O2 Sensor Location

**JetspeedCamry** · 22-08-2006, 06:45 PM

Hi,

Just a question regarding the prefered location of an O2 sensor. I have custom extractors made up and currently, my O2 sensor bung is located on one of the cylinder pipes, closeish to the head, rear bank, cylinder 1.
Looking at the stock manifold, the bung is located at the collector of the rear bank pipes (1-3-5). So its getting its signal collectively from 3 cylinders, as opposed to now, where its only gettign it from Cyl 1.
Is there any difference in the O2 sensors placement?

Thanks

**oldcorollas** · 22-08-2006, 07:08 PM

ideally.... get a heated sensor and stick it down at the final collector

then it doesn't matter how far away it is, and you sample all cylinders.

failing that, if there is an issue with the single cylinder it is sensing, then you could have issues.. at cruise/light throttle...

as long as that cylinder is representative of all, it is probably ok

**WindeX** · 22-08-2006, 07:48 PM

Mine has been running fine for over 6 months with the sensor only running off two cylinders, I guess for the finest tuning you could run one off each cylinder

But it runs fine as is.

**JetspeedCamry** · 22-08-2006, 07:55 PM

Ok sweet! Just making sure its not a 'critical' thing

Thanks

**AndyTTR** · 24-08-2006, 10:22 AM

The ecu responds to changes in A/F ratio faster if the sensor is closer to the head. On the other hand, it will be more accurate if it is placed in the collector - getting the average of two or more cylinders. It is normal for some cylinders on the same fuel rail to run lean / rich due to the distribution of fuel pressure between injectors.

Its not really relevant to this thread, but we've burned holes through pistons while trying to fine-tune A/F ratios on our engine dyno. The lambda sensor was placed in the collector for cylinders 2 and 4, but cylinder 3 ran lean (mostly due to bad fuel rail design). This happened even though the ECU was reading a healthy A/F ratio. Now we run two lambda sensors and thermocouples on each cylinder.

**JetspeedCamry** · 24-08-2006, 12:44 PM

Well at the moment, the sensor is quite close to the head, but only on Cyl 1. Since with the extractors, the collector is further down near the sump now (as opposed to the original manifolds which have it collect at the bottom of the head). I'll probobly just leave it for now as is, seems to be ok.

**oldcorollas** · 24-08-2006, 04:00 PM

i would suggest the response time of an O2 sensor, vs the number of cylinder events per second, is very slow.... however, a hot O2 sensor responds quickly.

when tuning my megasquirt, with a heated bosch sensor past the gearbox (ie, a total of about 5 feet from the head), any changes i made in fueling turned up on screen with no noticable response time....

i would not hesitate to run a heated sensor that far away. and having O2 sensor too close to head, esp for a car with work on it, might actually overheat the sensor (they have a reasonably small window of temperature operation.. like 400-600?)

unless you have issues like Andy, then samplign a single cylinder is ok.. but i would either run (as he said) two in collectors for a 6 or 8, or a single one at the final collector... but definitely run a heated one (you can use heated if it was not used stock, just set up a 5A capable cable triggered from IGN and have a good earth for it)

**JetspeedCamry** · 24-08-2006, 05:18 PM

Well this motor uses a 4 wire heated sensor stock, so i might get a boss welded down further past the final collector at some stage in the future. Is there any problems with extending the wires on these things? Id just cut it half way, and solder in an extra length of wire with some good heat proof insulation.

**oldcorollas** · 24-08-2006, 05:23 PM

not really.. as long as there is no massive increase in resistance, and the signal wire is shielded.

the O2 sensor is just a low voltage generator, so as long as the wire is of enough size (stock wires are smallish anyway... except for the heater wires.. which should be about 5A capable) it should be fine. any coltage produced by hot dissimilar wire junction will be small compared to the sensors voltage.

some of them have wires that are a little hard to solder.... but enough heat should do the trick

**JetspeedCamry** · 24-08-2006, 05:44 PM

Cheers, thanks.

**Mos** · 25-08-2006, 04:15 AM

Extending O2 sensor leads is fine. There is virtually no current flow in the signal, so the wire diameter does not need to be of any significant size (like brake lines vs radiator hoses- primarily transferring pressure with little volume flow versus emphasis on high volume). Many signal wires are only that size because it's a minimum practical gauge - needs some tensile strength without needing the current capacity.

Some shielded cables have plating that will not wet with conventional solder - turning up the heat can remove the plating, but my preference is using cables that solder properly

The factory sheilded cables [that I have extended] soldered without problem.

Originally Posted by AndyTTR

It is normal for some cylinders on the same fuel rail to run lean / rich due to the distribution of fuel pressure between injectors.

Its not really relevant to this thread, but we've burned holes through pistons while trying to fine-tune A/F ratios on our engine dyno. The lambda sensor was placed in the collector for cylinders 2 and 4, but cylinder 3 ran lean (mostly due to bad fuel rail design). This happened even though the ECU was reading a healthy A/F ratio. Now we run two lambda sensors and thermocouples on each cylinder.

Andy, I'm curious as to what pointed to the fuel rail rather than the intake track?

Mos.

**AndyTTR** · 25-08-2006, 07:21 PM

From memory it was the position of the pressure regulator, i cant remember exactly where it was now (either between injectors 3 and 4 or on the end after 4) but the result was a very uneven distribution of pressure between the injectors.

The engine was a F3 from a Honda CBR600 that we were running in our FSAE car. The engine was factory carby so the team designed custom intake / exhaust manifolds and a EFI system for it. I suppose burning holes in pistons makes for good learning experience. Plus the holed piston went on to become the 'Tool of the Week' trophy.

**kingmick** · 25-08-2006, 08:08 PM

Having it in the collector for all 4 pipes etc lets you run it at the right AFR for all pots as a total, you might have 1,2 even 3 pots rich but you will not have the 4th lean. Does not always work that way but it will in most cases.