DIY_EFI Digest V4 #696 O2 Sensors

Thu Dec 16 00:52:27 GMT 1999

The standard sensors are very poor analog devices. The
transition area is very abrubt, non-linear, and
probably not consistent from batch to batch. The
factory ECU depends upon the transitions from rich to
lean, 'crossings' over the threshold. The A/F ratio is
set by the injector pulse width, the system acting
like an integrator over time. Those of you who see a
'constant' .75 V are probably using an analog meter,
or a DVM with with long time constant or input
filtering. If you take a look at the O2 sensor with an
oscilloscope, you will see the crossings in clsed loop
mode, as well as injector pulse width 'jitter' as the
ECU continuously adjusts the A/F ratio. 
As has been discussed on this list, wide band accurate
O2 sensors are available, but expensive. If these were
manufactured in the same volume as the current auto O2
sesnors, they would probably be about the same price.
Think of the ECU and O2 sensor like a big switchmode
power supply, the A/F ratio is varied via pulsewidth
to achieve an average value that is integrated over
time.

I wonder what the effective loop damping and settling
times are for these systems? Are there systemic
oscillation conditions to be aware of, and to design
out, for DIY'ers like us? Anyone have first hand info?

--- DIY_EFI Digest
<DIY_EFI-Digest-Owner at efi332.eng.ohio-state.edu>
wrote:
> 
> DIY_EFI Digest      Wednesday, December 15 1999     
> Volume 04 : Number 696
> 
> 
> 
> In this issue:
> 
> 	CFM vs. lb/min
> 	re: Pulse Wdith.
> 	Re: DIY_EFI Digest V4 #695
> 	Re: 4 links
> 	Max injector size?
> 	Re: ECU pulse width calculations
> 	RE: Airflow Measurement
> 
> See the end of the digest for information on
> subscribing to the 
> DIY_EFI or DIY_EFI-Digest mailing lists.
> 
>
----------------------------------------------------------------------
> 
> Date: Tue, 14 Dec 1999 15:37:02 -0500 (EST)
> From: William T Wilson <fluffy at snurgle.org>
> Subject: CFM vs. lb/min
> 
> On Tue, 14 Dec 1999, DIY_EFI Digest wrote:
> 
> > Can someone tell me why cylinder heads, intakes,
> throttlebodys, etc.
> > etc. airflow numbers are quoted in CFM? Would it
> be better if they
> > were given in lbs/min? Why or why not? I
> understand why compressors
> > are rated this way just not why other engine
> componets are not.
> 
> Because CFM is the same regardless of ambient
> pressure, but lbs/min is
> not.
> 
> ------------------------------
> 
> Date: Tue, 14 Dec 1999 14:16:01 -0800
> From: "John Dammeyer" <johnd at autoartisans.com>
> Subject: re: Pulse Wdith.
> 
> >John Dammeyer <johnd at autoartisans.com> wrote:
> >
> >>If it's a standard O2 sensor why not replace it
> with a pot that provides
> >>0..1.0V and set the output to 0.5V which is
> supposed to be 14:7:1.  Then,
> >>because the O2 is telling the ECU that the mixture
> is perfect it won't keep
> >>screwing around with the mixture pulse width and
> you can see the real result
> >>of your tuning.
> >
> >I have little experiance in these things, but don't
> OEM controllers richen the
> >mixture until they see the O2 sensor make a
> "transition" from low to high, and
> >then they lean the mixture until the voltage goes
> back down.
> >
> >Holding the O2 sensor voltage constant would have
> an unpredictable effect.
> >Possibly the computer would flag it as faulty and
> use some sort of backup
> >system.
> 
> 
> I don't believe that's true.  The HEGO sensors have
> a voltage output of 0.5V at
> 14.7:1 and then the voltage changes rapidly on
> either end and so the sensor
> appears to have a on/off value.  Perhaps some engine
> controllers do just use a
> simple go/nogo approach but I doubt it.  I can
> adjust the mixture so the O2 puts
> out 0.55V or 0.6V etc.  I have no clue if 0.55V is
> 14.1:1 or 10.1:1 but I do
> know that with a properly set up engine with an
> Ellison Throttle carb. or else
> an SU carb that at maximum Torque the O2 sensor is
> at about 0.75V.  I've set my
> fuel injection to be roughly the same.  Exhaust Gas
> Temperature and a Post Run
> check of the spark plug colour all confirm the
> mixture is 'about' right.
> 
> Cheers,
> 
> John
> 
> ------------------------------
> 
> Date: Tue, 14 Dec 1999 17:28:28 -0500
> From: Scott_Hay at toyota.com
> Subject: Re: DIY_EFI Digest V4 #695
> 
> 4 Links
> 
> Van,  The length and distance apart are only a small
> portion of the equation.
> What are their angles relative to horizontal and are
> they parallel?   I assume
> that this car is for the road as opposed to drag
> racing (completely different
> setup).  The instantaneous center is the point where
> the torque of the wheels
> will act upon the body.  This is determined by the
> angles of the links.  If this
> projected point is too low (significantly below the
> CG) the front of the car
> will have tremendous lift on acceleration.  Too
> high, and not enough weight will
> transfer to the back and wheelspin (if you have that
> much HP).    As to the
> lengths that you are working with, that will
> probably be determined more by
> space availability more than anything else.
> 
>  There are some very good books like HP's "How to
> Make Your Car Handle" it is a
> starting point that goes into most of the questions
> that you need to ask.
> 
> Scott
> 
> Date: Tue, 14 Dec 1999 05:24:47 EST
> From: DemonTSi at aol.com
> Subject: Non- DIY EFI: question about four link/bar
> rear ends...
> 
> Hi folks, I was hoping to get some info from someone
> who has experience with
> fabricating a rear 4 link or 4 bar suspension setup.
> I'm working on designing
> and fabbing a 4 bar for my datsun roadster and was
> just wondering if there
> were any optimum lengths for the links, and distance
> between them that I
> should try and stick to. Right now I have the links
> at about 10.5-11" long,
> and 5" apart, center-to-center. Any
> help/advice/tips/comments would be great!
> Thanks in advance.
> 
> Van
> 
> ------------------------------
> 
> Date: Tue, 14 Dec 1999 17:30:16 -0500
> From: David Brode <dbrode at hereintown.net>
> Subject: Re: 4 links
> 
> Hello all,
>  Finally something I know a little about![humor].
> I've done a bit of
> chassis fabricating and help a friend with a 4 link
> equipped 3040 lb
> B-1 Dart that carries the wheels 90 ft and runs
> 9.20's @ 149
> [through the muffs].
>  Short "bars" will cause your pinion angle to change
> a bunch during
> travel. Most race car equal length "links" are 20"
> C/L to C/L.
> Unequal length deals are usually 20" on the lower
> bars and shorter,
> but as long as possible on top. For a street only
> car, Spread them
> out as far as possible to prevent bad body roll, and
> make them level
> front to rear. You can run them parallel to car's
> C/L, but you'll need a
> diagonal link or a panhard bar. Another option is to
> run the top bars
> [or all] at an angle, like many factory 4 bar
> setups. You could copy
> most any factory design. Running at least the tops
> at an angle will
> enable you to get away without a locator
> [panhard/track locator]
>  Fyi - Chevettes had a 4 bar set up that could be
> cheaply adapted.
> You can cut and re-weld them too.
> 
>  Now, if I could just get someone to tell me if 80's
> turbo 2.2 Chryslers
> had low or high impedance injectors.
> Good luck.
> Dave Brode
> 
> >
> > From: DemonTSi at aol.com
> > Subject: Non- DIY EFI: question about four
> link/bar rear ends...
> >
> 
=== message truncated ===

__________________________________________________
Do You Yahoo!?
Thousands of Stores.  Millions of Products.  All in one place.
Yahoo! Shopping: http://shopping.yahoo.com