|03-04-2014 03:22 PM|
|ZXthrizzle05||I'm going to sticky this. My gage doesn't match the livelink with the equation. I'm not a number cruncher or good with equations so I can't really help myself this way. But should be usefull to others|
|05-11-2011 09:38 PM|
Solution: Trouble getting AEM AFR gauge to match Livelink
I was having an issue trying to get my AFR measured in SCT Livelink via the analog output to match the AFR gauge itself... Tom can't tune your car unless they do, so it's an important thing to resolve quickly if you want to get your car running right.
This pertains to the AEM Wideband UEGO wideband O2 sensor and gauge.
When I first asked Tom about it he said some people have this issue, so I thought I'd write this up for everyone.
If you have this issue:
DO NOT go messing with the analog configuration equations unless you try this first- I was trying to tweak the equations and it's just impossible to do it by guessing numbers and going for a drive to see if the computer and gauge match at idle, cruise, and WOT. I was using my multimeter to record voltages and AFRs and trying to plot them in Excel and make a linear regression equation to fit my gauge's own quirky voltage output, but I thought my gauge was defective and who wants to deal with that? So I got a better idea...
Troubleshooting it properly:
1) Make sure LiveLink and the SCT XCAL3 firmware are both up-to-date.
2) Your gauge should be getting power from a 12V (use a digital multimeter to confirm the voltage) line that is "switched" with the ignition. I used a power from my interior lights and used a lightbulb screwdriver tester to verify it gets power in the same pattern your radio does (ignition on, or with ignition off before opening the door to get out of the car after having the ignition on). To troubleshoot the issue, I unplugged this from the overhead lamps and from my radar detector, which also gets power from the same line.
3) The ground for your gauge must be "clean" according to the AEM instructions. I had wired it to the ground for those interior lamps, but today I rewired it down to screw in where the shifter bolts to the transmission-- great spot to ground things. Easy to thread a wire from where your gas pedal is into the shifter console as well.
4) At this point, unplug the sensor (6-pin connector) from the gauge and leave the 4-pin power connector in. Turn on the ignition, and the gauge should read 14.7 according to AEM. Mine is reading 14.8 but I think that this is still allowable.
5) Using a digital multimeter set to "9V"... tape one probe to the WHITE line coming from the gauge (this is the analog output), and the other probe to your cigarette lighter (the outside ring of it is ground). So you'll get numbers in voltages, so AFR = Volts x 2 + 10. This is your conversion equation to check if the volts in the output are agreeing with the gauge display. For example, if your voltage is 2.33 the gauge should read 14.7 (rounded up from 14.66). If they match, go ahead and plug in your sensor again to the gauge and now it gets fun.
6) I video taped the gauge side-by-side with the multimeter (I have a cell phone mount that was able to hold the meter, which mounts on the vent) while idling, then driving.
I downloaded the video to my computer and played it frame by frame and using an excel spreadsheet I took down the time, gauge AFR, voltage on the meter, and a calculated AFR with the equation above. this took some time, but in short time you will realize that the multimeter's voltage lags behind the AFR gauge somewhere between 0.2 and 0.5 seconds or so... my video player doesn't go more specific than seconds so I can't tell you the exact time.
So the output is working properly. Now is there a problem with your firewire, the xcal, or perhaps the livelink program? to check this...
7) I started a data log with livelink as I would for tom (make sure your analog 1 or analog 2, whichever you're using, is configured to change voltage to AFR with the equation above or by selecting "AEM" under the option for the analog input configuration), but the most important things to watch are TPS, rpm, and AFR... the first 2 are to get your bearings on what you were doing at that particular AFR. I took a video of this as well: I took a few seconds of video of the computer data log to sync up the timestamps of my video and the log... then videotaped the AFR gauge so I could compare the numbers later, on Excel like with the multimeter.
Your first job is to make sure you get idle, cruise, normal acceleration, and some good pedal-to-the-floor time to see a wide range of AFR's on the gauge and computer for comparison... a safe time for the WOT is in 3rd gear going 30mph... you don't need to be accelerating, just getting that AFR down.
Next, when you compare the video and the log, you have to determine a) how far behind the log the video is. For example, at exactly 5 sec of video my data log read 7.229 seconds, so my video began 2.229 seconds later than the data log. b) how long it takes the AFR output to reach the computer and be reflected on the data log. You'll notice if you give the car some gas it takes a fraction of a second for the computer to show your RPM's go up, etc... so just like with the multimeter, it takes a little time to show up after your gauge shows a given AFR. It's tricky but using the idle section you can find a peak in AFR and then go to your videotape and pause it at where the AFR shows a high point.
For me, it looks like the computer gets the AFR reading about 0.5-0.6 seconds later, but it looks like it's getting accurate readings. At a given spot in your log, the AFR might sometimes be off a little, 0.1-0.2 at most, but keep in mind the computer only records an AFR about every 0.15 seconds... so don't worry as long as they're usually consistent.
If you ran into a problem at any one of these steps, you should recheck that the gauge is wired properly.
If you still can't solve it, your gauge my be defective. AEM told me "if you unplug the sensor and the gauge reads 14.7, the gauge is fine, and you should call SCT".
This takes time and patience. The stuff I just recounted to you all took me about 2 hours once I figured out how the heck I was going to try to solve the problem (which took me about 2 weeks).
Hope it helps someone...