Air fuel ratio 02 sensors for data

General engine tech -- Drag Racing to Circle Track

Moderator: Team

hoodeng
Member
Member
Posts: 82
Joined: Sat Mar 04, 2017 6:53 pm
Location: South Australia

Re: Air fuel ratio 02 sensors for data

Post by hoodeng » Wed Sep 06, 2017 8:06 pm

ALKYBURNER'S primary interest was to investigate a way of minimizing guess work out of a multiple input tune situation [ tangible inputs being ,fuel,NOS and ignition°] and yes there are lot of other variable factors that effect and alter these inputs [ RA, Hg, RATemp, AD ],and has rightfully identified exhaust data logging as the most practical and achievable to start with ,this for him is a great way to start , but as i said previously this will be for him as it was for us many years ago probably just the beginning of a quest for more knowledge.
Our ideal EGT and O2 were primarily derived from the experience and development of the engine manufacturer and supplier who supported us throughout the time we campaigned that particular bike,plus also through our own experience of what this particular bike liked and responded to , the target numbers for those two observations at the time we ran the bike were influenced partly by a class fuel regulation, so to give the raw numbers of our targets here i think would start a discussion of which numbers are correct , the numbers that are correct are the ones that make your race vehicle run optimal.
Earlier TPS was identified as an input and in many circumstance is , in our instance TP was only ever 100%, if it wasn't the rider would have a
'please explain?'

Sunoco maximal ,13.?:1 1250°-?. ponder the missing numbers!!

Cheers.

joe 90
Guru
Guru
Posts: 1972
Joined: Tue May 27, 2014 4:02 am
Location: The land of the long white cloud

Re: Air fuel ratio 02 sensors for data

Post by joe 90 » Thu Sep 07, 2017 5:24 am

Was having a look at the specs of a WB setup that we've got in the workshop.
It's a tech edge, the 3B1 which would have been top of the line about 10 years ago.
It's got datalogging etc as well as 4 thermocouple inputs.
It'll also read down to 23 or 24 to 1 which is handy for diesels.

http://www.wbo2.com/3b1/default.htm

smeg
Pro
Pro
Posts: 419
Joined: Fri Feb 09, 2007 12:43 am
Location: melbourne, Australia

Re: Air fuel ratio 02 sensors for data

Post by smeg » Fri Sep 08, 2017 6:33 am

I have the ability to log 8 exhaust temps on my dyno but after being led down the garden path a few too many times they are no longer used.
I now use the wide band, but that has confused me a lot as well so it is better to say, use them as a tuning tool no more no less, there is no magic wand, your results may vary.

David Redszus
Guru
Guru
Posts: 5642
Joined: Tue Nov 27, 2007 9:27 am
Location: Chicago
Contact:

Re: Air fuel ratio 02 sensors for data

Post by David Redszus » Fri Sep 08, 2017 11:09 am

smeg wrote:I have the ability to log 8 exhaust temps on my dyno but after being led down the garden path a few too many times they are no longer used.
I now use the wide band, but that has confused me a lot as well so it is better to say, use them as a tuning tool no more no less, there is no magic wand, your results may vary.
What, may I ask, was confusing about the use of wide band lambda sensors?
Did you replace the 8 EGTs with 8 wide bands with outputs fed into your data logger?

andyf
Guru
Guru
Posts: 1224
Joined: Tue Oct 05, 2004 12:55 pm
Location: Oregon
Contact:

Re: Air fuel ratio 02 sensors for data

Post by andyf » Fri Sep 08, 2017 11:51 am

smeg wrote:I have the ability to log 8 exhaust temps on my dyno but after being led down the garden path a few too many times they are no longer used.
I now use the wide band, but that has confused me a lot as well so it is better to say, use them as a tuning tool no more no less, there is no magic wand, your results may vary.
I know what you are saying. We used to only have EGT before wideband became available. It helped but didn't tell the full story since timing makes a bigger difference to EGT than AFR. So if you only have EGT you have to be careful to get the timing close before you chase AFR. Once you add a wideband into the equation things can get better but only if you have good wideband information.

We've used a bunch of different wideband setups over the years and some produce garbage. Not exactly sure why that is but my guess is it either has to do with sample rate, or it has to do with more simple things such as electrical integrity within the unit. If the grounding is not done properly the voltage levels will float around and so will the AFR readings.

I got a bit of an education when we started testing EFI engines that had their own wideband setup. The Holley systems with the high dollar NTK wideband will provide a dead solid AFR reading while the widebands that the dyno company sold us are wiggling all over the map. It is fairly common for us to see more than one point of difference between the NTK and the dyno widebands. The NTK might read 13.5 while the dyno wideband varies between 14.5 or 12.5.

Good data is very valuable but bad data is misleading and dangerous. Most dyno shops do not have full time engineering resources on site to help them sort this stuff out. NASCAR teams can hire PhD candidates to sit in the lab and run regression tests all day long but the average engine builder barely understands how to install a wideband much less figure out how to verify the data.

Not sure I have any good advice on the subject other than to buy high quality equipment and see if you can work with someone to get your dyno cell calibrated. Even that is probably beyond the ability of most small shops since the good equipment is expensive and there aren't a lot of people capable of providing calibration services and/or system training.
Andy F.
AR Engineering

n2xlr8n
Pro
Pro
Posts: 367
Joined: Tue Jun 28, 2005 9:01 pm
Location: Bama

Re: Air fuel ratio 02 sensors for data

Post by n2xlr8n » Fri Sep 08, 2017 11:59 am

I've used EGT sensors (for combustion timing verification) in conjunction with WBO2 (AFR) in all of my personal EFI engines.

I would never rely solely on EGT for fuel tuning. Wrong tool, imo.
He who is in me is greater than he who is in the world.

RaptorLou
New Member
New Member
Posts: 13
Joined: Mon Jun 26, 2017 2:47 pm

Re: Air fuel ratio 02 sensors for data

Post by RaptorLou » Fri Sep 08, 2017 12:15 pm

There are lots of valuable concerns posted. When all is said and done, they are all just tools to help tune. Dyno - tool, Flow bench - tool they are there to help make the best decisions on paths. I recently did testing on my flat 4 with a dual system with 2 wide bands in each tract. I run a single Weber carb and found that one throat is rich/leaner than the other. Did lots of swapping of jets, emulsion tubes, a/c, even flipped carb end for end. Now did I make a gain, YES I have learned and documented that one side is slightly different and can jet accordingly. Typically I run 4 into one and the wide band basically averages the 4, but now I can have the 4 better balanced from the get. Interesting as well is I put the carb on the flow bench and flowed it and measured jet signal at points and did not see a super variance, kinda weird, but now I know. I actually now color code my carb side for data and logging.
Long and short, they are tools, how we use the tools is the key.
Bolt on's are a gamble
If it isn't broke push it harder :D

David Redszus
Guru
Guru
Posts: 5642
Joined: Tue Nov 27, 2007 9:27 am
Location: Chicago
Contact:

Re: Air fuel ratio 02 sensors for data

Post by David Redszus » Fri Sep 08, 2017 12:22 pm

It is fairly common for us to see more than one point of difference between the NTK and the dyno widebands. The NTK might read 13.5 while the dyno wideband varies between 14.5 or 12.5.
If a lambda sensor shows a level, steady reading, it is probably wrong.

Virtually every engine, even in steady state, produces a range of outputs, often ranging over 3 AF ratios (14.0 - 11.0).
The steady readings are often smoothed while the jumpy reading reflect actual engine operation.

It is essential that logged readings are sampled at an high sample rate and not smoothed in any way.

Now the task involves making sense out of what appears to be data hash. But it is real data.
Data is smoothed or low sample rates are used for display purposes; true data would be too jittery to read.
That is why proper data logging is essential.

I would recommend the use of an AiM or MoTec data system for accurate results.

andyf
Guru
Guru
Posts: 1224
Joined: Tue Oct 05, 2004 12:55 pm
Location: Oregon
Contact:

Re: Air fuel ratio 02 sensors for data

Post by andyf » Fri Sep 08, 2017 12:33 pm

David, your reply only makes sense to people who work in labs and have access to experts and high dollar equipment. Put yourself in the shoes of an average engine builder who buys an expensive dyno. The guy doesn't understand data analysis, he doesn't have an engineering degree, he doesn't remember statistics from high school, he is just trying to get an engine out the door so he can pay the rent.

If you pretend to be an engine builder and you call up SuperFlow or Stuska or someone like that and ask them what the sampling rate is on their widebands you will most likely get a blank stare. They don't know what this stuff is. They just tell you to plug in the equipment and use it. They don't have any way to pull raw data dump it into Excel and run statistical analysis. Most engine builders struggle to use basic functions in Excel much less know how to import large data files and manipulate them.

I'm sure you're trying to be helpful but I think you work in a bubble......
Andy F.
AR Engineering

David Redszus
Guru
Guru
Posts: 5642
Joined: Tue Nov 27, 2007 9:27 am
Location: Chicago
Contact:

Re: Air fuel ratio 02 sensors for data

Post by David Redszus » Fri Sep 08, 2017 12:46 pm

andyf wrote:David, your reply only makes sense to people who work in labs and have access to experts and high dollar equipment. Put yourself in the shoes of an average engine builder who buys an expensive dyno. The guy doesn't understand data analysis, he doesn't have an engineering degree, he doesn't remember statistics from high school, he is just trying to get an engine out the door so he can pay the rent.

If you pretend to be an engine builder and you call up SuperFlow or Stuska or someone like that and ask them what the sampling rate is on their widebands you will most likely get a blank stare. They don't know what this stuff is. They just tell you to plug in the equipment and use it. They don't have any way to pull raw data dump it into Excel and run statistical analysis. Most engine builders struggle to use basic functions in Excel much less know how to import large data files and manipulate them.

I'm sure you're trying to be helpful but I think you work in a bubble......
Are you suggesting the perpetuation of ignorance?

Please do not throw bubbles around; they are fragile and likely to burst.

The bubble I work in involves teaching hundreds of racers how to use ECUs and data systems. Very few have difficulty learning how to use a new and better tool to improve understanding and performance.

To your point, there are some who can afford to write a check but are unwilling to pay the intelectual cost of learning.

Belgian1979
Guru
Guru
Posts: 3428
Joined: Sat Mar 26, 2011 11:34 am
Location: Belgium - Koersel

Re: Air fuel ratio 02 sensors for data

Post by Belgian1979 » Fri Sep 08, 2017 1:26 pm

A WB is a usefull tool. I have 8 of them on the engine and I can attest to the fact that they vary from one cylinder to the other and from one cycle to the other. Each engine is different in this respect.

The only thing confusing about them is a misfire (either rich or lean misfire) that gets indicated as lean.

I have been contemplating EGT's for possible trimming individual cylinder timing, but no idea if that would actually work in practice.

To go further, I'm also a firm believer of in cylinder pressure measurements for tuning purposes, but these are way to expensive for any regular tuner and even then some.

andyf
Guru
Guru
Posts: 1224
Joined: Tue Oct 05, 2004 12:55 pm
Location: Oregon
Contact:

Re: Air fuel ratio 02 sensors for data

Post by andyf » Fri Sep 08, 2017 1:43 pm

David,

I'm saying that your answers about data sampling are not useful to the average person. The average guy who buys a dyno package from Superflow or Stuska or someone like that can't get to that level of analysis. If you have a dyno system then what you get is a digital readout and a printout. Nobody can explain where the printout numbers come from. Pick up the phone yourself and call Superflow and ask them what the AFR numbers on the dyno printout mean. Ask them what the sample rate is on the wideband and which number out of the data cloud gets printed on the report. They don't know. I've sat on the phone and asked those questions, they don't know. Someone wrote a data capture program for them and they don't know how it works, or at least the people I've talked didn't know and didn't know how to find out.

The average engine builder understands analog electronics and can get reliable data from an EGT since it is very robust. Digital electronics are touchy and are susceptible to things like noise and ground loops and a bunch of stuff that engine builders aren't trained on. After mucking around with this stuff for the past 20 years I understand why engine builders are confused about wide bands. They just don't work that well in the field. They seem to work pretty well for the OEM's but once you take them out of the hands of experts I think the data integrity falls off pretty quick. If you really spend that much time teaching people this stuff you should understand that better.
Andy F.
AR Engineering

Warp Speed
Guru
Guru
Posts: 2060
Joined: Wed Jun 25, 2008 1:46 pm
Location: NC

Re: Air fuel ratio 02 sensors for data

Post by Warp Speed » Fri Sep 08, 2017 2:40 pm

Do we want to remain "average"?
If someone doesn't understand the functions, or are not willing to learn the aspects, they have no business buying a dyno, let alone charging people for dyno time! LOL

Warp Speed
Guru
Guru
Posts: 2060
Joined: Wed Jun 25, 2008 1:46 pm
Location: NC

Re: Air fuel ratio 02 sensors for data

Post by Warp Speed » Fri Sep 08, 2017 2:42 pm

And again, what are EGTs good for?
Certainly not much relating to this thread! LOL

David Redszus
Guru
Guru
Posts: 5642
Joined: Tue Nov 27, 2007 9:27 am
Location: Chicago
Contact:

Re: Air fuel ratio 02 sensors for data

Post by David Redszus » Fri Sep 08, 2017 2:53 pm

I'm saying that your answers about data sampling are not useful to the average person. The average guy who buys a dyno package from Superflow or Stuska or someone like that can't get to that level of analysis.
The average person does not and should not build engines much less buy a dyno. If you do not understand how to use a tool, find someone who does.
If you have a dyno system then what you get is a digital readout and a printout. Nobody can explain where the printout numbers come from.
If your dyno supplier cannot explain the readouts and how they were obtained, you are working with the wrong supplier.
Pick up the phone yourself and call Superflow and ask them what the AFR numbers on the dyno printout mean. Ask them what the sample rate is on the wideband and which number out of the data cloud gets printed on the report. They don't know.
I've have engaged in very technical conversations with Harold Bettes for over 10 years. He is quite knowledgeable and willing to share his knowledge with customers and engine builders. But he does not suffer fools gladly; don't waste his time.
I've sat on the phone and asked those questions, they don't know. Someone wrote a data capture program for them and they don't know how it works, or at least the people I've talked didn't know and didn't know how to find out.
Obviously you have spoken to the wrong people.
The average engine builder understands analog electronics and can get reliable data from an EGT since it is very robust.
Are you kidding? EGTs are very unreliable; they are slow to respond and do not provide accuracy at all.
Have you ever checked thermocouple cold junction voltage? Where would you go to obtain proper technical support for a thermocouple?
Digital electronics are touchy and are susceptible to things like noise and ground loops and a bunch of stuff that engine builders aren't trained on. After mucking around with this stuff for the past 20 years I understand why engine builders are confused about wide bands.
Don't muck around; you're wasting time. Get some training.
They just don't work that well in the field. They seem to work pretty well for the OEM's but once you take them out of the hands of experts I think the data integrity falls off pretty quick.
At the end of this month, the Indianapolis Motor Speedway will be hosting the SCCA National championships, You'll see 600 amateur racers, most using some form of data logging. Many will be using wide band oxygen sensors with great success. A walk through the paddock should convince any skeptic of the validity of wide band sensors.
If you really spend that much time teaching people this stuff you should understand that better.
If you have a problem understanding the instrument, why not simply ask this forum for help. There are a large number of folks who could provide you with any information you might seek. In fact, a search of previous post on this forum would produce tons of useful information on this subject.

Many community colleges offer courses on the use of automotive electronic diagnostics. Go back to school.

Post Reply