O2 Sensor Placement?

[Sumner]

.....................The chemical balance of something like iso-octane at a chemically correct (stoichiometric) burn looks like this:
C8H18 + 12.5(O2 + 3.76N2)     =   8CO2 + 9H2O + 47N2  Note that on the right side there is a gob of water that is produced. Sure it slobbers more on cold fire up. The heat from this stuff burning is what we try to turn into push on the piston(s) and eventually it becomes usable power at the output shaft. ...............

Hey thanks and I guess there is more to it than the water that condenses in the exhaust after all.  I was even able to finally take the left side of your equation and get the right side  ,  why couldn't I ever figure that out in Chemistry class  .  I hated that subject, looks like you didn't ,

Sum

[willieworld]

hay guys i dont know much about electrical stuff so i bought a air fuel guage that has a bank of lights green for lean and red for rich with orange and yellow in the middle--the o2 senser is a 1 wire so the exhaust has to heat it before it works and that works for me as i use like a temp.guage and know when the engine is warmed up --the o2 senser is  wide band and will show from 10 to 1 to 17 to 1  the o2 senser is mounted on the rear pipe about 6 in from the head on the outside bend and works great  thanks willie buchta

[Sumner]

hay guys i dont know much about electrical stuff so i bought a air fuel guage that has a bank of lights green for lean and red for rich with orange and yellow in the middle--the o2 senser is a 1 wire so the exhaust has to heat it before it works and that works for me as i use like a temp.guage and know when the engine is warmed up --the o2 senser is  wide band and will show from 10 to 1 to 17 to 1  the o2 senser is mounted on the rear pipe about 6 in from the head on the outside bend and works great  thanks willie buchta

Willie are you sure that it is a true wide band?? What is the brand??

I'm not aware of any widebands that work with one wire.  I have a narrow band that says it works from about 10 to 17 with a bunch of lights, so it implies that it is a wide band, but tries to use a narrow band sensor to give a wide band looking output.  I run it in the truck all the time and have run the LM-1 wide band at the same time.  The cheaper one is very accurate at 14.7, but when you get very far from that it is not all that accurate even thought it will tell you that you are running richer than 14.7 or leaner than 14.7.  You think it is accurate because of the lights and the scale, but it isn't.  It is good to tune a motor for the highway, but not accurate enough to be sure where you are at in the 10-13 range.  Put another bung in and talk Scott into putting his LM-1 on your bike at El Mirage and compare that to yours.  I'd loan you mine, but I don't plan on being at El Mirage until November,

Sum

[isiahstites]

I spoke with a a very trust worthy friend this evening and asked about the placement of the sensor. He said within 12" of the head. He also has a pretty detailed write-up on the O2 sensors and there placement and is going to see if he can find it for me to view.

Scott

[narider]

Todd that looks really good to me and as long as the logs look good that is all we care about. 

Not the clearest looking on here, but very clear in logworks and valuable  in many ways as you can see from just the few simple traces I have here a while back. Hasn't necessarily made me any faster(probably could of gone faster without it by now actually), but it's done me well on multiple levels of longevity which means I will eventually reach my goal... and even after a half decade of testing on the same motor I can still stick it right back in my street bike and keep on riding(well, unless I decide to ream her out "just one more time" this year and it decides enough is enough that is - lol).

Looks to me that you are plenty far from the end of the pipe.

Oh yea, no problem there. Deb run's on her bike (well past the collector) and I have used mine on literally hundreds of bikes now both on and off the dyno (even checked on of my pitbikes and my generator once).
The thing about WB sensors like these are their reaction to not only fresh air, but also to air pressure. The fresh air reaction is as one would think with a crack in a pipe or reversion(fresh air scavenge from the pipe end) causing lean readings.. fair enough and simple to check. You can use welding gloves to cover the end of the pipe to see if you're getting reversion(especially once you realize how the pressure effects the readings and can cause an increase or a decrease in afr readings), and you can use the gloves or even rags over the pipe in possible crack or leaking areas to see the decrease in afr if they exist(are sucking air).

Positive pressure on the sensor itself will give false readings as well (and although they are still relative to being lean or rich most of the time, they still won't be accurate). What will happen is that the increase of pressure on the sensor will make a lean condition appear much leaner, and a rich condition appear much richer then it actually is. This is with exception to something like a big amount of reversion being blocked by covering the end of the pipe like above(since that will not be increasing the lean reading from the fresh air, but actually decreasing the afr by blocking that fresh air), as opposed to an anti-reversion trap like in your picture(the scavenge pipe) should it cause a pressure increase on the end of the sensor... as this A-R trap is something that will show the afr enhancements faults as I described if not fully free flowing.
I have an external setup like that on a stand(made by innovate), and to make sure I don't get any pressure buildup I've connected a cheap air-vac(like to clear an a/c manifold) to help keep a negative pressure on the sensor all the time.

Here's the A-R trap



Here's the air-vac


I connect them together like this and run the air-vac form my compressor

That motor looks really nice, now if I could only hear it. 

YOU MEAN IT HAS TO RUN TOO!!!!! 

Ok, It's just a warm up run, but here ya go (turn up your sound) 
http://www.youtube.com/watch?v=NMCd0SXC9Po



Have a good nite Sum,
Todd

[narider]

Thanks for the pictures and your input. Curious as to what logging system you are using? Also, any reason you went with the back pipe instead of the front? I know the back tends to run a bit leaner, is that the reason?
Scott

My pleasure Scott. We have a couple of Innovate's LM1's , a LMA2(mobile) and a LMA3(stationary), with Logworks2 software.
I periodically sync the front and rear cylinders at the shop via single fire ignition adjustment, plug gaps/heat ranges and with adjustable baffles in the pipes with both afr's, cht's and egt's (which helps assure the lack of intake leaks as well). Then I only care about logging one cylinder and periodcially comparing plugs when at the track. I've found you can tune either cylinder to run hotter or colder then the other one, be it at idle or at full rpm.
Todd

[Sumner]

Have a good nite Sum, Todd

And a good nite it was.  That made the cost of high speed internet worth it right there.  I listened about a dozen times with the volume up.  You guys on dial-up make sure you take the time to download this one. The bike doesn't just sound good, but looks great!!  I added it to my land speed racing links page (  http://purplesagetradingpost.com/sumner/bvillecar/bvillelinks.htm ).

I'll have to admit when I was into bikes in the 60's-70's I wasn't a big Harley fan and most of the road bikes still don't do much for me, but your bike, Scott's, Willie's, Bob's twin engine "Bull Dog" and others sure are neat and there isn't anything that sounds as cool.

I'm sure missing the fact that I wasn't able to make it back there this year.  Now when I come I for sure have to put an extra day on the trip to stop by your shop if you have the time.

Thanks for the added input on the wide-band.  I like the way you have the remote wide-band setup.  Pretty neat.  Say hi to Deb and thank her for putting the video up if she did,

Sum

[narider]

hay guys i dont know much about electrical stuff so i bought a air fuel guage that has a bank of lights green for lean and red for rich with orange and yellow in the middle--the o2 senser is a 1 wire so the exhaust has to heat it before it works and that works for me as i use like a temp.guage and know when the engine is warmed up --the o2 senser is  wide band and will show from 10 to 1 to 17 to 1  the o2 senser is mounted on the rear pipe about 6 in from the head on the outside bend and works great  thanks willie buchta

I wouldn't trust that much more then idling or in a no-load less then 10% tps situation where you just want to tune for good gas mileage(right close to stoich). That's a narrow band sensor(no matter what the ebay ad said, jk ). Seriously, it's good for what it was made for, but not for what it's advertised for.
Todd

[narider]

Willie are you sure that it is a true wide band?? What is the brand??
I'm not aware of any widebands that work with one wire.  I have a narrow band that says it works from about 10 to 17 with a bunch of lights, so it implies that it is a wide band, but tries to use a narrow band sensor to give a wide band looking output.  I run it in the truck all the time and have run the LM-1 wide band at the same time.  The cheaper one is very accurate at 14.7, but when you get very far from that it is not all that accurate even thought it will tell you that you are running richer than 14.7 or leaner than 14.7.  You think it is accurate because of the lights and the scale, but it isn't.  It is good to tune a motor for the highway, but not accurate enough to be sure where you are at in the 10-13 range.  Put another bung in and talk Scott into putting his LM-1 on your bike at El Mirage and compare that to yours.  I'd loan you mine, but I don't plan on being at El Mirage until November,
Sum

Bingo!
Don't be discouraged with it though Willie, do like Sum suggests and compare it's accuracy with a decent wideband so you know where it's value is at, and then use it for it's intended and manufactured purpose.
Todd

[Sumner]

The really big value to these is the data logging like Todd has shown in his post.  Sure they are good to look at, but the real value is to have one like the LM-1 or other ones that can data log data logging the throttle position, rpm and boost if you have any along with an EGT if possible.  Now you can make real tuning decisions.  It is surprising how much non-WOT time there is at b'ville and it is good to tune for that also.  I feel confident that using ours has saved a motor and that way more than paid for itself right there.

In addition to the data logging we run a second EGT that is right in front of the driver and can be seen during the run.  If it gets much over 1500 in our case the run is aborted and we look at the data and go from there. 

c ya and Todd you'd better hit the sack it's late back there,

Sum

[narider]

And a good nite it was.  That made the cost of high speed internet worth it right there.  I listened about a dozen times with the volume up.  You guys on dial-up make sure you take the time to download this one. The bike doesn't just sound good, but looks great!!  I added it to my land speed racing links page (  http://purplesagetradingpost.com/sumner/bvillecar/bvillelinks.htm ).

Thanks.. I'm honored to have my old junk take up space on your pages.

I'll have to admit when I was into bikes in the 60's-70's I wasn't a big Harley fan and most of the road bikes still don't do much for me, but your bike, Scott's, Willie's, Bob's twin engine "Bull Dog" and others sure are neat and there isn't anything that sounds as cool.

Yeah, from Evil's 750's and Ironheads to the top fuelers at the drags and what I imagined Leo, Warner and Cook's bikes sounded like on the salt.... nothing was(or will) ever be as cool in my book!

I'm sure missing the fact that I wasn't able to make it back there this year.  Now when I come I for sure have to put an extra day on the trip to stop by your shop if you have the time.

As I mentioned when we talked last, the shop and our home are both open to you anytime and without question. Ya know, there's a solid handful of you wrong coasters over this last year that have really ruined the bad image I had of ya'll for all so long now.  

Thanks for the added input on the wide-band.  I like the way you have the remote wide-band setup.  Pretty neat.  Say hi to Deb and thank her for putting the video up if she did,
Sum

I just got home and woke her up to tell her you said 'thanks and sweet dreams'... she said she'll deal with you tomorrow - lol.
Sure took the heat off of me for getting in late, thanks  

G'nite once again,
Todd

[ddahlgren]

I have read this with some interest and have some comment to offer. All single to three wire sensors are basically digital in operation with a slight slope on the rich side of stoic.. They are universally unreliable due to a minimum of plus or minus 10% due to temperature changes of only a couple of hundred degrees. There is one 4 wire sensor I am aware of that can do some reasonable wide band but reguires sophisticated electronics to poll the sensor occasionally to determine the resistance and from there infer the EGT and hence the relative error. You are not going to find that in a couple of hundred dollar readout. The next thing is lose the concept of air fuel ratio as it is meaningless and ties you to the particular chemistry of the fuel. Use a lambda number that refers to excess oxygen or lack of it compared to stoic, typically 14.6 a/f ratio, most readouts are calibrated with this number to infer an a/f ratio. Any quality meter will read in lambda numbers so if you are shooting for 13.5/1 then a lambda of around 0.92 is the real target. When you tune on the dyno keep track of the oxygen content of the fuel you are using and specific gravity. For a change in specific gravity do a proportional change. 5% heavier fuel dial out 5% globally. If more oxygen add a bit of fuel though that gets a bit sticky as to the exact amount. Heat sinks are generaly never needed unless you are on the engine side of a turbo with big back pressure and the pressure error is greater than the temperature error.
As far as the light loads go dial the fuel out uptil you get a stumble or a slight backfire in the intake add it back in to get the best running and it picks the throttle up easy. Do this at several load points at the same rpm and do a linear interp between them and you are done for that part.
Dave

[Harold Bettes]

Howdy Eric and All,

I am pretty easy to confuse, so I will make a couple of comments and ask a couple of questions and hope for a reply that will enrich us all on this subject material. The strangeness of the chemistry of the fuel and the combustion phenomena is what most of the engine power game is all about.

A) The use of lambda reference is I agree somewhat easier to deal with but as you well know the base of Lamda = 1 = 14.6:1 is somewhat nefarious. "Gasoline" is what the long chain hydrocarbons that are represented by the Hs and Cs compounds are called and there are hundreds. To plan on using a base of Lambda = 1 = 14.6:1 means that the Carbon content is somewhere about 86% Carbon. That is based on the calculation of A/F ratio = 32.63 - .209 x (% Carbon) That is one reason why the gasoline fuel supplier is hopefully very open with the details of the fuel that he supplies.

B) Are you implying that the Carbon content percentage is the only reason that the SG varies? So if you measure a little more dense fuel, you automatically change the fuel mixture?

C) Oxygen content of the fuel? You mean oxygenates that are additives? To the best of my knowledge gasolines have only Carbon and Hydrogen and the % of Carbon must be found by chemical analysis, not by SG.

D) Is the tuning technique that you described related to finding RBT and LBT when testing on a dyno?

Not ragging on you Dave, as I said I am really easy to confuse and am looking for some clarification in all this. That's what gearheads are about isn't it?

Thanks in advance.
Regards to All,
HB2

[ddahlgren]

Here goes.. My thoughts YMMV...
A. Gasoline is a chemical soup of dozens if not hundreds of ingredients. All with varying amounts of carbon and all with different chain lengths and unless you work in the petrochem biz for someone like Mobile and know the blend of the day you really don't know exactly what you are pumping or pouring into the tank. You can find the structure in rough terms as given for an example in a text but that is just that typical but not specific. If we agree on that then we have to agree you really don't know the chemical reaction in great detail but a generality of what it is not the specific one for the blend we have right now and subject to change. Using lambda we know percentage wise how far rich or lean we are from a chemically correct oxidation of the fuel. Most engines of a given family give best performance at a range of lambda readings that are similar even though some of the details of the fuel have changed and an easier at least for me number to work with. At the end of the day it is I think very good to know the chemistry of the oxidizing your fuel, gasoline whatever, I think as well it is of little value other than the understanding in principle of what you are trying to accomplish. I have two reasons for this viewpoint. First you really don't know exactly what the blend is at the atomic level. Even if you did you don't know the atoms of oxygen you added exactly. In that case you would end up with a chemical formula that has exactly this amount of carbon and hydrogen and 'some' oxygen... If by some chance you did know exactly the number of the oxygen atoms you still don't know the exact conditions of combustion on a particular cycle and they are not all the same and that is not opinion on my part,volumes have been written on the subject of cycle by cycle variation in combustion. That is what the art and science of combustion chamber and piston design is all about, eliminate the variations. The best a tuner can hope for is to hit the happy spot where most of them are beneficial to producing power.

B.Fuel density can change for a number of reasons and the why it changes is a bit less important than what to do about it. Temperature alone can do this along with with blending as well. Engines require pounds of fuel per hour. The metering devices we use only can measure volume of fuel, carb jets fuel injectors and the like all flow fuel by volume not weight so density does matter and does change the tune-up.

C. Yes oxygenates are what I was referring to as opposed to the oxygen content of the fuel. In saying that and referring back to A. we hopefully still agree 'gasoline' is a petro soup there can be and most likely will be additives that contain oxygen in their structure added to the fuel. Ethanol MTBE and a host of others come to mind. If you want to do this on a chemical reaction level and account for all the atoms you can not ignore these as they do not pass through the reaction with no effect on it. Saying they don't make a difference would be like ignoring the chemical differences between a classical gasoline with no oxygenates and methanol. Empirically I do know that if a particular engine ran best at a reading of 0.92 on a fuel with no oxygenates I am probably going to find best power at around 0.90 with an oxygenated fuel. Possibly chemically someone else might add some light to this subject but personally I just accept it for what it is and move on to the next engine.

D. What I was attempting to express is to not get caught up in the details of what is going on with a light load at 2000 rpm and 1/4 throttle. Sure it all matters cruising down the road on a 4 door sedan. Hopefully that uses a wildly different camshaft exhaust system and no doubt almost everything else. We only visit that spot briefly in 1'st gear. At low rpm light loads and a long duration cam with typical amounts of overlap a considerable amount of air goes right out the exhaust as well as the results of oxidizing the fuel. Your lambda sensor gave you the results of this by reading lean. I have seen readings over lambda 1 at idle or a light load and the engine is quite happy, adding fuel just made it lazy and unresponsive just like any engine running too rich. In my mind under those conditions you really don't care what the number is but you do care how the engine is running and the device you are using to measure the amount of oxygen in the exhaust is of little value. Add to that cycle by cycle variations at low speed especially and you are wasting your time in the end worried about a lambda reading off idle rather than what the engine is really doing.
Dave

[Harold Bettes]

Dave,

Thanks for such an in-depth clarification of your thoughts. Good for the brain cells.

Your blending of the empirical as well as some theoretical was well done and I appreciate your efforts and am sure that many others do so as well.

Regards to All,
HB2