AE Test Runs & Data Q
Thread Starter
|
Senior User
Joined: Oct 2010
Posts: 407
Likes: 2
From: Mobile, AL
AE Test Runs & Data Q
So, I got the AE hooked up and working (darn on/off button 
). Logged a bunch of data and am working through it, so naturally I have a couple of questions-
1) It passed the buzz and CCT tests- good
2) PERDELS were zero on 1, 2, 3, 6, 7. #4 was around .2-.3, so that is pretty low. #5 & #8 logged some 1.5 up to 2.1. Anything to look for there?
My biggest question is related to the way AE logs the data. I'm getting "groups" of columns, one time and one sensor reading. But the times are not sequential across the columns. So, if column a is time and b is rpm, then c is time and d is speed. However, from a to c to e to g the times do not match nor are they sequential. So each sensor isn't being recorded at the same time and I can't seem to find a sequence. It makes analysis pretty difficult. I attached the mess to illustrate ....
From reading other posts, I logged vehicle power voltage, ICP, ICP DC, Engine RPM, Fuel Injector Pulse Width, EBP, Manifold Gauge Pressure, Mass Fuel Desired and Vehicle Speed. I threw in the temps, too. I was thinking of relating one variable to the other to cross check performance.
Any tips or tricks to sorting to help looking at it?
). Logged a bunch of data and am working through it, so naturally I have a couple of questions-1) It passed the buzz and CCT tests- good
2) PERDELS were zero on 1, 2, 3, 6, 7. #4 was around .2-.3, so that is pretty low. #5 & #8 logged some 1.5 up to 2.1. Anything to look for there?
My biggest question is related to the way AE logs the data. I'm getting "groups" of columns, one time and one sensor reading. But the times are not sequential across the columns. So, if column a is time and b is rpm, then c is time and d is speed. However, from a to c to e to g the times do not match nor are they sequential. So each sensor isn't being recorded at the same time and I can't seem to find a sequence. It makes analysis pretty difficult. I attached the mess to illustrate ....
From reading other posts, I logged vehicle power voltage, ICP, ICP DC, Engine RPM, Fuel Injector Pulse Width, EBP, Manifold Gauge Pressure, Mass Fuel Desired and Vehicle Speed. I threw in the temps, too. I was thinking of relating one variable to the other to cross check performance.
Any tips or tricks to sorting to help looking at it?
Cargo Master
Joined: May 2010
Posts: 3,047
Likes: 10
From: Snohomish, WA
Each time a sensor records data it does it at a specific time. Only one record can be done at a time. If you notice the times, it goes through each sensor and starts over again. The times will be separated by a few milliseconds. The more sensors you are recording data on the longer it will take to complete a cycle.
Your PERDELS look good and you are passing the CCT so that proves they are within tolerance.
After I record data, I go back through and remove each column with the time stamps, except the first one. If I need more precise times, I can test items specifically.
Here is a shot of what it looks like if you remove each column for the time stamps.
Your PERDELS look good and you are passing the CCT so that proves they are within tolerance.
After I record data, I go back through and remove each column with the time stamps, except the first one. If I need more precise times, I can test items specifically.
Here is a shot of what it looks like if you remove each column for the time stamps.
Last edited by woodnthings; Jan 6, 2011 at 02:43 PM. Reason: added screenshot
Thread Starter
|
Senior User
Joined: Oct 2010
Posts: 407
Likes: 2
From: Mobile, AL
Thanks, Ken, for the confirmation on the PERDELS and CCT.
Any thoughts on how to look for a boost leak? I'm trying to figure out how the RPM, manifold pressure, EBP relate to see if I can determine what may be going on.
As for the data itself, I'm looking through my data set and can almost put it in order. I think, and it goes with what you're saying, that it samples each sensor in turn. However, the sample order may not coincide with the column order. So, it won't resample the first sensor until all have been sampled for that round which gets hard to follow when I have 30+ columns.
From your experience treating the data as instantaneous sampling (by keeping the initial time and removing sub-times) is good enough to verify proper operation or id a problem area. I think I'll go that route and see how things match up. It seems to sample every 4-6 seconds, kinda slow as I'm used to 4 or 16 Hz sampling from work. I'll cut the fluff and see if that improves the rate.
Thanks again, and nice cradles
. You make some sweet custom pieces.
Any thoughts on how to look for a boost leak? I'm trying to figure out how the RPM, manifold pressure, EBP relate to see if I can determine what may be going on.
As for the data itself, I'm looking through my data set and can almost put it in order. I think, and it goes with what you're saying, that it samples each sensor in turn. However, the sample order may not coincide with the column order. So, it won't resample the first sensor until all have been sampled for that round which gets hard to follow when I have 30+ columns.
From your experience treating the data as instantaneous sampling (by keeping the initial time and removing sub-times) is good enough to verify proper operation or id a problem area. I think I'll go that route and see how things match up. It seems to sample every 4-6 seconds, kinda slow as I'm used to 4 or 16 Hz sampling from work. I'll cut the fluff and see if that improves the rate.
Thanks again, and nice cradles
. You make some sweet custom pieces.
Post Fiend
Joined: Nov 2007
Posts: 9,549
Likes: 14
From: Payson, AZ
Use the record button before you start a test run and play it back at the end of run. I find it easier to record the data and then play it back. When you go to play it back you can pick what sensors or parameters you what to see on a moving graph (2 or 4 parameters at a time), the live data meters (all nine can be seen at a time) or on the live data grid. The only problem with this is you have to disconnect vehicle (use the dropdown menu if you don't want to remove scan tool from the DLC) to run playback. Using this method you can also playback any recorded data anytime, even sitting at your desk. I just use the AE software for graphs because I don't have any other software to do that.
Cargo Master
Joined: May 2010
Posts: 3,047
Likes: 10
From: Snohomish, WA
Thanks! I enjoy building a variety of things. Mostly all custom pieces to fit where someone want's. I built the cradle for a good fishing friend of mine, when his wife was having their first child. It was a pretty cool gift for them. The other one I built for my sons soccer coach who didn't have a kid on the team. I figured he coached those kids for several years and gave up many hours just to pass on the sport to younger kids. I figured his wife needed a little something for giving up all that time when she was having their first child.
The easiest way to check for boost leaks is with an adapter to pressurize the CAC system. Dieselsite makes one like shown below, or if you are crafty and visit the local home improvement store, you can make your own.
The easiest way to check for boost leaks is with an adapter to pressurize the CAC system. Dieselsite makes one like shown below, or if you are crafty and visit the local home improvement store, you can make your own.
Thread
Thread Starter
Forum
Replies
Last Post
DIYMechanic
1994.5 - 1997 7.3L Power Stroke Diesel
175
Jul 9, 2019 09:18 AM
bobbyrjw
1999 - 2003 7.3L Power Stroke Diesel
2
Jul 29, 2011 09:07 PM
