engine monitor log analysis

[TheTurtle]

what software do you use to look at your log data?  I have an Insight G3.  It dumps to csv.  I uploaded a few logs to savvy analysis and it looks nice and stuff but I dont see a good way to compare flights or look for trends.  it will automatcally warn you when a sensor was over an alarm level.  But for instance my oil pressure seems low.  looking at previous flights it seems normal.  I would think the software should see the trend of lowering pressures and tell me.

I see there are a couple paid programs out there.  I also saw an excel spreadsheet with picot tables (although now I cant find it again).

So what do you use?  And what do you look for?

[carusoam]

JPI supplies an app/program for it's users.  Really easy to scan through multiple flights of data...

see if your instrument manufacturer does the same for yours...

Best regards,

-a-

[TheTurtle]

Just now, carusoam said:

JPI supplies an app/program for it's users.  Really easy to scan through multiple flights of data...

see if your instrument manufacturer does the same for yours...

Best regards,

-a-

nope they link to savvy, some 1980's desktop app and an excel spreadsheet. 

Im thinking of throwing something together in Access.  Or maybe finally figure out how to use splunk.

[kortopates]

Allow me to ask if you are already familiar with how to analyze your data? I. e. How you would use it to assess your mixture or the health of your ignition system, or your induction system or even the health of your valves? I ask because the really big ticket items can't be assessed by merely watching trends. If not I can point you to some stuff that will help. But for starters normal cruise flight data isn't all the interesting.


Sent from my iPhone using Tapatalk

[TheTurtle]

Just now, kortopates said:

Allow me to ask if you are already familiar with how to analyze your data? I. e. How you would use it to assess your mixture or the health of your ignition system, or your induction system or even the health of your valves? I ask because the really big ticket items can't be assessed by merely watching trends. If not I can point you to some stuff that will help. But for starters normal cruise flight data isn't all the interesting.


Sent from my iPhone using Tapatalk

I have no real idea what Im looking for as far as trends, or anything else for that matter. But the device does log data, and I like nice graphs, and computers are smart so there must be something that analizes log data and tells me x y and z.. 

What can you point me to?  Im always looking to learn more about my plane.

[Oscar Avalle]

I am using savvy analysis and their pro option. You send the data from one flight and they analyze if for you. They are really great and follow up on you. Highly recommended. Otherwise, I must say that I really don't know what to look for in the data...You know... I am just an MBA. 

[gsxrpilot]

There's a lot of information in the data. Either sign up to Savvy Pro or go take the Advanced Pilot Seminar at GAMI when it's offered. Savvy will do it for you or APS will teach you how to do it yourself. 

[kortopates]

2 hours ago, TheTurtle said:

I have no real idea what Im looking for as far as trends, or anything else for that matter. But the device does log data, and I like nice graphs, and computers are smart so there must be something that analizes log data and tells me x y and z.. 

What can you point me to?  Im always looking to learn more about my plane.

What I am trying to suggest is different perspective that rather than looking at trends or peaks of individual values or channels like CHT, that you need to look at the data in terms of how different parameters behave relative to something else. Because so much is interrelated, we have to recognize very few single sensor values are important independently; especially with respect to their absolute values. An example is EGTs, we don't really care what their actual values are - they aren't even a real measurable physical phenomenon, like Cylinder temperature. Instead EGT is a measure of the exhaust gases for only a very small period of every 2 crankshaft revolutions when the exhaust valve has briefly opened. They aren't anywhere near the temperature of actual combustion. But what makes them important is their temperatures relative to peak. The next thing to realize which goes against the value of just tracking trends, is that because everything is so interrelated we really find very little diagnostic capability by just monitoring cruise data statistics. In order to really learn about how our engine is doing we need to fly some specific kinds of regimens that allow us to look at different things in isolation. Some examples of this are mixture and ignition. We may see things in cruise that make us suspicious of one of the other but to apply the science so to speak we can tell much more when we collect data for a specific flight profile that allow us to isolate exactly what we want too see. So next to actually see how we put this to practice I suggest you read this primer at Savvy on in-flight-diagnostics. It will step you through the big ones. (Of course I am largely simplifying, there are some exceptions to the above, but that's the big picture.)

After that I suggest you read any of the many find articles available to you on the internet for free. This includes John Deakin's Pelican Perch articles on Avwebs, Mike Busch's article on Avweb, and in the EAA monthly magazine and all of webinar's archived on EAA's site, and lastly but not least the Advanced Pilot Seminar site. The latter site has an excellent online and in-person class you can take that Paul above also recommended. But I would encourage you to read all about mixture, how it affects the combustion timing and especially from the perspective of ICP (or Internal Cylinder Pressure). By doing so, you'll also learn the relationships between EGT, CHT. ICP, HP and BSFC - just google "Red Box" if the latter is new too you and you'll find some good articles with the graphs to show these together. With that knowledge I think you'll want to concentrate more on actually periodically capturing data for what Savvy refers too as the Savvy test Profile. Savvy didn't invent it, but they put the simple instructions together and that's the kind of data IMO you want to collect and monitor for changes in lieu of crunching cruise data.

Hope that helps your perspective on how to use the data and as a disclaimer one of my jobs is that I work for Mike B at Savvy. 

 

Edited November 9, 2016 by kortopates