Re: OBDII, O2 sensors, etc.

From: Jon Steiger (stei0302@cs.fredonia.edu)
Date: Sat Sep 12 1998 - 16:23:21 EDT


At 08:24 AM 9/12/98 -0400, you wrote:
>Jon,
>
>This is my first actual post to the digest, although Tim May and Gene Beaird
>have been indirectly posting some things for me in the past (I checked the
>archives <g>).

   You aught to post more often; great info!! :-)

>
>I have some experience with PCMs and OBDII in general, and O2 simulators
>specifically, so I thought I should post some corrections and additions to
>the recent post(s) on the subject.

  Ahh, so you must be the guy someone referred to who has a GC O2
simulator. :-) You obviously know a heck of a lot more about this
stuff than I do. I'm just shootin' in the dark, armed with a FSM. :-)

>
>>(Basically, the computer compares the readings from the downstream sensor to
>the upstream sensor, and it expects to see more oxygen by volume at the
>downstream sensor than what the upstream sensor is reporting, since the cat
>is supposed to remove the extra junk.)
>
>Not exactly. The upstream sensor is defined as the one(s) BEFORE the cat,
>while the downstream sensor(s) is/are AFTER the cat. The cat functions to

  Yep, I know.

>REMOVE oxygen from the exhaust, which is what the sensor is looking for. The
>downstream sensor therefore expects to see LESS oxygen.

   That I DIDN'T know. The cat removes oxygen? I figured that oxygen
was OK, so a cat would let it by, while trapping the hydrocarbons (or
whatever)

>
>>Each sensor has 4 wires. Two are for the heating element (to make the sensor
>readings more accurate I'm guessing), one is a ground provided by the PCM,
>and the other is a voltage to the PCM (0 to 1volts)
>
>The heater is to heat up the sensor so it begins to function more quickly
>after a cold start. O2 sensors don't function properly below 600 degrees, and
>the sooner they function the quicker the vehicle can go into closed loop
>(sensor controlled) operation. Cold start emissions is one area the EPA has
>been attacking recently, so all O2 sensors now must have heaters to get them
>functioning more quickly.

  Ok, that makes sense. I figured it would be something like that; many
sensors have to be temperature corrected, or only give accurate readings
at a certain temp, so this makes sense.

>
>>Its been a while since I looked at this stuff but if I remember correctly, 0
>volts would mean 100% oxygen, and 1 volt would mean 0% oxygen.
>
>Exactly the opposite is true. More voltage indicates MORE oxygen.

  Hmmm... You sure about that? My FSM says otherwise:

     "Two heated O2S sensors are used. The sensors pro-
      duce voltages from 0 to 1 volt, depending upon the
      oxygen content of the exhaust gas in the exhaust
      manifold. When a large amount of oxygen is present
      (caused by a lean air/fuel mixture), the sensors pro-
      duces a low voltage. When there is a lesser amount
      present (rich air/fuel mixture) it produces a higher
      voltage."

         (1996 FSM, page 14-32)

  Its possible that the FSM is wrong I suppose; stranger things have
happened. I assume that if the FSM was wrong, CC would have sent me
a correction? OK, how do we find out for sure? :-)

>
>>its very easy to see what would happen if you removed it; just unplug it and
>go driving. I did exactly that, and I drove for about 45 miles with no CEL
>(check engine light). However, as I was going up a grade, I was behind a
>slow car which I was about to pass. As I was waiting for an oncoming car to
>go by, I was in 3rd gear at about 3,000rpm for several seconds. That's when

>the CEL came on.
>
>The Catalyst Monitor (that's what the test function using the downstream O2
>sensor is called) is a "once per trip" monitor, which means that sensor's
>function is tested once per vehicle trip, after certain enabling conditions
>(such as reaching closed loop operation) are met. A trip is basically defined
>as a cycle where you: start the engine, drive (reach closed loop and meet
>some other criteria), and then stop the engine. The other enabling criteria
>make it a little more complicated than that, but for now this description
>should be adequate.

[...] (lots of good info deleted)

   Interesting. :-) I didn't know about the trip stuff or the pending
codes; that's good to know. (BTW: Where'd you find that info? Is there
a good place to look this stuff up?)

>
>>There are a few Mustangs without cats that I have heard of who are NOT
>getting CELs.)
>
>That simply cannot be true on stock '96 and up vehicles. If it were true,
>Ford would be in SERIOUS trouble and fined by the EPA, BIG TIME. Honda was
>recently fined many millions of dollars (I seem to remember hearing a figure
>over $200M) for inappropriately turning off the misfire monitor under certain
>conditions in some of their vehicles, so they take this very seriously!

   I don't have any personal experience with the cars in question, so I
may not have all the info, but one of the cars in question is a '96 Mustang
which was owned by a former DML member. As far as I know, he removed his
cat and installed an "H" pipe. I don't know wether or not the O2 sensor
was left in the pipe. If so, maybe the emissions were within spec, so he's
not getting a MIL? (Jarrod Pillone, I think it was. Jarrod, you still
on the list?)

>
>>So, there is at least one documented case of a CEL as a result of no O2
>sensor.
>
>I can assure you that in ANY '96 and newer (OBDII) vehicle, simply removing
>or disconnecting the downstream O2 sensor will result in at least one, if not
>several, DTCs. It would notice the open heater circuit, as well as the
>failure of the sensor itself.

  You're probably right. I spoke without having all the info; I don't know
for sure that they removed the O2 sensor. All I know is that they removed
the cat, and they had an O2 sensor simulator that they were going to use,
but they never got a trouble code, so they didn't use it.

>
>>If a CEL doesn't bother you, you can just reset the computer whenever you get
>one, but I'd rather not see it at all since then I've got to pull the codes
>and make sure the CEL is because of the O2 sensor and not a more serious
>problem.
>
>Sure, you could disconnect the battery, or use a Scan Tool to delete the
>pending or stored DTC (and thus shut off the MIL) after each first or second
>trip. However there are several reasons why IMHO this is NOT a good
>technique. <g>

   Agreed. I definitely don't want to go this route. I'd put the
cat and sensor back in first. :-)

>
>>So, I'm working on a module to simulate a downstream O2 sensor behind a
>functioning cat.
>
>No need to re-invent the wheel. I've been testing one for some time now, and
>so far it works perfectly. It was developed by a friend of mine several years
>ago initially for GM products, but a simple wiring and connector change has
>it working beautifully for Chrysler as well.

  Bummer; I already have most of the stuff I need to do it. :-) I might
give it a shot anyway, just out of curiosity.

>
>BTW, the circuit either may not be as simple as you might imagine, or in some
>ways may be simpler than you think. One thing to know is that most are
>looking for true O2 transitions, not just a steady voltage.

  Yep. A constant voltage would be nice and easy. :-) Basically, I was

just planning on spitting back the voltage from the front O2 sensor,
divided by ten. (simple voltage divider circuit)

>
>>This module will plug into the wiring harness using the same connector as an
>O2 sensor, and will also have to read the voltages being sent from the
>upstream sensor to the PCM.
>
>I can assure you at least for now, that type of connection is not necessary.
>The properly functioning circuit I am testing plugs directly into the
>downstream O2 sensor connector, and is powered by the heater circuit on that
>connector so no extra connections are necessary. It is a completely sealed
>(in fact potted) unit about the size of your index finger which is totally
>impervious to dirt and moisture intrusion. It has a hole drilled through the
>potting material for tiewrapping securely to the undercarriage.

    How are you simulating the exhaust stream without knowing what the
front sensor is reading? From what you've said, I'm gussing you're
spitting back a voltage that varies occasionally?

>
>>If anyone knows of a place to get the type of snap-together connectors used
>in the wiring harness and O2 sensors, that would be helpful, since this would
>allow someone to simply plug it in without having to do any cutting or wiring.
>
>In lieu of purchasing the connector, terminals and seals separately, I used
>the original stock downstream sensor, cut off the wires, attached a Packard
>Electric 4-pin weatherpack (sealed watertight) connector, which mates to the
>same type of connector on the finger sized simulator unit. It then becomes a
>direct plug-in to the Chrysler wiring harness.

  Yep, I was planning to cannibalize the connector from the downstream
sensor for the module, but I also wanted a male and female which I could

use to make a nice neat plug-in up at the front sensor. Also, in the future,
I plan to make a mixture guage which will read the voltage from the front
O2 sensor, so I'm going to want some of 'em for that project too.

>
>>As far as killing low end, I don't have enough info to be sure.
>
>The cat delete is normally worth 10 HP or more on most vehicles. I agree that
>with headers AND a low restriction DUAL exhaust, subsequently deleting the
>cat may produce so little restriction as to somewhat compromise low end
>torque. However with the stock manifolds, and either a stock or modified
>SINGLE exhaust, the cat delete really wakes things up. Not only power is
>improved greatly but the sound as well!

   Currently, I've got a set of JBA headers with 1.5" primaries which
feeds an ATR Y pipe with 2.5" arms and a 3" "collector", then a 3" pipe
to the back, where everything exits through a Gibson 3" single exhaust.

>
>>Once I get my truck back, I'll install my module to do some additional
>testing. Once I'm satisfied with it, I'll post the parts and info required
>to make one yourself on the DML page.
>
>I'll be interested to hear IF your module works, and how it was constructed.
>Remember that you'll need to successfully navigate at least two complete trips
>to verify no MIL, and you'll really need access to a scan tool to see pending
>codes, as well as to view the PCM's interpretation of what your module is
>doing. Without a scan tool you'll be testing blind.

   You're right about that; I'm just a poor backyard mechanic, so I'm
going about it blind. Just planning on building something, testing, if it

doesn't work, going back to step 1, etc. :-) I haven't actually tested it
yet (I haven't put it together yet), but here's my basic idea so far,
maybe with your prior O2 sensor experience, you'll notice a problem I
haven't:

   Upstream O2 Vout ------> 10K resistor -------> PCM downstream O2 Vin
                                           /
           PCM Ground ----> 1K resistor --/

      Heating element -----\
                          6 ohm resistor
      Heating element -----/

  The voltage output of the upstream sensor would go through a 10,000 ohm
resistor, while the PCM ground runs through a 1,000 ohm resistor. These
are joined at the other side, downgrading the upstream O2 sensor voltage
by approximately a tenth. This voltage is then fed to the PCM via the
downstream O2 sensor Vout wire in the harness. As far as the heating
element wires go, I'm planning to stick about 6 ohms of resistance between
them. (The FSM said that to test the O2 sensor, you should be reading about
5-7 ohms between the two wires, so I figured I'd just put the same
resistance there and maybe it would fool the computer.) Pretty crude
I know, but I figured it'd be worth a shot?

  Speaking of scanners, do you know where I could find the protocol for
the OBD-II connector under the dash? I'd like to try building my own
scan tool, since I can't afford a commercial one.

>
>>My eventual plan is to have a "quick connect" type of arrangement which I can
>use to swap the cat in and out with a minimum of hassle. (I'm planning to
>use stainless band clamps.)
>
>I found that band clamps unfortunately wouldn't give enough clamping force to
>prevent the muffler and tailpipe tip from rotating. Part of the problem is
>the anti-seize which I used to coat the pipes where they join, but that is
>important to prevent them from rusting together, thus facilitating an easy
>removal later. I ended up using some nice stainless steel muffler clamps,
>which have the advantage of never rusting, which looks better and allows them
>to be easily removed in the future.

  You're right about that. My tailpipe does rotate a little bit, though
not enough to hit on the truck or the ground. I suppose I could just
put a tailpipe hanger in there somewhere to stabilize it a bit. I don't
care for exhast clamps because they crimp the pipe and make it impossible
to take apart. :-( That's the reason I have my 2.5" -> 3" adaptor still
in my system. I tried to take the crimps out with a pipe expander, but
no dice. I'm going to have to cut it off. :-( I like the band clamps
because they make a nice seal and they don't crimp the pipe. They do
allow it to rotate though.

>
>Another thing I did as much for function as appearance was to have the pipe
>Jet Hot coated inside and out to prevent any corrosion from occurring. With
>this coating, the anti-seize, and the stainless clamps, it should be no
>problem to separate these pipes even years in the future. Scott.
>

  :-) Yep, that thing should come apart just as easily 100 years
from now! Nice setup.

                                               -Jon-

  .--- stei0302@cs.fredonia.edu ------------------------------------.
  | Affiliations: DoD, EAA, MP Race Team, NMA, SPA, USUA. RP-SEL |
  | '96 Dodge Dakota v8 SLT CC (14.80@92.97), '96 Kolb FireFly 447 |
  `----------------------- http://www.cs.fredonia.edu/~stei0302/ ---'



This archive was generated by hypermail 2b29 : Fri Jun 20 2003 - 12:09:49 EDT