| Author | Message | ||
     Stuuk |
Hi All Just starting to do some datalogging using tunerpro on my 1125r (BUEZD) and have a couple of q's if you dont mind. When I compare my history table to my fuel map the rpm scale is the wrong way round on the fuel table Left is high rpm and right is low On the history table its the other way round. Is there anyway to change the rpm headers on the fuel map so its the same as the history table? In the history table drop down box i can choose front accel correction...... or rear accel correction..... it says something after this as well but i cant read it- What does it say? How do i interperate the data in the history table boxes? Should i use these numbers to replace the numbers in the fuel map in the same cells? What view should i use in the history table? There is quite a few Most recent sample Running average History average etc etc Ive only done a short test run and im confused with some of the numbers im seeing eg most recent sample history table entry at tps 7 rpm 400 shows 442.943 this position on my fuelmap shows 72 Can someone help clarify some of this? Cheers Stu | ||
| Stimbrell |
Hi Stu, New to this as well so if someone with more experience says something different take their advice over mine. To change the rpm scale go to, "tools", "preferences", and in the general tab on the right put a tick in the reverse columns box. As far as I can see after Accel correction it says "VS RPM" Yes, it is a case of replacing the cells as you say. On the you tube "how to" it shows using the "history average" The anomalous readings you see are in areas you do not run in, in your example, closed throttle at 400 rpm, the bike will only ever see this during start up, any numbers that are way different to what the fuel map says I guess the data logging has not managed to capture, so I ignore them. One other thing, I only data log the bike when it is at operating temperature so I do not capture warm up tables etc, this is a guess on my part and may not matter, I am just being cautious. Let us know how it goes. Hope this was of some help. Steve. (Message edited by stimbrell on September 23, 2010) | ||
| Stuuk |
Hi Steve yes, thanks for the help I managed to reverse the rpm I'm trying to understand what we are looking at in the history table with a heading like front accel correction vs rpm and why i use this as a direct replacemet for the same cell in the fuel map? I'm assuming these numbers are derived and calculated from the o2 sensor data? Are we aiming for stoichiometric or a little richer? Steve Have you done many data runs Have you seen any improvement using this method? Cheers Stu | ||
| Stuuk |
ps Have you noticed when monitoring the guages in tunerpro that every 5 secs or so the guages flick to maximum? I noticed this happening on opti's youtube video as well. Wont this be producing spurious data in the logging? | ||
| S21125r |
Stuuk, Yes you will find random "blips" in your data log where either the RPM drops out or shoots to the moon. At these same moments you will see some quirky stuff on the load side i.e. TPS 7 etc ---- ignore them. Most recent sample is what the engine required at that moment in time when fuel was metered (like a snap shot). Running average takes the last 10 samples and averages them (big picture). History average takes every sample it has in memory and averages them (ah... several pictures taped together..). So if you took 3 different logs and played them back end to end, it would average the data from all 3 logs. I mostly use running average and verify the result with the history average, but I changed my log def file to save 50 running average samples instead of the default 10. Important things to look for is your sample size and standard deviation. If you only have 2 samples for a cell and it's suggesting that you lean out that cell value, I would be leary about it until you have more samples. The more samples the better. As for standard deviation... Hard to explain but I'll try. If you have a cell value of 50 for example and the standard deviation is 5, then that means the 68% of the time the engine needed a fuel value in the range of 45 - 55. The other 32% of the time it either asked for something below 45 or above 55. So in a nutshell, you want the deviation as low as possible. I'm usually content with a deviation that is less than 5-10% of the cell value. So a cell value of 50 and a deviation of 2.5, or a cell value of 100 with a deviation of 5, etc. From what I understand, the values are based on volumetric efficiency calcs at stoich and that accel, warm up and WOT enrichments are multiplied on top of the cell value. So generally speaking you should be able to data log and plug the numbers in as long as you use good judgment with respect to the sample size and deviation - also doesn't hurt to compare logs from other runs as well for a sanity check. Measure twice, cut once. BTW - when I did some logging, I found that the logs for 2500 - 4000 RPM at cracked throttle (the "surging zone") were all over the board in a couple of cells. Big standard deviations with 50+ samples. Almost as if there were some funky harmonics going on in the intake track/air box at these parameters that was bouncing the value around. | ||
| Stuuk |
Hi Mate Thanks for the info I'll try and digest it slowly. When you said "So if you took 3 different logs and played them back end to end, it would average the data from all 3 logs" How do you load 3 logs at the same time? One more question I guess the goal is to keep modifying the maps until the maps and history tables start looking very similar (after many data runs)? | ||
| S21125r |
Load one log and play it back. The without closing tunerpro, you load the second log and play it back. So on and so forth. History average is also really handy if you have a really big log with a couple of hundred samples. Running average will only calculate the last 10 (or 50 like mine is set up) but history will calculate all the samples. | ||
| Xoptimizedrsx |
there you go glad you guys like it. another way to do it is to move the older logs out and place then in a different folder not linked to tunerpro. then you can pull randomly from it for sample checks. I made the adx as a plain version you can add more details and other functions to it as well. please make sure you set the rpms in the history to match your ecm or it will not be accurate t all. its taking the full electronics into play in the history tables. starting with the maps and ending with the actual fuel being injected. its using the O2,emps and ect to see changes and so forth. the blips of bad data is something you cant change due to data cycles and cable speed. also make sure you always check for newer versions of tunerpro as it changes it gets better and better. the last one really cleaned up the adx stuff for even better logs. its 7.1## number now. plus the buell plug in now has a are you sure you want to burn button (yes or no) so no mistakes are made on burning... you can also run the datalogs in MLV if you like. both ways work. a few of the values are names funny but those are not needed in the mlv. you will change the mlv setup to work. I dont remember exactly what but its just basic settings you change. from there it sees accel and decel data time and ommits that from adjustments. based from the o2 and maps ego and ect. you will need a msq file maker i have though. not posted but I can post it. with inst. ctrl a,c,v is what you use to move the data once done. you will also delete one data range in the log to run in MLV. I have to dig back in the cranium and get the steps for you guys. I really never posted much of that way as I was working on the history way mainly. Mike | ||
| Stuuk |
Managed to do some serious data logging today (dodging the rain) Did 3 20 mile runs and got logs. Due to weather etc. concentrated on the low to mid tps 3000 - 6000rpm regions (where I was getting lots of data. Ran all 3 logs through log history and used historic average. I only changed fuelling on the maps where I was getting a good sample count on the history log. I was surprised how close the maps were to the history log data. At least 25% of the region of the map I was concentrating on was within one unit of the history log and none of the readings were miles from each other I think the biggest difference was 8 but the average was more like 3 or 4. I interperate this as meaning the maps in this region are not too far out and that the collected data is probably good. Well ive adjusted the maps and waiting to see the results. I'm using a small fujitsu touch screen tablet pc strapped to the pillion seat. This is running all the s/w so if necessary I can annalyze and change mapping while out on the bike. Just waiting for some decent weather now so I can log higher rpms and tps's Cheers Stu | ||
| S21125r |
Stu Most of the logging you did is more than likely in the closed loop zone so your adjustments may not gain much. Never hurts though. One thing I found hard was getting a meaningful number of samples @ high rpm. Popo and my neighbors frown upon me ripping up and down the road a dozen time at a buck+. Good luck and let me know how things go for you. I've got about 2/3 of my map logged so I'm still tweaking mine too. | ||
| Stimbrell |
Originally posted by Optimizedrsx, please make sure you set the rpms in the history to match your ecm or it will not be accurate t all. Many thanks for this, did not realise this was even possible, learning all the time. Stu, Only question left to answer, I think, "Have you seen any improvement using this method?" Well I would say a definite yes, especially in the low rpm, low throttle area I have been working on, the improvement in riding in city traffic and at steady cruise is pronounced and well worth the effort. Hopefully more people here will get in to using this software and we will all be able to pool knowledge. (Message edited by stimbrell on September 25, 2010) (Message edited by stimbrell on September 25, 2010) | ||
| Xoptimizedrsx |
Dont thank me for tuner Pro that program is Marks and has been use in the Auto world for a long time (before 2003 actually) before ecmspy,Directlink,FuelSolutions and so on. As far as writing any Of the programs I have not written any of them. They all belong 100% to other people. I did learn how to utilize one for the bikes and extended that data for free. But I did not make the program itself. TunerPro is just a eeporm chip reader thats very accurate and used by auto tuners from many manufacturers. you set the start page and the end page and it gets every byte in between. It reads chips that are not ecm/ecu chips as well. note: all of the programs out as aftermarket programs for the Bikes are not hacks of the buell program but most are... One program is a hack of more than one and I wont point fingers but Its in his story How he hacked it all by his self. Just a FYI on hacking... Many top companies get many things and hack it. Then base development from existing systems or Items. Its all part of peoples jobs behind the scenes. They get a program open it up see how it works, then rewrite it in a different language with a new look doing the same thing as a end result but taking a new path making it different. one for instance is phones, another is computers and chip companies. Its done everyday in the electronic world. Everything is based from something from the past by adding to or modifying something existing. In the real world They(people and Companies) call this R&D. lol |
