When does the ECM switch from closed loop to open loop mode?

[Totenkopf]

Originally Posted by HDMD88

Your tuner is only as good as his 02 sensors accuracy. When I go to Dealers I take a new wide band sensor with me and my first question is ...''when is the last time your sensor was checked for the correct calibration and when is the last time you changed it out''.
You would be surprised that in 80% of the tuning Dealers are working with a sensor that is anywhere from .5 to 2.0 off the correct calibration.....now the 2.0 is the worse I ever saw but I see .5 to .8 quite often. There are so many things effecting a sensor and it's life span so as a rule I check the accuracy of the sensor every 5 tune ups and replace the sensor every 20 to 25 tunes. If you know of someone that is changing the 02 every 6 tunes he knows about the weak link in tuning and is doing the right thing, however in a production shop like Dealers are every 6 tunes will get expensive this is why I check the calibration every 5 tunes.
Bottom line, the answer to your question is simple....YES, the rest of the tuners are fooling them selves.

Doc,
Would it be fair to say in open loop tuning that if you're tuning to a sensed AFR value, sensor accuracy is important but if you are tuning to peak the torque and hp and finding what each motor wants, the AFR line is going to be whatever the offset is from calibration and at the end of the day, doesn't really matter if you found all the power.

I've also seen one dealer that their sniffer was off at least one whole point and they were really interested in my RSR AFR gauge on my bike and put it on their dyno just to compare AFR readings. I know that mine was really close to being right because of recent dyno time and good correlation then. My sensor, albeit just a generic single wire Bosch, was very fresh at that time and not contaminated and showed what they suspected, theirs was off by alot. Even today after 12K miles on that sensor, readings have not varied since it was new.

---------------------------------------------

For ToBeFrank.... regarding the TwinScan II+.......

all I can tell you is that the DTT TwinScan II+ with the WEGO III is a really good supplementary tuning aid and isn't all that hard to use on SERT or PC VE tables. Have used it for pre-determined test regimes on the street and on the dyno and it really helps get the driveability sorted out quickly. Like I mentioned before, the final WOT and other tweaks are done on the dyno but this method really helps reduce total dyno time. For someone that has tuning skills and wants to do it themselves and get most of the efi tuning in order on their engine build with the SERT or PC, if they have a good idea of what a spark curve should be, this is a good way to get AFR under control and keep the motor safe. One of our members, GPO03Fatboy, did a great job with his bike using this method and then cross checking it on a dyno a year or two ago.

all the best!

[HDMD88]

Originally Posted by Totenkopf

Doc,
Would it be fair to say in open loop tuning that if you're tuning to a sensed AFR value, sensor accuracy is important but if you are tuning to peak the torque and hp and finding what each motor wants, the AFR line is going to be whatever the offset is from calibration and at the end of the day, doesn't really matter if you found all the power.

I've also seen one dealer that their sniffer was off at least one whole point and they were really interested in my RSR AFR gauge on my bike and put it on their dyno just to compare AFR readings. I know that mine was really close to being right because of recent dyno time and good correlation then. My sensor, albeit just a generic single wire Bosch, was very fresh at that time and not contaminated and showed what they suspected, theirs was off by alot. Even today after 12K miles on that sensor, readings have not varied since it was new.

---------------------------------------------

For ToBeFrank.... regarding the TwinScan II+.......

all I can tell you is that the DTT TwinScan II+ with the WEGO III is a really good supplementary tuning aid and isn't all that hard to use on SERT or PC VE tables. Have used it for pre-determined test regimes on the street and on the dyno and it really helps get the driveability sorted out quickly. Like I mentioned before, the final WOT and other tweaks are done on the dyno but this method really helps reduce total dyno time. For someone that has tuning skills and wants to do it themselves and get most of the efi tuning in order on their engine build with the SERT or PC, if they have a good idea of what a spark curve should be, this is a good way to get AFR under control and keep the motor safe. One of our members, GPO03Fatboy, did a great job with his bike using this method and then cross checking it on a dyno a year or two ago.

all the best!

Yes Tote you are right on....the AFR numbers mean nothing in the map and if your tuning for a AFR number and you don't know what your sensor is reading your tune is dangerously bad.
If you tune for power you might see a overly rich AFR reading which should tell you your sensor is reading wrong....ie.. if I'm getting the best power from an AFR of 13.5 by experience and the dyno I'm on is getting the best power with a setting of 12.2 I know the sensor is bad.
You brought up a good point and a good example of AFR numbers are meaningless in the AFR Table if you don't know what your sensor is doing.

[ToBeFrank]

Originally Posted by Totenkopf

all I can tell you is that the DTT TwinScan II+ with the WEGO III is a really good supplementary tuning aid and isn't all that hard to use on SERT or PC VE tables. Have used it for pre-determined test regimes on the street and on the dyno and it really helps get the driveability sorted out quickly. Like I mentioned before, the final WOT and other tweaks are done on the dyno but this method really helps reduce total dyno time. For someone that has tuning skills and wants to do it themselves and get most of the efi tuning in order on their engine build with the SERT or PC, if they have a good idea of what a spark curve should be, this is a good way to get AFR under control and keep the motor safe. One of our members, GPO03Fatboy, did a great job with his bike using this method and then cross checking it on a dyno a year or two ago.

I'm not saying it doesn't work. I'm merely saying it doesn't work as good as it could, or in my opinion as good as it should. I can't speak for using it with a dyno, but I do have first hand experience with it street tuning. The link I posted earlier showing the spikey 3D surface plots are not my plots, they are from another guy who was street tuning. However, my surface plots looked just like those when I was using the TSII software. There is no way the plots should be all spikey like that. If you read that thread, you'll see that Steve Cole has the same opinion. I got so fed up with trying to hand smooth them to get rid of the spikes that I wrote my own software to do it the right way. After that, the results with the TSII hardware when street tuning were pretty good.

[strokerjlk]

WOW
This is good stuff thanks guys for all the input!
Now when I ask this understand that it is just like a student asking there teacher to explain.I am not doubting just trying to understand.
Everyone seems to agree that the sensors only last 10,000 miles. mine are at 30000 and seem to be functioning fine. they were only in closed loop for about 10000 of these miles . Are you saying that they wont function correct or totally inoperable?
Are you just talking about there ability to control using CLB settings? I understand that in open loop they are not controlling or (switching). But they are still in the heat of the exhaust. I also understand that at a high CLB they would not last as long. So if in open loop how long is the life as far as collecting correct data? (data logging)
10000 miles is unacceptable as far as I am concerned
This thread is awesome for those of us trying to learn.
THANKS AGAIN!!!!

[BVBOB]

Originally Posted by strokerjlk

WOW
This is good stuff thanks guys for all the input!
Now when I ask this understand that it is just like a student asking there teacher to explain.I am not doubting just trying to understand.
Everyone seems to agree that the sensors only last 10,000 miles. mine are at 30000 and seem to be functioning fine. they were only in closed loop for about 10000 of these miles . Are you saying that they wont function correct or totally inoperable?
Are you just talking about there ability to control using CLB settings? I understand that in open loop they are not controlling or (switching). But they are still in the heat of the exhaust. I also understand that at a high CLB they would not last as long.

10000 miles is unacceptable as far as I am concerned
This thread is awesome for those of us trying to learn.
THANKS AGAIN!!!!

They will not collect data and are totally inoperable in open loop, if you are asking how long they would be functional if they were basically left in while running open loop and then turned back on by switching the afr tables to 14.6 to collect data that would be a tough question to answer as the probability of anyone actually doing that would be rare.

[BVBOB]

Originally Posted by ToBeFrank

I'm not saying it doesn't work. I'm merely saying it doesn't work as good as it could, or in my opinion as good as it should. I can't speak for using it with a dyno, but I do have first hand experience with it street tuning. The link I posted earlier showing the spikey 3D surface plots are not my plots, they are from another guy who was street tuning. However, my surface plots looked just like those when I was using the TSII software. There is no way the plots should be all spikey like that. If you read that thread, you'll see that Steve Cole has the same opinion. I got so fed up with trying to hand smooth them to get rid of the spikes that I wrote my own software to do it the right way. After that, the results with the TSII hardware when street tuning were pretty good.

Hey Frank, I sat down tonight at my computer and opened up a pile of different maps, some canned, some custom, a portion of both showed some pretty spiky areas in the graph of the VE tables, as I understand you are using software to smooth the VE tables out to make the colored graphs show a smooth transition, my question would be how do you know this is what the motor wants, to me, smoothing for the sake of the smoothly colored graph is bascially making an assumption as to what the motor wants for fuel delivery, after a cell by cell tune I just dont see how smoothing for the sake of a nicely colored graph can help the performance of the bike. If it does, I will be the first in line to use this method but tell me "Why" it would work better.

[AZFXSTB]

Originally Posted by BVBOB

Hey Frank, I sat down tonight at my computer and opened up a pile of different maps, some canned, some custom, a portion of both showed some pretty spiky areas in the graph of the VE tables, as I understand you are using software to smooth the VE tables out to make the colored graphs show a smooth transition, my question would be how do you know this is what the motor wants, to me, smoothing for the sake of the smoothly colored graph is bascially making an assumption as to what the motor wants for fuel delivery, after a cell by cell tune I just dont see how smoothing for the sake of a nicely colored graph can help the performance of the bike. If it does, I will be the first in line to use this method but tell me "Why" it would work better.

The software he is using he created himself..as to the spiky VE tables there is no way it should be like that. The VE is a predictable extrapolation from a known value that is correct. Dame that made my head hurt. Now there are factors that will skew the results but with a base plot it is fairly predictable thing with environmental constants.

Now I do not question anyone in this area I don't know enough myself but when it comes to the geek side of this stuff, trust me when I tell you there are VERY few that can compete with TBF.

[HDMD88]

Originally Posted by BVBOB

Hey Frank, I sat down tonight at my computer and opened up a pile of different maps, some canned, some custom, a portion of both showed some pretty spiky areas in the graph of the VE tables, as I understand you are using software to smooth the VE tables out to make the colored graphs show a smooth transition, my question would be how do you know this is what the motor wants, to me, smoothing for the sake of the smoothly colored graph is bascially making an assumption as to what the motor wants for fuel delivery, after a cell by cell tune I just dont see how smoothing for the sake of a nicely colored graph can help the performance of the bike. If it does, I will be the first in line to use this method but tell me "Why" it would work better.

Bob
You are correct in what you are saying, the VE table will not be smooth when you install a cam, pipe, heads ,etc that alter the VE in different areas of the map. It would be nice to have a smooth transition in the 3-d map but reality says cam reversion in the manifold will spike the VE tables alone. Once you start changing the configuration of how a motor breathes and trying to get the AFR synchronized the VE tables are no longer a concise value...the numbers become what yo need to get the fuel your asking for so the 3-d VE table will look like a section of the rocky mountains. If you try to smooth this out for a pretty picture your negating what the engine is calling for. I see many shops try to tune like that from learning from another person that travels the states doing what I do and there is pinging/detonation going on and it's because the VE tables are off which throws the whole map reading off. Leave the mountains alone.

[HerkoBagger]

Originally Posted by HDMD88

John has been working hard on this 02 sensor dilemma, he has spent countless hours documenting figures for me plus he has spent a small fortune in test gasses, LC1 sensors, Bosch wide band sensors, an the latest Dyno equipment to assure a accurate and collective data from every stand point. My hat is off to you for your efforts John.....and.....thanks for the Christmas Cookies, lol.
The best sensors used in todays top performance shops and colleges for EFI map developers cost about $1000.00 and up. These sensors are calibrated at the beginning of a tune for free air and again after lunch....lol...really they are re calibrated for free air most every time they cool down. These kind of sensors are not an option for tuners of our caliber so checking the accuracy of the sensor being used often and replacing the sensor the first sign of it getting lazy is all we can do to assure our tunes will be the best we can give.
Thanks again John for all your research and money you spent in this endeavor.
Doc

Thanks Doc.
Glad it helped and now the other tuners that read this thread may benefit as well.
Christmas cookies...you're quite welcome. Black walnut, it's kind of a love or hate variety lol.
Great to see you posting regularly again.
John

[strokerjlk]

Originally Posted by BVBOB

They will not collect data and are totally inoperable in open loop, if you are asking how long they would be functional if they were basically left in while running open loop and then turned back on by switching the afr tables to 14.6 to collect data that would be a tough question to answer as the probability of anyone actually doing that would be rare.

Thanks Bob
so then I assume the switching (controlling) in closed loop is what wears them out and not so much the heat?

[HDMD88]

Originally Posted by strokerjlk

Thanks Bob
so then I assume the switching (controlling) in closed loop is what wears them out and not so much the heat?

No it's not the switching that kills an 02 sensor, it's carbon, fuels, oil, heat over 1350 degrees..(I should make this clear, at 1650 degrees the Lambda 02 sensor is done, cooked, finished...some racing fuels will kill a sensor in one hours time) Like Bob said there are several things that alter the readings of an 02 sensor...we have to keep on top of them by checking their operation as much as we can...they are a weak link andhave way too muc h control over the fuel delivery for being such a weak link. There are companies working on this problem as we speak. You will see a BIG difference in Harleys 02 sensors next year and you might not see AFR as you know it now....start looking for Lambda unstead of AFR ratio's....you can go on line and read up on the fuel measurment of Lambda.......the 09 ECM all ready has provissions for the Lambda ratios and a new 218 calibrations for them. SURPRIZE!!!!!!

[BVBOB]

Originally Posted by HDMD88

Bob
You are correct in what you are saying, the VE table will not be smooth when you install a cam, pipe, heads ,etc that alter the VE in different areas of the map. It would be nice to have a smooth transition in the 3-d map but reality says cam reversion in the manifold will spike the VE tables alone. Once you start changing the configuration of how a motor breathes and trying to get the AFR synchronized the VE tables are no longer a concise value...the numbers become what yo need to get the fuel your asking for so the 3-d VE table will look like a section of the rocky mountains. If you try to smooth this out for a pretty picture your negating what the engine is calling for. I see many shops try to tune like that from learning from another person that travels the states doing what I do and there is pinging/detonation going on and it's because the VE tables are off which throws the whole map reading off. Leave the mountains alone.

That particular guy is exactly who I was thinking of as I was writing that post, while not exactly the same method(as Franks), his method uses block tuning (large sections of VE adjustments at the same time) which IMHO is making a very large assumption, it bascially takes the fine adjustablity of the tool before you (SERT/SEST/TTSetc.) and makes it rather coarse. I have seen one of his tuning charts and as I recall it took a spreadsheet area that had over 200 potential points of adjustability and lumped it up into 7 or so. There are even more things to consider, such as the area where the ecm changes it reading from one column or row to the next, all can have an effect on what that colored graph ends up looking like. I think Frank has a great idea interpolating the cells but not to the extent that the 3-d graph having a smooth color transition is the only thing that matters, how bout some of you other dyno owners on this site, check out your maps, I'm bettin you see some spiky areas on some of your maps also.

[ToBeFrank]

Originally Posted by BVBOB

Hey Frank, I sat down tonight at my computer and opened up a pile of different maps, some canned, some custom, a portion of both showed some pretty spiky areas in the graph of the VE tables

See below for what I define as spikey.

as I understand you are using software to smooth the VE tables out to make the colored graphs show a smooth transition, my question would be how do you know this is what the motor wants, to me, smoothing for the sake of the smoothly colored graph is bascially making an assumption as to what the motor wants for fuel delivery, after a cell by cell tune I just dont see how smoothing for the sake of a nicely colored graph can help the performance of the bike. If it does, I will be the first in line to use this method but tell me "Why" it would work better.

This is NOT what I'm doing. I am not smoothing the VEs after they are generated. I have completely replaced the Twin Scan VE generator software with my own algorithm. That algorithm includes a smoothing algorithm. There are no assumptions being made. The data logged from the bike tells it what to do.

Originally Posted by HDMD88

You are correct in what you are saying, the VE table will not be smooth when you install a cam, pipe, heads ,etc that alter the VE in different areas of the map.

So do you think this surface plot looks acceptable? This plot is what you get when street tuning with the Twin Scan software. BTW, Steve Cole agrees with me that the answer here should be "no" (supporting link posted earlier in this thread).

Originally Posted by BVBOB

his method uses block tuning (large sections of VE adjustments at the same time) which IMHO is making a very large assumption, it bascially takes the fine adjustablity of the tool before you (SERT/SEST/TTSetc.) and makes it rather coarse.

This is NOT what I'm doing.

I think Frank has a great idea interpolating the cells but not to the extent that the 3-d graph having a smooth color transition is the only thing that matters.

When did I say "having a smooth color transition is the only thing that matters"? My point is if the algorithm to generate the new VEs is correct, you automatically end up with a smooth graph. And for the nit pickers, yes, you'll have some spikes, but in general the map will be pretty smooth.

BTW, interpolating cells is mandatory if you want a correct algorithm. An example: suppose your VE table has a TPS axis with 10 and 15 and a RPM axis with 2000 and 2250. Now suppose you log a data item at 12.5 TPS and 2125 RPM. Which cell does this data item contribute it's data to? The next data item is logged at 12.0 TPS and 2150 RPM. Which cell does it go into?

[AZFXSTB]

I'm just going to take my seat with my can of Shut The Frack up and follow the rest of this becaue I oviously have no Idea what I am talking about......

[claydbal]

Originally Posted by AZFXSTB

I'm just going to take my seat with my can of Shut The Frack up and follow the rest of this becaue I oviously have no Idea what I am talking about......

you can sit by me, that way you wont be the dumbest. i win that trophy! can i fetch the popcorn and coke?