This.

That page was a good read, and I discovered that the GM fuel sensors actually measure both ethanol percentage as well as fuel temperature, but I'm still curious how it works. The article seemed to hint that perhaps it's measuring oxygen content (that is, the oxygenated fuel) to determine ethanol percentage, but I still don't know how that is actually measured.

When I used to switch between the two I'd always try to run the tank down to the last gallon or so of fuel (in a Talon...16gal tank), so the mix was relatively innocuous with respect to the ratio. I never had an issue with either tune on my dual map setup when I approached it this way.

This is the method. I've seen a pic somewhere...I'll post it if I can find it again.

Ok, serious time. Some research into the subject has produced a couple of interesting datapoints.

First, my first guess (which was actually intended to be sarcastic) was right. At least some of the sensors are capacitive in nature, using the fuel as a dielectric. Specifically those supplied by Siemens and Delphi are of this design, which are apparently the ones used by GM. There are some newer designs which are refractive in nature, but I find no references to their actually being used in an OEM application. Source: http://www.eng-tips.com/viewthread.c...=161127&page=7


Second, it would appear that these sensors are unreliable as hell. Lots of complaints on E85 forums about them failing. Apparently, the situation is so bad that people are now building "sensor replicators" that simulate the signal produced by a tankful of conventional E10. This eliminates the ability to run E85, but folks don't seem to care. Here's a link to one such product: http://www.tandselectronics.com/


Third, GM doesn't use fuel composition sensors of any kind anymore. Instead, they're doing the analysis purely in software (the call this V-FFS) using the primary O2 sensor and a lookup table which compares expected injector durations to various fuel compositions. In other words, the ECU knows that, at a given load condition, it should take xx milliseconds of injector duration to produce 1λ on gasoline, and yy milliseconds to produce 1λ on E85, so by looking at what the actual injector duration is at a point in time when the O2 sensor is giving a stable reading of exactly stoichiometric, it can make an educated guess as to what the fuel composition is.

Sources: http://www.autos.ca/auto-tech/auto-t...ex-fuel-system
http://www.gm-trucks.com/forums/inde...howtopic=52132

Clever.

I wonder if the MS3 inner-circle is cognizant of these facts?

I wonder how immediately it applies that change -- real-time data from the O2 sensor, I know, but how rapidly does it modify the fueling?

Hypothetically, if you shut it off running 100% gas (ethanol-free), and then emptied the tank and started it up on E85, how many combustion cycles would occur before the fueling was adjusted to the appropriate level?

As soon as the residual fuel is out of the lines, rail and injectors and burned. Pretty quick I'd say. Unless you meant that by draining the tank.

Right, but how many combustion cycles will have occurred by the time the O2 sensor has measured the change and adjusted the fueling?

Probably several hundred or several thousand, if I interpret what little information is available correctly. Of course, I expect that feedback from the O2 sensor and knock sensor would at least get the engine "roughed in" prior to actual recalibration, and perhaps a large detected variance in gEGO or gKNOCK is enough to initiate a recalibration by itself?


This is an interesting question. If I interpret the information I have found literally, the system would not recalibrate at all unless you filled the tank to a higher level than it had previously been filled with prior to being drained. I'm sure there must be some method for dealing with the contingency, but I've not seen it specifically enumerated.

That was awesome. :laugh: I would have drove along side of them longer to get more footage. What are they gonna do, get me?

Related video -

"Mommy, why don't they make a Barbie like me?"

"Sorry honey, real girls walk on the floor with their feet. They don't scoot around using their fat layers like a blob".

I swear, if one more person posts advice on my thread about Rotrex with stories that pertain only to roots blowers....

what is the average price to get a rear axle fixed in a older truck... say a 1986 ford bronco?

I don't know. Don't you work for NERV? Ask Balthasar or something.

Seriously though got pics of your Auska ex gf? Plez

I got banned from CR for a week. WHAT WILL I DO?????
Apparently it's unacceptable to tell somebody to army crawl through a razorblade factory then swim in a pool of AIDS. Fuckin crybabies

god dammit faeflora... god dammit.

What was that retarded magazine that decided to turn a Miata into a racecar so they bought a 600 dollar eBay turbokit and a 1.1k 5Zigen exhaust?

All of them.