Widebands and exhaust backpressure.
 

I spent saturday on the dyno and my wideband and the shops were confirmed as reading the same at the start of the day. Get through all the low load tuning and start doing full throttle pulls and we notice that the two widebands are reading different by as much as .8 AFR, mine being the richer reading and 8" back from the turbo with a copper heatsink, the shops in my tail pipe. Some testing and both widebands would read the same in the same location. What happens is exhaust backpressure can cause a false rich reading!

Did some looking on the subject and this shouldnt be big news, but its something that doesnt seem to get talked about and is a big deal as I thought my safe 11:1 tune on my last motor was likely a borderline tune, it sure did make good power tho lol. Definatly an issue to keep in mind, especially for those that have a "restrictive" exhaust.

I wouldn't think back pressure would effect on the sensor.

The air at the exhaust manifold is the same at the muffler.

What really throws the readings off are exhaust leaks and airpumps/cats.

http://wbo2.com/lsu/lsuinstal.pdf look at the very end of the document. many install guides mention this, its just easily overlooked.

nice find, now we need back pressure sensors to tell our widebands how much we are off :D

i was pretty suprised at how much off it was off. I have the racing beat catback and a resonator in a 3" exhaust, wastegate is rerouted. pretty much from the point the wastegate opened is when the reading went off. Makes me wonder if placement after the reroute pipe would help any.

what about those just running an open dump? couldn't you put the sensor in the dump tube being there would be no backpressure? obviously this would only work during WOT tuning, or maybe just tune from the tail pipe and use the in car wb02 as an insurance guage.

Were you running racegas or any kind of leaded fuel? Also did you have a Bosche sensor or the NTK.

Thanks,

Anthony

Nothing special 91octane from the pump, I have a bosch sensor with an innovate. I didnt note what brand the shop had.

for those wanting the best data while dynoing i suggest adding a second WB bung for the dyno. there are a number of factors that work against the tailpipe clip-on. best place is before the midpipe connection and a good distance from the turbine housing.

hc

How much pressure were you seeing in the downpipe? At what boost level?

Not sure what the actual pressure in the downpipe was, we were talking about making a fitting to test this but I was more interested in getting the tune finished. The dyno session was mainly to get the car reliable and driveable, we only tuned it up to 10psi. I have been planning on changing the exhaust for some time now, but before I do it wouldnt be too hard to fab up a fitting and use the haltechs onboard map sensor to log the downpipe pressure.


Howard, most manufactures seem to reccomend around 1m back from the turbine housing on turbo cars for wideband placement, this is for heat reasons. The tailpipe does have problems, it works the WBO2 heater hard and may not get consistent readings at low loads because there is potential for air dilution, for a dyno session especially the full load tuning there shouldnt be any drawbacks of a tailpipe reading, assuming there is no emmisions equipment hooked up and the sensor is in good health.

EDIT: point I wanted to make with this thread is just becuase your wideband says something and its calibrated as accurate doesnt mean there arent other factors influencing the reading. In the past I blindly accepted my wideband as being accurate as I did check the calibration periodically, and that likely factored into my last motor letting go.

You were most likely seeing reversion on the dynos sensor and may or may not be overheating the sensor that is installed in your downpipe.

When you try to compare one wideband to another, there is a tendency to take one as "correct" and one as "incorrect" or "off." The problem is that no matter where you locate a wideband o2, there's going to be some factor working against its accuracy. A vehicle is not the same as the controlled lab where these sensors were designed and tested.

To give a simple explanation of the sensor's output: basically the further the mixture is from lambda = 1 (about 14.7:1 AFR), the more current produced.

https://www.rx7club.com/attachment.p...1&d=1272402657

The polarity of that current depends on whether it is lean or rich of stoichiometry.

https://www.rx7club.com/attachment.p...1&d=1272403188

In operation richer than 14.7:1, higher exhaust temps (relative to standard lab conditions) result in a richer reading and so does higher backpressure. This is because the current signal increases (absolute value of the signal) but the polarity doesn't change. So at a given AFR richer than lambda = 1, the signal would be say 1 mA under standardized lab conditions. With high pressure and temperature the sensor would read say 1.2 mA, skewing it richer:

https://www.rx7club.com/attachment.p...1&d=1272404354

When the engine isn't operating in steady state cruising conditions the exhaust Pressure and temperature are constantly changing, so your sensor's accuracy relative to its reading under standard lab conditions is always changing to some degree.

I have put a lot of study into this issue as I have recently been getting into Subaru tuning. On most turbo Subarus, the factory wideband is located before the turbo (yes it can take that kind of heat). You see on almost all Subarus the two cylinder banks collect (boxer engine) and then that collector pipe "up pipe" feeds the turbo inlet. The up pipe has a cat in it on a lot of them, and the factory wideband goes before that.

With the stock location, the Subaru factory wideband will basically peg rich once you get into boost. If you eliminate that up pipe precat and locate the wideband after the turbo you will get a leaner and more believable reading. Here is somebody's comparison of the readings from a factory Subaru preturbo wideband (no preturbo cat) on a Subaru and a post turbo Innovate/Bosch sensor:

https://www.rx7club.com/attachment.p...1&d=1272403188

You can see the exhaust backpressure and heat skews the reading (Subaru factory sensor pegs rich at 11.2:1). But as I said before, no matter where you put it there is some concern about temperature, pressure, or airflow effects. Widebands are just tools, they have their limitations. That's why it's important to build a safety margin into your tune--both fuel and timing--and you will be fine. In some cases that requires the discipline to say "ok, I'm going to eschew internet bragging rights and intentionally make less power than I could, for the sake of reliability 1, 2, or 3 years from now."

How far into the exhaust stream does the sensor need to protrude? I recently got a LC1 and it came with the "extended" bung. The sensor tip doesn't stick out of the bung but 1 or 2 MM.

Any general rules?

Exhaust on a turbo car is extremely turbulant so it shouldnt matter how far out it sticks. The extended bung is more heat reasons and if 1-2mm is exposed you can be confident that its getting enough exhaust to have a reading.

I'm going to assume you are recirc your wg.

If you measure it before it comes back into the exhaust that would make a difference also since one wb wouldn't see all the exhaust and the other one would.

Yes its rerouted... I did think about one wideband would be seeing the reroute exhaust where the other wouldnt, but it all comes from the same source so its all the same exhaust. My tuner mentioned that hes seen false rich readings before when the sensor is in a high pressure area. and then the manufactures confirm it in their manuals. As Arghx also posted the high heat zone can effect the reading as well.