|
|
|
Car has 195k miles on it 5 speed with Original bosch sensor. Just passed NJ emissions test 2nd toughest after Cali. Car runs great 23-25 mpg around town which is all it gets lately. Highway was (28-30) 16 inch tires. Tested sensor when just warming up and watched it rise from .2 volts to .88 volts in a few minutes and thats pretty much where it stayed. Tested again after good and hot and it was at .9 and never came down again both tests are at idle and with no vaccum leaks created. Used a beckman multimeter diconnected signal wire and reading from the sensor. I thought they were supposed to settle back into the .5 volt area. Again both tests were at idle only and the car is running great. So what am I missing?
|
|
-
|
|
|
You may be able to revive the O2 sensor by burning off any stuck on contaminants with a propane torch. There are YouTube videos.
--
1980 245 Canadian B21A with SU carb, M46 trans, 3:31 dif, in Brampton, Ont.
|
|
-
|
|
|
For the $12 bucks it will cost for a bosch cut and splice new one why would I bother?
|
|
-
|
|
|
it has to swing between .2 to .8v about twice per second. 195k-it's dead. get a direct replacement-no splicing or soldering. good luck, chuck.
|
|
-
|
|
|
Is my MM going to show those fluctuations? It should come down some I would think even at Idle? I have a direct fit I bought yrs ago and also purchased some bosch units for older fords which will require splicing. 1/6 the cost of the original.Now just have to get the CORRECT socket for it. Evidently there are lesser quality ones out there. Puzzled as to why the mileage and performance are fine stiil?
|
|
-
|
|
|
A few comments about your multimeter...
1. Not all meters are created equal, and you need a high impedance meter in order to respond to the oxygen sensor output. I'm just guessing here, but I think you would want a meter with 5 mega-ohm or better impedance.
2. I like analog meters better because you can watch the needle swing as the voltage changes - the digital ones update periodically and may not represent the signal very well.
3. I bought a cheapie analog meter and it was NOT up to the task - impedance was not high enough, so it overwhelmed the sensor output...didn't really show anything.
4. If you are using an "auto ranging" multimeter (pretty common in digital meters) you might find that you have to put it in "range hold" mode to keep it from bouncing around trying to dynamically change the range. You might hook it up to a 1.5 volt battery and let it auto-range to that voltage then press the range hold button...otherwise it might lock onto a range that is too small.
5. Wouldn't a handheld scope be great for this job?
|
|
-
|
|
|
My MM is probably as old as the car so there is no load ranging feature.Im going to break out the (Simpson 260)20,000 ohm impedance if i recal correctly? next and give that a go as that big needle will bounce back all over the place if the sensor is working properly.The way this car is running I dont think it has failed. But if it has and I can improve mileage, its worth replacing.
|
|
-
|
|
|
Simpson 260, Geez that is old! Its input impedance is too low, you need a couple mega ohms, which most $20 DVMs have these days. And you will see the volts bounce around on one of those.
That O2 sensor is probably bad, and if you replace it your car will probably run the same. If you check the voltage with the O2 sensor connected to the ECU, which is just a fine way to do it, you may read about 0.5 volts, which the ECU puts there as a default if the O2 sensor is bad (probably an open circuit).
The car will run fine.
YOur biggest problem is getting the old sensor out of the Cat. After this many years it may just rip a hole in the case if you can grip it and apply enough force.
In which case you are in for a new converter.....
Been there, done that..
--
93 245 Classic, M47, IPD, 91 245SE AW70, 89 745, M46, IPD, Dependents-00 S/V 40s
|
|
-
|
|
|
Testing it from the signal wire from the sensor so the ECU is not in play. I am thankfull that its mounted on the exhaust manifold and not the cat or anywhere next to it. If the car isnt going to run any better or give me better mileage than why would I change it? simpson 260 should be fine to use as I am just reading from the sensor itself as I am not probing into the circuit
|
|
-
|
|
|
regarding the other post-if you want an ee, then look for art benstein's posts. he's an ee and may have---in previous posts---gone over this. seriously, stop then debate, 195k, replace the o2. good luck, chuck.
|
|
-
|
|
I agree, Chuck. If this is anything but a quest for curiosity, get a new sensor in there so there's a baseline of expected performance.
However, if this is for "educational purposes," then look up the propane torch test for an oxygen sensor. Must be done removed from the car, of course, but it works. You can use the torch to clean the soot.
I also have connected one to a Simpson 260 series 6 - 20,000 ohms per volt, finding the load of the meter reduces what you read in voltage produced by the sensor, compared with a 10M input meter.
Analog or digital makes no never mind; the analog is easier to follow than bouncing numbers, but who (besides old hoarders like me) keeps a an old VTVM (vacuum tube volt meter) around these days. A scope is the best method, but for car folks likely to have a meter of any kind, it is their DMM (digital multimeter) that will have the 10-meg input spec.
The sensor has to be hot - something like 600F IIRC - to output a useful voltage.
Another thing that fools people a lot on these Bosch systems is the ECU maintains a weak (current limited) 0.5V voltage on the oxygen sensor input sometimes called the reference voltage. The sensor has to be hot to overcome this with its own output and then vary it lower for lean and higher for rich.
By the way, I killed another sensor with an oil leak. This harkens back to our discussion of where the reference air is drawn from. There's this nice picture in the LH2.4 green book... showing where the oil wound up. The Bosch universals come with a really sweet splice box these days.
--
Art Benstein near Baltimore
"He's the best mechanic in town." - 1992 Annemarie Powell said of General Colin L. Powell, a Volvodad
|
|
-
|
|
|
i get why analog is easier, just worry about voltage when connected to the ecu. the 10m spec has been around for years. for the propane test, well, it's not connected to the ecu-no problem.
i found this scope a few years ago for 35, been a big help because of it's size. easier to carry than my 4channel interros-
https://www.storeslider.com/craftsman-professional-multimeter-231333378106e.html
seeya, chuck.
|
|
-
|
|
|
20KOhm impedance is not enough. The problem is that the O2 sensor doesn't really have any power behind the signal...so if you try to drive a 20k load with it, the voltage will droop. That's my limited understanding, anyway.
FWIW my cheapie analog meter was (*I think*) 20KOhm impedance, and it didn't show jack shit.
Try it..but if you don't get a signal, don't conclude it's a bad sensor.
Here they say 10 MOhm is the 'typical' recommendation:
http://www.swedishbricks.net/faq/fio2.html
Maybe an EE will come along and 'splain us.
|
|
-
|
|
|
make certain the meter has 20k impedance. i'm no ee but was told never use an analog meter, only digital when dealing with ecus. analogs allow to much voltage into the circuit. good digitals have 10k built into them and prevent that.
yes, a dvom should be fast enough so show the signal swing. all i had before i got scopes. good luck, chuck.
|
|
-
posted by
someone claiming to be bob
on
Thu Nov 6 20:47 CST 2014 [ RELATED]
|
|
It has been a while but I seem to remember that the signal wire needs to be connected to show the fluctuation. Should flex between about .2-.8
|
|
-
|
|
|
Haven't read all the posts here yet, but Bob, you hit the nail on the head. The "fluctuation" is a result of the feedback SYSTEM in operation -- in as some call it, closed loop. And if it isn't fluctuating, that does NOT mean the sensor is at fault.
--
Art Benstein near Baltimore
"Well, brand my terahertz DNA unzipper!"
|
|
-
|
|
|
http://www.ehow.com/how_8017505_test-o2-sensor-preobd.html
So this test is faulty as its an open loop test? Sensor wire disconnected just reading from the sensor. ECU not involved
|
|
-
|
|
|
I read through that e-how article twice, and yes, it sure looks like Johnathan forgot the step where you connect the sensor back to the harness.
Easy to do too, when you write for a medium that's mostly advertising and not seriously subject to peer review. That's one reason we like the BB.
I see in your last reply to Trev's suggestion, you are not really interested in the details but just looking for the easiest way to get from A to B. In that case you are right. Just buy the $12 universal and be done with it. I use impact to remove the sensor from the head pipe or the cat. I use a stubby 22 combination wrench and a 3-lb hammer. That's my advice.
On second thought, if it runs great and passes emissions, why test it? Wait until it stops doing that, unless you're thinking the sensor will become one with the pipe or cat in that period of time. Or just test the system responds to it, like you're supposed to, closed loop.
--
Art Benstein near Baltimore
"Any ship can be a minesweeper....Once."
|
|
-
|
|
|
Thanks Art for the clarification of the test procedure.I dont think the sensor has failed and will test it closed loop.At this price to get a 3 wire bosch to cut and splice I was thinking why not replace the original one?
http://www.ebay.com/itm/380876501041?#shpCntId
|
|
-
|
|
|
I got some of those exact ones for me and a friend. I have not used one, he did. He backprobed it and the voltage stayed low, did not range at all. I wanted to try one of mine but haven't yet.
I don't know why, maybe someone has a thought?
|
|
-
|
|
|
i should have said 10 meg, not 10k. much higher. sorry, chuck.
|
|
-
|
|
|
One word, fluctuations!
It appears lazy or the system is not responding to the signal.
Phil
|
|
|
|
|
