test to see if she runs cooler. I will also check the 5 volt ECU feed to the HEGO to determine where the voltage loss is coming from.

This data has huge implications for this thread, as mine was still running in closed loop mode at these low voltages!

And if I am right, this has huge implications for those with overheating coolant systems that have defied diagnosis, like mine was doing.

Also, note that my last tank of gas got a huge 30 to 40% increase in local mileage that have never gotten from this jeep in 5 years and 40,000 miles. I got about a 25% increase over my best highway miles, and half of it was idling in the driveway and local driving.

Note that I am basing this on the Renix 4.0 L Multi-port injection FSM, page 41, which says it is lean at voltages above 2.5 and rich below 2.5 volts, but keep in mind that is for a properly working sensor and ECU!

 

[1bolt]

Wait let me see if I have this straight... you've finally succeeded in "fooling the computer for better mileage" albeit completely by accident and the results of random overheating due to the lean condition excite you?:looney:

 

[gradon]

:laugh: at 1bolt

 

[Ecomike]

OK, turns out the data below is wrong.

Had the red probe in the wrong spot on my analog meter. My HEGO is working, getting 4.99 volts from the ECU to the HEGO sensor, 12 volts to the HEGO heater, & ground is good. It is running just a little bit lean on average (average is about 3 volts, running from 1.4 to 4.4 volts) but is switching back and forth across the 2.45 volt stoich point so the computer is working, and in closed loop mode. It should be averaging 2.5 volts. The swing narrows at higher RPM, but still averages 3 volts (still slightly lean).

But the fact remains that my mileage has shot way up, and it is running hotter as a result.

Not 100% sure yet, but my intake manifold seems to be a lot hotter than it use to be, especially near the head.

MAP sensor is OK, it passed the tests. I will post up here again when I figure it all out.

 

[xjtrailrider]

You may have to move the MTS (ACS, whatever we are calling it tonight) to the air tube now to richen it up some, hopefully not too much. I would just stick another one in there and have the 2 on a toggle so you can instantly tell the difference and switch back and forth for winter/summer.

Just a thought or are you thinking another direction from me?

 

[Ecomike]

You may have to move the MTS (ACS, whatever we are calling it tonight) to the air tube now to richen it up some, hopefully not too much. I would just stick another one in there and have the 2 on a toggle so you can instantly tell the difference and switch back and forth for winter/summer.

Just a thought or are you thinking another direction from me?

HUh?

 

[xjtrailrider]

HUh?

You say its running leaner and you need to richen it up correct? You say the intake is hotter than normally. Moving the MAT (I had it wrong in my post) to cooler air would richen the mixture some but maybe too much. That would be a simple solution. Or are you trying to go even leaner?

I'm trying to jump on your train of thought here!

 

[Ecomike]

You say its running leaner and you need to richen it up correct? You say the intake is hotter than normally. Moving the MAT (I had it wrong in my post) to cooler air would richen the mixture some but maybe too much. That would be a simple solution. Or are you trying to go even leaner?

I'm trying to jump on your train of thought here!

OK, Got it. Only problem is the working O2 sensor will compensate in closed loop mode as the MAT, MAP and CTS data is only used for approximation during start up, decel and acceleration until the O2 sensor stabilizes again.
I am hoping my problem is a bad thermostat, or clutch, and not a sticking intake valve or dying head gasket.

 

[1bolt]

Mike why do you not suspect the water pump? I've had three 4.0's that had random overheating and all of them turned out to be water pumps that were going bad... One of them was a new pump even... it went bad in short order (like a couple weeks IIRC) I couldn't believe it but I pulled it anyway and sure enough it was all rough when the shaft was turned... Duralast reman'ed job with the lifetime replacement warranty. It's replacement is still going strong after two years with no overheating at all.

 

[stevieb]

sorry to hijack
while in the subject of overheating etc. tell me would a cracked outlet have reason to lower my mileage to 10.5 mpg or 4.5 km/l (as we use here). this running highway at around 65mph. i have had the vehicle at the agents and to no avail. my xj does seem to run slightly warm as well. it is a '96 4.0. the cat has been removed some time back and have done a barrel air filter conversion. i run on 30's so i can't see this having such an influence.
any help will be very much appreciated. where can i start looking. any sites will help.
thanx
steve