Wideband O2 Question

[Cutlassefi]

Most evac fittings I've seen are in roughly the same location as O2 sensors are usually. Since the evacuated gases are a mixture of exhaust and possibly some fresh air from external leaks, I'd say you should definitely compare readings with/without the evac working and to play it safe, tune without.

X2.
In addition, OX2 sensors are temperature sensitive and if evac gases affect temp it will alter O2.
Use a sensor with an internal thermocouple.

Uhhh isn't that why they're heated? It is by the way . They're only inaccurate vs temp when too cold, i.e. first start up. Once they're up to operating temp they'll work fine.
.

[David Redszus]

Most evac fittings I've seen are in roughly the same location as O2 sensors are usually. Since the evacuated gases are a mixture of exhaust and possibly some fresh air from external leaks, I'd say you should definitely compare readings with/without the evac working and to play it safe, tune without.

X2.
In addition, OX2 sensors are temperature sensitive and if evac gases affect temp it will alter O2.
Use a sensor with an internal thermocouple.

Uhhh isn't that why they're heated? It is by the way . They're only inaccurate vs temp when too cold, i.e. first start up. Once they're up to operating temp they'll work fine.
.

Yes but it works the other way as well. A sensor placed too close to the exhaust valve cannot use the heater to cool down the sensor. By the way.

Heater current has been known to vary due to wiring and power supply and will affect sensor accuracy.

[Cutlassefi]

"Heater current has been known to vary due to wiring and power supply and will affect sensor accuracy".
Then you most likely have a lower quality O2.

[nickpohlaandp]

I don't know how accurate the o2 sensors I use are, but I take them at face value. With every wideband I've ever set up I use a Bosch UGEO sensor. Maybe it's a little naïve on my part, but I figure when it comes to sensors, Bosch is going to put out a better product that "Chang's sensors". This has me wondering though, has anyone ever tested the actual accuracy of a Bosch UGEO vs. a very well calibrated, known sensing device to see how accurate the Bosch UGEO is?

[user-23911]

WB o2 sensors are waaaaaay over rated and nowhere near as good as everyone tries to make out.


Go get yourself a proper 5 gas emissions box and compare it back to back.









But they have a really nice display and they look pretty.


I still stick to my(calibrated) NBs but have a WB alongside.
I've done a back to back on the dyno about 15 years ago.
WB vs emissions analyser.
Miles out.


But if you want to use it, mount it in one pipe close to the head.
Then use 8 EGT sensors as well.

Afterall, EGTs get used on engine dynos so they must work better?

[stealth]

I remember some engineering saying that fatique can be calculated... Yep, that's reassuring.

VERY TRUE...but to different degrees of accuracy. Unfortunately when the wing breaks off or the bridge collapses is the only time 100% accuracy is achieved........

[Warp Speed]

WB o2 sensors are waaaaaay over rated and nowhere near as good as everyone tries to make out.


Go get yourself a proper 5 gas emissions box and compare it back to back.









But they have a really nice display and they look pretty.


I still stick to my(calibrated) NBs but have a WB alongside.
I've done a back to back on the dyno about 15 years ago.
WB vs emissions analyser.
Miles out.


But if you want to use it, mount it in one pipe close to the head.
Then use 8 EGT sensors as well.

Afterall, EGTs get used on engine dynos so they must work better?

I guess every professional motorsports organization in the world is doing it wrong by relying on the information from a wide band sensor! LOL

[nickpohlaandp]

I remember some engineering saying that fatique can be calculated... Yep, that's reassuring.

Engineers can calculate all they want. The FAA or EASA WILL NOT certify any PSE (primary structural element) until it has been tested to the point of breakage. The automobile industry does something very similar with any new car that is designed and marketed to the public. They have to build "x" amount of the cars AS DELIVERED and crash them in a variety of different manners to prove to the DOT that they meet certain safety standards. When Ford re-released the GT back in 2005 my older brother was one of the chassis engineers on that project (it was a small group of guys that Bill Ford Jr jr jr... whatever wanted hand picked). I think it was around 2003 or 2004 when he took me to the building where they were assembling them and it was the first time I had ever seen one. It was badass! At the end of the very short production line there were two or three completed GT's just loaded with telemetry sensors all over and inside of the car. These were cars that were complete with engines, stereos, etc. I asked him what the deal was with them. He just told me, "These are the ones we're going to crash". That floored me.

Anyway, I got way off base there. If you want to see some interesting video footage, go on YouTube and look up "Boeing 777 wing test". It's about three minutes long and basically a bunch of engineers who had worked long and hard to develop the wing for the then-new Boeing 777. They're standing in a room watching a 777 wing be bent by hydraulic rams until the main spar breaks. There's a calculated point at which they "think" it'll break. Where it actually breaks is what the FAA cares about... and they don't do it just once. It's pretty amazing how far they can bend a wing before it fails.

Years back I worked for Gulfstream corporate jets in Savannah, GA. Whenever they would design a new aircraft (G450, G550, G650) they would take one into the "stress hangar" and bend the living hell out of it in every direction imaginable. That hangar had no windows and only one door in. They didn't want anyone seeing what was going on in there I guess. Anyway, Gulfstream aircraft are certified to 20,000 cycles (basically 20,000 times of pressurizing and depressurizing the cabin). In the stress hangar they would blow up the fuselage to simulate pressurization and then dump the pressure. While doing this, the entire aircraft is mounted to hydraulic rams all over the place that bend it like a circus performer throughout the entire testing phase. I don't know the exact number they pressurize and dump the cabin, but I know it's substantially higher than 20,000. It's a pretty insane thing to witness, and uncomfortable, because you know what's going on, and you see the aircraft, and your brain just tells you, "metal shouldn't bend like that", but it does.

[racear2865]

and some people think all engineers are dumb/ I think about that every time I go up 30+ floors in an elevator.
reed

[David Redszus]

"Heater current has been known to vary due to wiring and power supply and will affect sensor accuracy".
Then you most likely have a lower quality O2.

We perform an enormous amount of data collection of vehicles and engines.
I have only used Bosch oxy sensors for over 40 years.

Like any other sensor, oxygen sensors can become contaminated by fuels and oils, silicone, lead, and they will age with time.

While Bosch sensors are very accurately calibrated on a synthetic exhaust gas, they will perform differently depending on the composition of the exhaust stream gases.

If one is serious, locate the Bosch technical specs for the specific sensor.
And get a five gas analyzer to calibrate an oxy sensor being used on unknown fuels. (All fuels are unknown.)

[David Redszus]

"Heater current has been known to vary due to wiring and power supply and will affect sensor accuracy".
Then you most likely have a lower quality O2.

We perform an enormous amount of data collection of vehicles and engines.
I have only used Bosch oxy sensors for over 40 years.

Like any other sensor, oxygen sensors can become contaminated by fuels and oils, silicone, lead, and they will age with time.

While Bosch sensors are very accurately calibrated on a synthetic exhaust gas, they will perform differently depending on the composition of the exhaust stream gases.

If one is serious, locate the Bosch technical specs for the specific sensor.
And get a five gas analyzer to calibrate an oxy sensor being used on unknown fuels. (All fuels are unknown.)

[Firedome8]

I remember some engineering saying that fatique can be calculated... Yep, that's reassuring.

Engineers can calculate all they want. The FAA or EASA WILL NOT certify any PSE (primary structural element) until it has been tested to the point of breakage. The automobile industry does something very similar with any new car that is designed and marketed to the public. They have to build "x" amount of the cars AS DELIVERED and crash them in a variety of different manners to prove to the DOT that they meet certain safety standards. When Ford re-released the GT back in 2005 my older brother was one of the chassis engineers on that project (it was a small group of guys that Bill Ford Jr jr jr... whatever wanted hand picked). I think it was around 2003 or 2004 when he took me to the building where they were assembling them and it was the first time I had ever seen one. It was badass! At the end of the very short production line there were two or three completed GT's just loaded with telemetry sensors all over and inside of the car. These were cars that were complete with engines, stereos, etc. I asked him what the deal was with them. He just told me, "These are the ones we're going to crash". That floored me.

Anyway, I got way off base there. If you want to see some interesting video footage, go on YouTube and look up "Boeing 777 wing test". It's about three minutes long and basically a bunch of engineers who had worked long and hard to develop the wing for the then-new Boeing 777. They're standing in a room watching a 777 wing be bent by hydraulic rams until the main spar breaks. There's a calculated point at which they "think" it'll break. Where it actually breaks is what the FAA cares about... and they don't do it just once. It's pretty amazing how far they can bend a wing before it fails.

Years back I worked for Gulfstream corporate jets in Savannah, GA. Whenever they would design a new aircraft (G450, G550, G650) they would take one into the "stress hangar" and bend the living hell out of it in every direction imaginable. That hangar had no windows and only one door in. They didn't want anyone seeing what was going on in there I guess. Anyway, Gulfstream aircraft are certified to 20,000 cycles (basically 20,000 times of pressurizing and depressurizing the cabin). In the stress hangar they would blow up the fuselage to simulate pressurization and then dump the pressure. While doing this, the entire aircraft is mounted to hydraulic rams all over the place that bend it like a circus performer throughout the entire testing phase. I don't know the exact number they pressurize and dump the cabin, but I know it's substantially higher than 20,000. It's a pretty insane thing to witness, and uncomfortable, because you know what's going on, and you see the aircraft, and your brain just tells you, "metal shouldn't bend like that", but it does.

A wile back ,,before YouTube,, saw a video of a wing stress test on a 747 "I think" ,,,,they bent it back and forth to and fro until it failed. Cool video not sure if Utube has it.

[adam728]

5 gas is nice, but there are lots of wide bands out there doing a "good enough" job of keeping an eye on things, from the home tuner to OEM's. They all have pro's and cons, and are all susceptible to false results due to fuel, tuning, engine conditions, operator error, etc. Wide bands are cheap and simple to use, and in most cases are plenty good enough.

As for engineers vs real world testing: I don't know any engineer that thinks a quick design/calculation is the end of the story. We setup targets that way, but then test / test / test. Even the simplest of systems seem to turn up many extra variables and a lot more testing. Safety factor depends on application, historical variation, etc. Something may see 2 million cycles in a lifetime, but is tested to 40 million. Think of something as simple as a interior panel spring clip, with a 180 page test report summary.

Then there's always the bean counters that fight for the goal that everything should fail all at once at the products expected end-of-life, like the Blues Brother's car. Once sat in a meeting with a customer that wanted 40% cost taken out of an electronic component, because it was being used in a product with a 1,000 hr life rating, and they were going to utilize it in a new line with a 600 hr life rating. Joked about adding cost and writing code so it would quit functioning at 601 hours (one guy seemed less worried about cost and more focused on "it's built too well and needs to fail sooner").

[MadBill]

5 gas is nice, but there are lots of wide bands out there doing a "good enough" job of keeping an eye on things, from the home tuner to OEM's. They all have pro's and cons, and are all susceptible to false results due to fuel, tuning, engine conditions, operator error, etc. Wide bands are cheap and simple to use, and in most cases are plenty good enough.

As for engineers vs real world testing: I don't know any engineer that thinks a quick design/calculation is the end of the story. We setup targets that way, but then test / test / test. Even the simplest of systems seem to turn up many extra variables and a lot more testing. Safety factor depends on application, historical variation, etc. Something may see 2 million cycles in a lifetime, but is tested to 40 million. Think of something as simple as a interior panel spring clip, with a 180 page test report summary.

Then there's always the bean counters that fight for the goal that everything should fail all at once at the products expected end-of-life, like the Blues Brother's car. Once sat in a meeting with a customer that wanted 40% cost taken out of an electronic component, because it was being used in a product with a 1,000 hr life rating, and they were going to utilize it in a new line with a 600 hr life rating. Joked about adding cost and writing code so it would quit functioning at 601 hours (one guy seemed less worried about cost and more focused on "it's built too well and needs to fail sooner").

Like this? https://rpo.library.utoronto.ca/poems/d ... ical-story

I was reading the fine print on a carton of name-brand tri-light bulbs; they claimed an average life of 1.1 years, used 3 hrs./day. In my experience they last about as long as a soap bubble..

[racear2865]

When you fart in the bath tub is about as short.
reed