Batman we have lean burn! 49.6 mpg!

[cfg83]

popimp -

Originally Posted by popimp

cfq83:

Did it work for you?

From the limited data I was seeing, I don't think it was working for me. I predicted seeing the A/F ratio getting more lean with the mod, but it was still in the 14.7 range. I also wanted to capture the other data (MPG/IGN/TPS/MAP) to see if the ECU/PCM was modifying other settings in reaction to the IAT mod. Here is why my experminent failed :

Attachment 812

The A/F meter is legible, but the scangauge is not. If I had prepped the setup to get the lighting under control, then I think it would have worked. My plan was to enter each change in value in a spreadsheet and average the totals.

Note : Filming the scangauge is a kind of "poor man's carchip" :
http://www.davisnet.com/drive/produc...p_products.asp

Note : When the engine is cold and in open loop, I do think the IAT mod works, because I was definitely observing a bias in the A/F.

CarloSW2

[popimp]

So can we put the resistor mod in the busted field for Dodge Grand Caravan and Saturn?

[cfg83]

popimp -

Originally Posted by popimp

So can we put the resistor mod in the busted field for Dodge Grand Caravan and Saturn?

Most probably, but I don't want to say for sure because my test failed .

CarloSW2

[unstable bob]

Originally Posted by popimp

So can we put the resistor mod in the busted field for Dodge Grand Caravan and Saturn?

I'm trying an IAT mod right now in my Cavalier, using a potentiometer to set the incoming air @ 222 degrees. Doesn't look too promising so far, and I may in fact be losing MPG. I should have some more concrete evidence later in the week.

[itjstagame]

Originally Posted by cfg83

Note : When the engine is cold and in open loop, I do think the IAT mod works, because I was definitely observing a bias in the A/F.

CarloSW2

That's what I was expecting, when in open loop the ecu has to use the inputs it has to determine fuel and IAT is am important part, so it will see it's hot out and use less fuel.

Once closed though it will adjust almost solely based on the O2 sensor, some ECUs have only a 'range' of adjustment to a fuel calculation (that's the same as open loop), so the calculation will say to run at 16:1 and then in closed loop it'll know it's lean and adjust as much as it can but can only add a certain amount of pulsewidth and so will end up lean still (like 15:1).

I could see other ECUs adjusting totally on the O2 or allowing more pulsewidth adjustment and holding you at 14.7.

That's why I always thought the resistor inline with the O2 sensor was a better mod, since it change the target AFR itself (of course you'd better get a digital AFR gauge or something so you can really tune how much resistance yields whichever results). This also shouldn't affect the AFR at WOT since it will just ignore the O2 sensor and calculate fuel just like it always has.

Of course best would be to have a stand alone (programmable) ECU with a wideband O2 sensor so you can just dial in the AFR you'd like.

For your digital gauge did you install a wideband sensor or is it just reading your stock sensor?




Also, asked way above about how to know if you're too lean, install an EGT gauge (Exhaust Gas Temp) that will show you if you're getting to dangerous levels (like the melting point of iron).

[cfg83]

itjstagame -

Originally Posted by itjstagame

That's what I was expecting, when in open loop the ecu has to use the inputs it has to determine fuel and IAT is am important part, so it will see it's hot out and use less fuel.

Once closed though it will adjust almost solely based on the O2 sensor, some ECUs have only a 'range' of adjustment to a fuel calculation (that's the same as open loop), so the calculation will say to run at 16:1 and then in closed loop it'll know it's lean and adjust as much as it can but can only add a certain amount of pulsewidth and so will end up lean still (like 15:1).

I could see other ECUs adjusting totally on the O2 or allowing more pulsewidth adjustment and holding you at 14.7.

That's why I always thought the resistor inline with the O2 sensor was a better mod, since it change the target AFR itself (of course you'd better get a digital AFR gauge or something so you can really tune how much resistance yields whichever results). This also shouldn't affect the AFR at WOT since it will just ignore the O2 sensor and calculate fuel just like it always has.

Of course best would be to have a stand alone (programmable) ECU with a wideband O2 sensor so you can just dial in the AFR you'd like.

I wish I knew a year ago what I know now. I would have gone with something along the line of DRW's Apex SAFC fuel controller that he has in his Plymouth Laser.

For your digital gauge did you install a wideband sensor or is it just reading your stock sensor?

It's reading the stock narrow-band sensor. After it arrived, I was worried that it was wide-band only, so I e-mailed the manufacturer, and they assured me that it was narrow-band compatible.

I have also read in the net that it can be modified to read wide-band, but I don't have the URL in front of me.

Also, asked way above about how to know if you're too lean, install an EGT gauge (Exhaust Gas Temp) that will show you if you're getting to dangerous levels (like the melting point of iron).

I also have a high temperature probe kit for just this purpose :

High Range Adjustable Temperature Switch with LCD
http://www.jaycarelectronics.com/pro....asp?ID=KC5376

But, I haven't built it yet.

CarloSW2

[lovemysan]

Well I'll update my findings. I removed my IAT resistor. I had to recalibrate the SG. I reworked my HAI for more consistant hot air. I also work to keep the temps above 150 and below 190. I'm getting better mileage. I'm 3 days away from my first mixed driving 50mpg+ tank. For me the resistor mod is busted! I'm getting 57mpg around town. I can get 59mpg @50-55mph. Next up is a block heater and sealing the engine compartment.

[itjstagame]

cfq83: Those are pretty sweet kits. I bought a kit to learn programming a pic controller myself but just haven't had the time to play with it. I think what I'd really prefer is the exact voltage that being outputted by the sensor, but I guess the LEDs are enough to give an idea of where you're running.

[cfg83]

itjstagame -

Originally Posted by itjstagame

cfq83: Those are pretty sweet kits. I bought a kit to learn programming a pic controller myself but just haven't had the time to play with it. I think what I'd really prefer is the exact voltage that being outputted by the sensor, but I guess the LEDs are enough to give an idea of where you're running.

Yeah, I really like the jaycar stuff. Cool kits that are car focused and at a decent price. They seem to link their car-mod articles with the kits, so there is more explanation than just the assembly instructions.

If you want the exact voltage, then all you need is a multimeter, but it would need a major refresh rate if it is digital.

CarloSW2

[itjstagame]

Hmm, yeah, I just found these:
http://www.allelectronics.com/cgi-bi...810&type=store
http://www.harborfreight.com/cpi/cta...emnumber=90899

At those prices it's cheaper to just hook or hack one of those up as a permanent install. I'm actually picturing pulling a couple different + lines (like from O2, Water temp, IAT, etc) and a negative line all to under the dash with some kind of convenient connectors and just plug the probes on my dash mounted multimeter on whatever I want to test.

You're right though about the refresh rate, the one say about 2.5sec! I don't really want to see it bouncing all over anyway, I think what I need is two displays, one showing the most recent max and the other showing the most recent min.

What does your do, average over the past however many microsecs?

If only I weren't so cheap :-).