mmm.xdl When it goes lean at a slow cruse there is a surge, and that happens with the timing at distributor set to 20 BTDC as well. It backfired thru the throttle body if I gave it to much throttle on this log
mmm.xdl When it goes lean at a slow cruse there is a surge, and that happens with the timing at distributor set to 20 BTDC as well. It backfired thru the throttle body if I gave it to much throttle on this log
OK that mmm xdl was a lot better, jumping around wise. I would work on getting your block learn numbers to run closer to 128. Lower is the computer pulling fuel, higher is the computer adding fuel. Some have mentioned the wide band does not always agree, but at least with my TBI they were very close. Watch the counts if the data seems out of wack. If its low on counts I would probably discard and hand edit that cell. Its worth a shot to see if you can make improvements in the drivability. I doubt this is all your problems but its a direction.
mmm.jpg
I was looking at the VV.xdl. You are doing a lot better not jumping around. See how all the BLM numbers are slowly counting down? It is adjusting. Spending enough time in each cell to get a good average is important. Put that data into the spread sheet and see what it starts to look like. I think as you work on the table the drivability will improve. When you reburn the chip, disconnect the battery each time to start fresh.
Last edited by donf; 09-18-2019 at 06:00 AM.
What happened between logs VV.Xdl and ass.xdl? are those right after each other with no changes? ass.xdl looks a lot worse blm wise. It must be really late in FL. I will catch up with you tomorrow. We will get it figured out.
Last edited by donf; 09-18-2019 at 06:11 AM.
Going over the VV and ASS files. Maybe they are not as far off as I thought at first. In the early file, there are more populated cells closer to 128 but the sample count on a lot of them is low.
Also, the data recorder is not that advanced. I did not know this before this morning. if you run two log files back to back its just going to add them together. It will not clear data unless between files unless you hit the clear button on all screens or close the program and start fresh. So if you are playing pre recorded data, skipping between files, fast forwarding, etc. Its pretty easy to get data log screen data that is complete jumble mess of all the files and not what it would have been if the screen had just been left open on a test drive and then stopped. When messing with it, there are definitely some crude parts to the freeware. We get what we pay for ha ha.
What I would do is open TunerPro new each time I did a test drive. I would let it warm up to 180 or so then start the log and then at the end of the drive copy and paste all end data to a blank spread sheet as the end average numbers are what I was most interested in anyway. Then I would look at the counts if there were low count cells or cells set in transition, or some hangars that seemed way out of place I would remove those or edit them by hand. Once that is done I would put them into the other spreadsheet for adjusting the fuel table values. Its definitely not an once and done process. You can get a good base though and sneak up on better as time allows.
I was also looking at your wideband data, the min and max numbers for each cell are pretty far apart. Also the narrow band o2, sensor, how old is it? Just curious as after looking at your BLM # and the wide band data there may be be some discrepancy. Others have mentioned this before. With mine I started with a fairly new sensor but with about 30k on it, then wondered the same thing and swapped to a new AC delco. There was some differences in the numbers I got after the swap, but not a huge amount. If the sensor was original, I would have definitely replaced it with a good quality senor before starting the process. I was just wondering if an old narrow band with a lot of miles feeding info into the computer may be part of the problem? Just wondering, it may be nothing.
After the tables are roughed out a little closer you can work on AE and PE. Some here recommend disabling those for the first part, but for myself just being careful with crazy throttle movements and starting low on the table and gradually working into the higher load areas seemed to work fine. My set up was more conservative though, it actually took a second or two to transfer in to PE at WOT. Not crazy long, but I got very good mileage even with the 4.10 gears. Your choice on if you need to lock it out or not for the rough out of the tables.
Last edited by donf; 09-20-2019 at 08:00 PM.
Hey donf, yes I was back to back logging on those files. I have been having and was experiencing the worst of my problems so far and started driving in a way as to get the timing issue to show its face. It's not a glitch, it's the knock sensor. The sensor does not seem to be bad as far as testing and never set a code, but I thing it may be over sensitive so I ordered a new acdelco. I figured this out last night so I removed the knock sensor, Teflon taped the threads, reinstalled it hand tight, and so far today it's running like it should. I hope this is the fix and that I can have fun with tuning instead of fighting with it. I'll log data and post it tonight. Thanks again for the info, your last post cleared up a couple problems I was having.
That doesn’t sound like the sensor. The sensor is likely performing exactly as designed. In addition, a knock sensor code is only set if the internal resistor fails (for 84-95 vehicles) and almost always has nothing to do with the operation of the sensor element itself. If adding Teflon tape to artificially dampen the sensor’s ability to do its job fixed the issue, it means the knock module in the PCM is incorrect for your application. This is what takes the raw reading from the sensor and interprets what is knock and what is normal noise. By defeating the sensor you are just masking the issue.
Personally I would want to go the route of finding a proper knock module, but scalpers on eBay are selling LT4 modules for $300+ and other modules have been largely untested as to what frequency ranges they are effective at in terms of knock detection. So short of creating your own module based on dyno-tested research of the sensor outputs, that’s not an option it would seem.
Personally, if you’re going to defeat the sensor anyway, I would just go into the computer and disable both the knock sensor codes as well as set maximum knock retard to 0, disabling the ESC system. I mean, you’ve already crippled it, so why not?
Sincerely, someone whose loud exhaust causes his PCM to retard timing too.
You were right about a problem with the O2 sensor readings. After some good log's I noticed the BLM's dropping as you said and noticed that when the fans turned on they dropped to 105. So I logged with the AC on (which turns the fans on) and found a problem. I used the pink wire that was left from the EGR delete to power the key on power to the relay for the fans. After disconnecting from the pink wire (that also powers the O2) it stopped and made things work a lot better. Thanks for bringing that to my attention.
I have no diagram for your truck so I could be off on this, but it is not a good practice to power any accessory with the wires from the engine control harness. If you need a keyed power source for the aftermarket wideband or any other accessory its best to go to the accessory side of the fuse box. I am glad you found that, it could have led you in circles.
Posting Permissions
Reply With Quote
Bookmarks