NetNews Usenet Archive 1992 #23/NN_1992_23.iso/spool/rec/autos/tech/13898

Text File | 1992-10-12 | 6.1 KB | 130 lines

Newsgroups: rec.autos.tech
Path: sparky!uunet!walter!porthos!eve!dje
From: [e-mail] (Don Eilenberger)
Subject: Re: oxygen sensor
Reply-To: [e-mail] (Don Eilenberger)
Organization: Bellcore, Livingston, NJ
Date: Mon, 12 Oct 92 14:06:11 GMT
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Sender: [e-mail] (USENET System Software)
Lines: 117

In article <[e-mail]>, [e-mail] (Chris Elmquist) writes:
|> So, there are only a couple different O2 sensor designs ?  Do they all
|> exhibit the same O2 vs. Vout curve ?  I suspect that mine (in my '87
|> Integra) may be failing and causing the rotten idle problem I've had
|> for a couple years(!).
|>
|> I'm wondering if I can disconnect the sensor and fake a nominal Vout
|> using a battery and a couple resistors (or a pot)...  If they have a 0v
|> to 1v range, can I feed .5v say, into the computer in place of the O2 sensor
|> and see if the idle behaves differently (better maybe) ?  Or is the output
|> of the O2 sensor much more dynamic and related to engine RPM such that a
|> constant Vout (or Vin to the computer) would really dork up the idle...?
|>
|> It would be nice to diagnosis the O2 sensor as bad before trying to pull
|> it out...
|>
|> Chris


Chris.. and all:

1. This is a very BAD idea... if you happen to accidently cause a spark,
or some other electical disturbance - check out the price of a new
computer...

2. Testing Ox sensors: Basic procedure is actually quite simple, and only
really requires a brain and a good quality VOM (volt-ohm-meter).

Step 1: Measure the output of the sensor while connected at idle, after
the engine has warmed up (especially important on sensors that are 3 wire
[ie:heated] and located somewhere down the exhaust from the manifold). The
output should be in the range of .1 to 1 volt - usually a correctly working
sensor on a well tuned engine averages about .5v. The output SHOULD be
fluctuating - about +/- .1 to .2v at a rate of about 1 to 2 times per
second. If these conditions are met - increase the engine speed to about
3,000 rpm.. the rate of fluctuation should increase - but may be masked
by the inertia of your VOM.

Step 2: Providing the sensor passed the above test - pull your oil dipstick
up an inch or two while the engine is running. This will induce a lean
condition in the engine (via excess air entering the PCV system). The
voltage from the sensor should DROP momentarily, indicating that the
sensor has detected the lean condition. If your sensor and computer
are working correctly - the voltage should quickly recover (the sensor
told the computer that the engine is running lean - and the computer
richened the mixture to compensate).. at this point, with the dip stick
out an inch or so, the voltage should have returned to the point it was
at before you pulled out the dipstick.

Step 3: Push the dipstick back in. The voltage on the sensor should rise..
usually about .3-.5v over the "normal" voltage. This means the sensor
has detected a rich condition (you've closed off the air leak). The voltage
should again return to normal after a second or so.

If the sensor passed the three tests above - it's about 90% certain that
it is still working.. but if you'd like to go further:

Stop engine, and disconnect sensor (this is called the "open-circuit"
test)...

Again - attach voltmeter to sensor output wire and ground, and:

Start engine: voltage should be about where it was in the tests above,
somewhere between .1 and 1v - with some fluctuation seen in the output.

Repeat test 2 above: response on sensor should be immediate - and the voltage
should STAY low (no feedback by computer to injectors)...

Repeat test 3 above: voltage should return to normal - without peaking
over the normal voltage (again - no feedback)...

If ALL these tests were OK, here is the general scoop on Ox sensors:

1. They rarely fail entirely - ie: giving NO voltage

2. The usual failure is slow response, with a slight loss in output
voltage... if you don't see fluctuations in output at idle, but just
a steady voltage - the sensor is probably worn out.

3. Factory recommended replacement intervals are probably fairly accurate..
my BMW recommended replacement at 50,000 - I replaced it at 60,000 and
immediately noticed a large improvement in how it runs...

4. The effects of a bad sensor are NOT most noticable at idle.. a slow
sensor may be able to respond to the engine at idle - but NOT at running
speeds. Symptoms of a bad (slow) sensor: engine surging at speed.. stalling
when coming down from speed to idle - but these symptoms can also be
caused by other problems.

Also - most modern automotive computers do not measure the
output from the sensor for the first 30-120 seconds after
an engine is started, since it takes this long for it to
warm up and start working.. so if the car works fine during
this warmup - and then runs poorly, this could be a hint
that the sensor is shot (or lots of other things could also
be wrong)..

As far as testing one out of the car - unlikely since they work at
exhaust temperatures (500 deg. and hotter).

I noted an earlier post mentioning the basic design of sensors isn't
really much different - except from 1 wire to 3 wire (the 3 wire
has a heater for faster response from a cold start).. well.. a long
story short: BMW wants $180 (list) for the 3 wire Ox sensor.. I replaced
it with a Ford sensor (DV606) - which happens to also be a 3 wire
sensor made by Bosch for about $37.. aside from the connector on the
wire they appear identical - and the new sensor passed all the in-car
tests that are in my manuals, and the car runs much better. Of course
YOUR results could be different.

Don Eilenberger ([e-mail])

DISCLAIMER: Any advice is worth what you pay for it, in this
case nothing. Your results may vary, or you may be a total
klutz who blows up your car. The information given here is
JUST information - not a recommendation or suggestion as to
actions you may take. So there.