Driveability help posts and replies (check engine lt and/or rough running)

[Autolabs]

This is for all the driveability help requests

I’m frustrated with the auto parts and service industry. Why must so many people guess at what is wrong with their car and spend their hard earned money buying parts that didn’t fix the vehicle? My take is that they come in contact with influential people that don’t know enough to even realize what they don’t know. Maybe it is the part store counter guy, bogus forum threads, or a friend that works on cars, either way it doesn’t matter. I have seen many, many people experiencing hard times throw their time and last bit of money out the door from ill informed advice. I was a technician for 8 years, a part store manager for 4, and now a repair shop owner for 3 years, and yes I did have to scan codes for customers while managing the part stores. After retrieving codes for them and then walking back into the store, this is how the conversation sometimes went:

Me:
“Just so you know, this is not a diagnosis, a code only gives you a direction to start testing to find out what is setting it.”

Customer: (while looking me right in the eyes)
“So what do I need to buy?”

Me:
“A proper diagnosis”

Customer: (getting frustrated)
“I thought that is what you guys did here for free, that’s why I drove all the way here!”

Me:
“I’m sorry Mr./Mrs. customer, but my best advice is to hold off buying anything right now. It would make me feel terrible if you purchased anything on a guess and it didn’t fix it, because once installed, I can’t return it for you”.


You know, sometimes they didn’t even say thanks in a sincere tone on their way out the door, but that’s cool.
And how many times have you read the original post that goes something like this, “I need help please! My vehicle is idling rough, stalling, hesitating, misfiring, backfiring, surging, etc. I have already replaced plugs, coils, all four O2 sensors, cleaned the MAF, replaced the MAF, replaced the TPS, replaced the coolant sensor and changed the oil.” (OK, maybe that’s a little far fetched for the norm, but the theme is still there.)

So let me sum this up and give some advice to those who love to give theirs here. If you can help these people by giving them things to try that are free, or almost free then that is great! You are then contributing to the greater good of actually helping people with cost saving tips and adding value to the knowledge base of this forum. But ask yourself this before replying to any driveability help request,“If I’m wrong, would I be willing to hand over my money to the one who lost theirs, following my advice?”

I love helping people help themselves as well. The 1st time I heard you could get a little wireless OBD adapter for less than 20 bucks, and download a mobile app for free that can read codes and view live data, I got pretty excited. I thought, “Now I can actually help the average DIY with their ‘data’ instead of just their codes and human perceptions.” For 20 bucks, there is no excuse not to have access to your vehicles live data, even if you don’t understand it. Codes are just your vehicle’s sticky notes it left for you. Live data is your vehicle’s speaking language, and it will tell a more detailed story. In OBD generic mode 1, the data doesn’t lie, unless of course it’s a VW diesel. Some manufacturer specific communications are allowed to lie by substituting PID values, but that’s a different story for the ones who will be typically seeing it.

I want to help the DIY mechanic use their live data to at least steer them in a logical (based on facts) direction. I am working on a project that automatically compares each PID from 2 different data logs under similar operating conditions. It searches all the data frames from both files and approximately matches user selected inputs, and graphs the outputs in a paired bar graph for each PID. So ideally, the 1st data log will be a known good running vehicle, and the second will be the vehicle in question. If anyone wants to email me their data logs in a csv file with a description of the vehicle’s condition, I would be very appreciative. Troy@pidfusion.com please help me give some power back to the DIY mechanic.

 

[decipha]

engines are too dynamic to compare datalog values

best bet would be to make sure the datalogging values are within range for the given condition

 

[Autolabs]

Do you mind elaborating on what you mean by too dynamic. I need some good info and facts to help develop or abort this project. I appreciate your wisdom and time.

 

[decipha]

the chances of having two datalogs that are close enough to be compared line by line is almost impossible

here is all the details you will ever need to know on the ecu, read through all of these, its important that you 100% understand all of them otherwise you'll be lost

http://www.efidynotuning.com

all the write ups and reference

 

[Autolabs]

Thanks for the link!

 

[Autolabs]

I agree that it is impossible to have any line from 2 datalogs match up exactly. KOEO maybe. Not all lines we would be even be a partial match. But let's say a stable operating condition of hot idle? tps 0-0% rpm 750 - 810 coolant 195-205, vehicle speed 0-0mph. To find the frames within each log that contain these inputs is definately not a problem.Perhaps that data logs are already tagged by density altitude, and water vapor which can be easily calculated from gps and local weather API's.

Let’s say someone is an absolute beginner with scanning and reading live data and doesn't even know what PIDS to log? Their goal is not to become a custom ECU tuner, but to get that p0300 fixed by themselves in a timely matter. What if they just logged all supported PIDS and selected the conditions like hot idle, and the program is preset to at least partially match and join certain PIDs and values from each log and compare the rest. If they posted the compared logs here, are there people here with a strong automotive background, but unfamiliar with typical PID values for that model, that would be more inclined to offer their help?

A more advanced version would allow the user to slice and dice the logs anyway they want. A User can choose what PIDS to match and then have the program find the best matches for lets say 10% increments throughout the min and max values. Then take the joined frames and create a movie to play, pause, ff, rewind. Do you think it might spark up some conversation and insight from more knowledgeable people without any data logging experience, because the data presented makes sense to them?

Any and all criticism or suggestions appreciated.

 

[2000StreetRod]

worth of datalogs

I use datalogs to evaluate engine performance after modifications and for troubleshooting problems. It is much easier since I have a baseline to compare to. It is much more difficult when there is no standard for comparison. If owners would make a datalog of normal driving and performance driving when their vehicle is operating normally then they would have something to compare to when they have a problem.

Some vehicles have wideband O2 sensors but many don't. That makes it difficult to diagnose problems when the PCM is in open loop and not using the narrowband O2 sensors.

Some vehicles have fuel rail pressure sensors but many don't. That makes it difficult to diagnose fuel pump problems.

The MAF sensor transfer function varies significantly for vehicles. Displacement, compression ratio, valve size, camshaft lift and timing, boost, intake configuration, exhaust configuration, etc. all affect how much air flows thru the MAF sensor.

Ignition advance is dependent on engine load, engine coolant temperature, intake air temperature, engine rpm, knock sensor output, and the calibration constants in the tune making comparisons between different tunes difficult.

Engine coolant temperature is dependent upon thermostat setting, water pump flow, coolant temperature at engine start, time since engine start, effectiveness of cooling system, heat generation of engine.

Most vehicles have an oil pressure switch to detect extremely low or non-existent oil pressure but no sensor of actual oil pressure.

There are numerous more examples when there is inadequate instrumentation of engine parameters needed to diagnose a problem.

It would be difficult to establish useful ranges for pids applicable to a wide array of different vehicles. However, I have found that software tools to compare a single vehicle's pids at different times is quite useful. Most datalogs are a collection of comma separated variables and many people use spreadsheets to compare different logs.

 

[Autolabs]

I agree that it would be useful for an owner to record a good running baseline for comparison for when/if they do have problems. Highly unlikely that would happen since most people and their vehicles tend to be more reactive than proactive. Since now their vehicle has a rough idle, and they didn't record a good running baseline, couldn't somebody with another stock vehicle, same year, make, model, powertrain compare their PIDs to theirs? What would be so different between the 2 that you can't find the frames of hot idle and compare them. I use the software Shopstream by Snap-on all the time to analyze logs. I like it. I keep saving more and more vehicle logs to it for down the road for comparison purposes. I also work on all makes and models so it comes in handy to not have to remember what the norms are through out all operating conditions.

Is this an impossible task? To quickly show a newb what values they see with their new app are normal? What other ways can it be quickly done?

 

[decipha]

no its not an impossible task and by having a few simple selections buttons for the specific vehicle it would be very simple

however, having after the fact datalogs is futile, it would be BEST and most ideal to have the program take live streaming PIDs in realtime from the ecu, that way you can instruct the operator to perform specific tasks and get the feedback you need for the data to be most accurate

 

[2000StreetRod]

diagnostic displays

Frankly, I wish the new vehicles would provide some diagnostic displays instead of devoting so much effort to climate control, navigation and communication displays. Our 2014 Grand Cherokee has an 8 speed automatic transmission and I'm thankful for a digital readout of the current gear on the instrument cluster. Vehicle manufacturers could easily provide access to pids via the installed touch screen.

I use an SCT X3 PowerFlash and a laptop to datalog preselected pids. I use SCT's LiveLink to plot the data. I think plots give a better idea of trends but the actual values are also important for analysis. Many people prefer using a spreadsheet because it allows plotting the same pid from two different datalogs on a single page. However, since time is usually the x-axis for pids I don't know how useful dual plot comparisons are for other than acceleration related plots.

Sharing of datalogs for similar vehicles is already being done on some user forums but it could be greatly expanded with the availability of user friendly inexpensive datalogging devices. I've read that the youngest driving generation is much less interested in how vehicles work and how to fix them than older generations. I assume those who are less able to pay others to fix their vehicles will always be a market for diagnostic devices.

 

[Flash]

Frankly, I wish the new vehicles would provide some diagnostic displays instead of devoting so much effort to climate control, navigation and communication displays. Our 2014 Grand Cherokee has an 8 speed automatic transmission and I'm thankful for a digital readout of the current gear on the instrument cluster. Vehicle manufacturers could easily provide access to pids via the installed touch screen.

.

I just installed VAGCOM on my laptop to use with my Touareg.
I have a big learning curve ahead of me.

 

[Flash]

This thread should be a sticky or maybe start a new thread that's designed to be a sticky.

 

[1998Exp]

Hmm, they could, but probably won't. The basic premise of all automakers in the last couple decades has been to isolate the user from almost anything under the hood. When was the last time you saw an actual oil pressure gauge in a US made car or light truck? Or an actual temperature gauge? Even transmission fluid dipsticks disappeared around 10 years ago. Some will claim that this trend is intended to make more money for the dealers ($100 to diagnose a loose gap cap, anyone?), but I think that it's meant to prevent the average uninformed Joe from doing silly things in response to information that he does not understand. How many drivers even heard of PIDs? Compare that with the number of drivers who are totally comfortable with pairing their smart phone with the car's entertainment system or talking to their GPS, and you'll see why the average new car is essentially a bunch of consumer electronics on wheels. I will leave the question of what impact all this has on traffic accidents for some other time...

Frankly, I wish the new vehicles would provide some diagnostic displays instead of devoting so much effort to climate control, navigation and communication displays. Our 2014 Grand Cherokee has an 8 speed automatic transmission and I'm thankful for a digital readout of the current gear on the instrument cluster. Vehicle manufacturers could easily provide access to pids via the installed touch screen.

 

[2000StreetRod]

original request

. . . If anyone wants to email me their data logs in a csv file with a description of the vehicle’s condition, I would be very appreciative. Troy@pidfusion.com please help me give some power back to the DIY mechanic.

Getting back to the original request. Do you want datalogs for stock vehicles with make, model, year, engine size/type and transmission type to build a data base. Or, do you just want datalogs to test the program you want to develop?

 

[Autolabs]

Thanks Streetrod! I need logs for both reasons actually, to add to a database for R&D. I know beggars can't be choosy, but if the logs could be tagged with altitude, ambient temp, relative humidity, and airspeed in knots, that would be awesome! Ok, maybe a airspeed indicator mounted to the front bumper is a little far fetched and useless below 45mph, but any and all logs with descriptions are very appreciated. Again, that is troy@pidfusion.com

Decipha, thank you and I appreciate and I expect the criticism. And yes, real time data comparison is ideal and being used in connected cars, but mainly for safety features, dealer service, and probably insurance companies. The live data that we see on the scan tool is not real time though. Mode 01 generic is slow, real slow, and what I propose would only be helpful during steady state operating conditions because of the slow PID request sampling rate. Compared to what the CAN is capable of, even the manufacturer enhanced live data is slow, but a lot faster than generic mode 01. You mentioned "having after the fact data logs is futile". I'll admit that I had to look up what futile meant after reading that a couple of times. I'll just say this, comparing real time live data between 2 vehicles will have to wait until the fully autonomous vehicle.

Thanks again, and I appreciate all the criticism from everybody. Let the public beating begin! It helps me to identify all the holes in my idea.