What can we measure ?

[Jonathan R. Lusky]

Steve Baldwin writes:
> I mentioned before that we should determine what we wanted a controller
> to do before we specified what should do it.
> Perhaps prior to that, we might discuss what parameters can be be
> measured, and what the relative usefulness is.
> For example, the exhaust pipe O2 sensor. What is available is pretty
> well restricted in its range of uses. I'll be using leaded fuel, so that
> rules that out. What about in an alky engine. Is an O2 sensor any use ?
 
Since you are using leaded fuel, is it safe to assume this is a racecar?
If so, then part throttle and idle fuel economy and emissions are near 0
concern for you and you won't run closed loop, and if anything you'd
want to run leaner than the O2 sensor can read accurately
when you are under those conditions (although the M85 judges
at F-SAE '93 majorly disagreed with me on that).  An O2 sensor is
useable with alky, CNG, propane, etc.  One thing to remember about the
O2 sensor and closed loop is that it is on top of the open loop
system...  closed loop is useless if you don't have the open loop
calibration correct.

> As I understand it, the O2 sensor gives us an indication of the
> efficiency of combustion but with a fairly slow response time. What
> other measurable parameters are there that could be used over a wider
> range of fuel types ?
 
Not directly.  The O2 sensor tells you whether you are at stoichiometric
or not, and which side you are on if you aren't.  It doesn't tell you a
whole lot about how far your A/F ratio is to either side of stoich, tho.
When you ask what else could be used, I have to ask used for what?
The O2 sensor is used for feedback for closed loop control at idle and
part throttle to mantain stoichiometry for best overall catalyst
efficiency when running on gasoline and using 3-way cats.

> My interest in DIY_EFI is not to re-invent the standard factory device
> by effectively cloning it out of a mixture of other peoples bits. We
> have a slightly different perspective than the auto manufacturer. ie. We
> are not solely driven by the emmisions vs. $$$ equation. Yes, it has to
> be realistic in the $ department and we don't want to be belching black
> smoke, but we have some lattitude in the way the money is spent. I don't
> mind having a thermocouple in each exhaust port if that is going to give
> me more information about what's going on inside.
 
You don't need an FI system to have a thermocouple in each exhaust
port...  they provide nice info on cylinder balance, and can quickly
point out problems.  UEGO's are also realllly nice, but they're kinda
expensive.  

As far as reinventing the OEM ECM goes, I see it as something to shoot
for!  The OEM stuff is considerably more advanced than the current
aftermarket ECM's that I've played with.  I'd like to be able add custom
stuff in addition to the features already found in factory ECM's.
Now I'm not saying that we should reverse engineer an OEM ECM--from what
I've seen of them, there are enough proprietary pieces that it would be
simpler to build something from scatch.


-- 
Jonathan R. Lusky  --  lusky at knuth.mtsu.edu
 "Turbos are nice but I'd rather be blown!"
   89 Jeep Wrangler - 258 / pile of junk!
       80 Toyota Celica - 20R / 5spd