• Welcome to the new NAXJA Forum! If your password does not work, please use "Forgot your password?" link on the log-in page. Please feel free to reach out to [email protected] if we can provide any assistance.

O2 sensor testing, ruminations on Renix closed loop mode

n1ywb

n1ywb

NAXJA Forum User
Location
The Green Mountains
So today I broke out the old analog volt meter and hooked it up to my O2 sensor. At idle, it swings rhythmically from about 2v to 3.5v at a frequency of about 0.5hz give or take. Based on what I know this indicates that the sensor is probably operating correctly. My understanding is that the O2 sensor doesn't really measure the mixture, but rather when the voltage crosses 2.5v it indicates that the mixture is stoichiometric, and the ECU continuously tweaks the mixture so that it swings back and forth over 2.5v constantly and thus maintains something close to stoichiometry. Can anybody confirm or deny this?

Also when I ground the O2 sensor, the engine starts to run rough, which I assume indicates that I am in closed loop mode. If I was in open loop mode, then grounding the O2 sensor should have no effect, correct?

It's also worth noting that my knock sensor has been knocked off the engine block. It still seems to run in closed loop, so I guess the ECU doesn't care. According to the official Renix manual, it's not very useful anyway. It just retards like crazy whenever it detects knock, then slowly ramps up again, it's not smart enough to settle at a timing value thats just below the knock threshold, like some more advanced FI systems. Again can anybody confirm or deny this?

According to the Renix manual, only four sensors are tested at key-on for closed loop operation: MAP, coolant temp, intake temp, and TPS. The ECU tests these at key-on and will bork to open loop if it detects an open or short. I'm not sure why it leaves O2 off the list, obviously it needs that to run closed loop. It's possible it doesn't get tested at key-on, perhaps it can't be tested, at key-on it wouldn't put out a useful value anyway.

One very interesting thing in the Renix manual:
Note: If the ECU does not receive input from the coolant temperature sensor, it will simulate the sensor input by defaulting to the input from the air temperature sensor and add 1/2 degree C for every 85 engine revolutions.
Click to expand...

So I guess the coolant temp sender isn't even really required for closed loop, it probably just takes longer to enter closed loop if the engine is already warm.

I guess this means that problems with the other sensors (CPS, TDC) won't force closed loop although obviously that would still affect drivability by messing up injection and spark timing.

Does anybody have a table to corrolate MAP sensor resistance to barometric pressure? The Renix manual says the DRB-II tool can be used to test the MAP by displaying it's reading in millibars and comparing that to the current pressure at your location corrected to sea-level.
 
n1ywb said:
So today I broke out the old analog volt meter and hooked it up to my O2 sensor. At idle, it swings rhythmically from about 2v to 3.5v at a frequency of about 0.5hz give or take. Based on what I know this indicates that the sensor is probably operating correctly. My understanding is that the O2 sensor doesn't really measure the mixture, but rather when the voltage crosses 2.5v it indicates that the mixture is stoichiometric, and the ECU continuously tweaks the mixture so that it swings back and forth over 2.5v constantly and thus maintains something close to stoichiometry. Can anybody confirm or deny this?

Yes, that is what you want to see, means it is working properly, but it is not a true stoich, it is 14.7:1 which is a prefered ration, in other words there is lots of excess O2.

Also when I ground the O2 sensor, the engine starts to run rough, which I assume indicates that I am in closed loop mode. If I was in open loop mode, then grounding the O2 sensor should have no effect, correct?

Yes.

It's also worth noting that my knock sensor has been knocked off the engine block. It still seems to run in closed loop, so I guess the ECU doesn't care. According to the official Renix manual, it's not very useful anyway. It just retards like crazy whenever it detects knock, then slowly ramps up again, it's not smart enough to settle at a timing value thats just below the knock threshold, like some more advanced FI systems. Again can anybody confirm or deny this?

Yes, correct.
According to the Renix manual, only four sensors are tested at key-on for closed loop operation: MAP, coolant temp, intake temp, and TPS. The ECU tests these at key-on and will bork to open loop if it detects an open or short. I'm not sure why it leaves O2 off the list, obviously it needs that to run closed loop. It's possible it doesn't get tested at key-on, perhaps it can't be tested, at key-on it wouldn't put out a useful value anyway.

Not correct. It starts in a rich open loop mode, until the O2 sensor starts reading in a reliable range. The O2 sensor has a built in electric heater that takes 15 to 30 seconds to get the O2 sensor into operating temperature range.

One very interesting thing in the Renix manual:


So I guess the coolant temp sender isn't even really required for closed loop, it probably just takes longer to enter closed loop if the engine is already warm.

It will start and run fine with out a CTS, but I don't think it will go into closed loop with out a CTS. It will start with out a TPS, but does not accelerate well with out a TPS.
I guess this means that problems with the other sensors (CPS, TDC) won't force closed loop although obviously that would still affect drivability by messing up injection and spark timing.

TDC? No spark with out a CPS signal.

Does anybody have a table to corrolate MAP sensor resistance to barometric pressure? The Renix manual says the DRB-II tool can be used to test the MAP by displaying it's reading in millibars and comparing that to the current pressure at your location corrected to sea-level.
Click to expand...

I thought the table was listed in the Renix MPI manual?
 
Thanks for the feedback.

TDC, whatever the hall effect thingy in the distributor is called.

No MAP table in the manual, just says to look at the DRB-II screen to get the reading in millibars and to compare that to a handy barometer. Oh well it passed the ohmeter test, it's probably fine.

I didn't think it would go into closed loop without the CTS, but the manual makes it sound like it will, after an essentially predetermined delay for the engine to warm up. It sounds like the air temp sensor is actually more critical. If I were more motivated, I might test this hypothesis.
 
n1ywb said:
Actually according to Wikipedia 14.7 : 1 IS the stoichiometric ratio for gasoline and air by weight: http://en.wikipedia.org/wiki/Stoichiometric#Stoichiometric_air-fuel_ratios_of_common_fuels
Click to expand...

Yes it is the actual ratio used, but it is not a true stoichiometric ratio. In chemistry a stoichiometric ratio implies that all of the reactants are consumed. At a 14.7:1 ratio there is about 5% oxygen left unused, which is what the O2 meter is reading.

I am one of those picky AH's ( Chem Eng) that gets annoyed when an industry (EPA and Automotive) missuses a well defined scientific word. EPA redefined the word Solid as "any Liquid, Solid, semisolid, or containerized gaseous material, when it wrote the waste regulations. Another good example.
 
Last edited:
So here's one other question. Even though the O2 sensor seems to be working, is it possible that it's still goofed up and throwing off the mixture and thus hurting mileage? How could I verify that, would a sniff test at the smog station show it?
 
I've heard of that happening, but more to the point that the sensor is slow to respond, therefore causing bad mileage. I tested my Renix by watching that analog meter while I was driving on the open road....counting how many times it crossed the 2.5v mark every second. I'm thinking it should be 5 times per second, might be wrong.

I'll refer you to my thread dealing with this, and diagnosing a bad sensor here. Granted, it's a few months old by now, but it should be relavent.
http://naxja.org/forum/showthread.php?t=947676
 
I don't have any hard facts, but the O2 sensor is read as a resistor in renix, so bad contacts in connectors could add resistance and it seems reasonable that an O2 sensor could be biased high or low just like any other sensor, they upper and lower error ranges and no sensor will read exactly the same even new out of the box.

Also, something new I just thought of, is the reference resistor and reference voltage in the Renix computer could be off a little one way or the other, like 5.3 or 4.7 volts instead of 5.0 volt supply, and say 1000 instead of 1100 ohms on the reference resistor. Either might make the system run a little rich or lean.
 
Sure. Component tolerances are engineered into electronics, most resistors are only 10% accurate, although some are very high precision, like 1% or better. The DAC in the ECU probably operates in differential mode though, one input is the +5 reference signal, the other input is the O2 output, and it measures the difference between the two. That way fluctuations in the +5 reference line, no matter how large, don't affect the relative output.

That in no way eliminates the possibility of a biased O2 sensor. Bad connections could definately throw it off, the added series resistance would throw the value. Even in a voltage generating O2 sensor, extra series resistance would probably cause problems, since the voltages and currents are so low.
 
Ecomike said:
I don't have any hard facts, but the O2 sensor is read as a resistor in renix, so bad contacts in connectors could add resistance and it seems reasonable that an O2 sensor could be biased high or low just like any other sensor, they upper and lower error ranges and no sensor will read exactly the same even new out of the box.

Also, something new I just thought of, is the reference resistor and reference voltage in the Renix computer could be off a little one way or the other, like 5.3 or 4.7 volts instead of 5.0 volt supply, and say 1000 instead of 1100 ohms on the reference resistor. Either might make the system run a little rich or lean.
Click to expand...
Reference resistor 3900 ohms.:shhh:
 
You must log in or register to reply here.

Similar threads

O
2000 XJ Failed Colorado Emissions high CO
2
Replies
27
Views
2K
Ken M
K
9
Renix lowish idle
Replies
10
Views
1K
90xj06
9
my2monkeys
Bosch o2 sensor #12009 for Renix. Still all metal body?
Replies
1
Views
1K
Saudade
Saudade
F
Curious MAP sensor failing - symptoms
Replies
0
Views
2K
fenns7
F
F
Stumped on this one!
2
Replies
29
Views
4K
Bent
Bent
Back
Top