OBDII Data

[RobinPZ72]

Not sure if there's a thread on here and there probably is, so sorry in advance!

The RX8 ecu does give a lot of data and maybe we can have a big of fun!

Pic is very crude and sample rate a bit low, so you can tweak that in the app or software!

[rameshr]

Which app did you use?


Sent from my iPhone using Tapatalk

[13Black]

I literally suggested a thread like this to someone earlier but more as a resource to use as a baseline for others to compare against to see if anything flags up as being erroneous/problematic with theirs... whether it would be useful or not who knows, but wouldn't be hard to do.

[RobinPZ72]

Yeah it gives a little insight on what's going on and a bit of good and obviously the odd LOL banter!

The app is just the OBDlink and the OBLink LX dongle! Then export data to spreadsheet!!

[Nerdstrike]

We could conceivably generate a confidence interval in which "normal" cars reside. Not easy though, as it would need quite a few healthy and un-mapped cars to establish normality.

I've still got the scatter plots from collecting MAF vs RPM data, but not the R workbook (real men don't use Excel...), and it takes a lot of filtering to get the right throttle regime and to have enough data in it to draw some conclusions. We did establish that one car with an AEM intake was reading much higher (erroneously or not), and that you could observe the intake stages stepping in on the MAF trace. For example, three RX8s, one is supercharged: https://drive.google.com/open?id=0B7oOk ... TJYOHZxcHM

What it did show is how variable everything is and that generic ELM bluetooth loggers have a shockingly poor sample rate. In practice, MAF, AFR, STFT, LTFT and catalyst temp are all calculated from just two sensors, so there are limits to what can be derived.

Your estimated horsepower looks excessive, and the AFR looks a shade rich!

[warpc0il]

Using the Torque Pro app you can switch to a higher (than default) sample rate, if the dongle can take it.

[13Black]

You'd be surprised what my dongle can take.


"rEaL mEn DoNt UsE eXcEl"
-Also doesn't label units or data

I kid, I really wish I had more time to learn R
D:

[RobinPZ72]

AF actual is just fine I'm sure!

It's mpg is not bad either and as long as you know it's maybe not super spot on its what the ecu is doing on the road! Short runs 20mpg!

Bit of fun and maybe not everyones cup of tea, but hey impossible to keep everyone smiling !

[13Black]

AFR seems to be about what the actually run at.
They always miss their target by a mile, up at the spicy end without a remap.

[RobinPZ72]

Yeah it's bog standard apart from a K&N panel filter and proper oil!!

[Nerdstrike]

You'd be surprised what my dongle can take.

"rEaL mEn DoNt UsE eXcEl"
-Also doesn't label units or data

I kid, I really wish I had more time to learn R
D:

My dongle failed dismally when I pushed it harder

I hate R as a language, and for a few years it was super-hard to Google, but its plotting libraries are very capable. My time again, I'd try to do it with a Python.

[RobinPZ72]

I've 2 pythons!

Anyhoo it's an idea and could be helpful to some who are looking for advice or purely just to cast others eyes over data!

Also I'm not sure if this thread is in the right place and maybe should be in technical?

Robin

[rameshr]

As we have recent dyno results, we can compare to find the error margin.

I use inCarDoc but I don’t think it can give me the same data


Sent from my iPhone using Tapatalk

[13Black]

The problem with OBDII horsepower calcs is that they introduce way more uncontrollable variables (over and above the error dynos have, which are about as 'lab-controlled' as we can hope for), so the figures are meaningless at best.

[warpc0il]

A roll-on acceleration test, given correct input for the vehicle weight (including fuel etc) logged using ODB2 data for time and road speed, is a reasonable indication if a modification on a specific car has increased power, or not.

However, there are too many variables to be much use for comparison between cars and even on the same car it should be repeated on the same piece of road, in the same direction, in the same weather (especially wind) conditions.

It should, at least, be more accurate than the butt dyno.

[RobinPZ72]

Would be interesting to see how much of a difference if people that have dyno time! The biggest problem is the Speed limits!!!!!!!!

Fun track day comparison!

[warpc0il]

A roll-on acceleration test is probably more accurate in 2nd gear than 4th.

Peak revs on 2nd is only 66mph in a 192 and 68mpg in a 231

[RobinPZ72]

Yes I know as 3rd at the beep about cough cough 90

[warpc0il]

For roll-on tests the RX8 has one great advantage over less sophisticated vehicles...

When we used to do such tests back-in-the-day, there could be a wide variation in the results depending how lead-footed you were.
For example, if you just floored the accelerator then the carb would open fully and the intake charge would stall, so the engine ran rough and acceleration was poor. The trick was to keep the throttle pedal just ahead of the actual speed, which meant that some of us could achieve a much cleaner acceleration, in exactly the same car, compared to other drivers, making the results significantly more variable.

The RX8 has a fly-by-wire throttle, which means you can bury the pedal into the carpet and the PCM will just take that as a signal to provide maximum acceleration, regardless of who's driving. Making the results significantly more repeatable.

[350matt]

that and not much torque so its rare to be traction limited

[TheManc]

Not sure if there's a thread on here and there probably is, so sorry in advance!

The RX8 ecu does give a lot of data and maybe we can have a big of fun!

Pic is very crude and sample rate a bit low, so you can tweak that in the app or software!

send me the raw data excel file via email carl300367@aol.com

also what gear was this in ? iirc dyno runs were done in 5th gear or was it 4th ?

piccy is my car on a base line map in 4th with the data grabbed from a cobb port

[RobinPZ72]

I'll do another one and see if I can get a better sample rate then send it!

Robin


Carl that's it emailed to ya!

[TheManc]

you need to enable the following options

[RobinPZ72]

I'm going be hook the laptop up as I think it going to flip out the phone! I'll get back to you probably Monday as my kids have booked me for the weekend!

[AndyBrad]

robin, see if you can get it going on your phone so fools like me an upload more data

so long as its idiot proof and we get similar stuff.