How do you account for the delay and how do you set it up? Does that mean that the afr read vs the dc at that same time will correspond with each other?whittlebeast wrote:I just looked at the data coming of my motor and the AFR is delayed about .25 sec. I happen to have fairly long tube headers.
Reading MLV Data Logs
Moderators: jsmcortina, muythaibxr
-
- Helpful MS/Extra'er
- Posts: 37
- Joined: Thu Feb 23, 2017 3:33 pm
- Location: 808 HI \[]]]/ ¤shaka¤
Re: Reading MLV Data Logs
Efi Tuning Noob, so please go easy on me.
AE92 GZE'd 85 AE86 3dr(blue "cancer" pill)
ATS Racing 2.1 stroker GT3071R 92 MR2t t-top (red pill) DIYPNP w/5sgte in the works
"Nobody listens to Turtle"
AE92 GZE'd 85 AE86 3dr(blue "cancer" pill)
ATS Racing 2.1 stroker GT3071R 92 MR2t t-top (red pill) DIYPNP w/5sgte in the works
"Nobody listens to Turtle"
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
Keep in mind that most of the stuff I tune are fairly hyper motors used for racing only. Taking the laptop for a ride is generally not possible. I do use VEAL on the dyno but that is used to get the motor close enough to get it to the track. I also use the dyno to find what AFR this motor wants to make the best power on the wideband installed in the car. I also find the best WOT timing on the dyno.
I use use three views to verify the tune
http://www.nbs-stl.com/tuning/Tuned%20T ... 20Fuel.png
I normally look at lots of data and create filters to toss out all data where the [Engine] >1 I am looking for no clear pattern of lean or rich. I tend to hand tune. Clear rich areas get adjusted leaner by about 5% Clear lean areas get adjusted richer by about 5% I like to have SD primary table 1 with a multiply by AN table 2.
The way I verify the lambda delay is I hold the motor at light accel and then stab the throttle. I look in the logs for how long it took for the AFR to notice the sudden change in airflow and then recover.
http://www.nbs-stl.com/CRX/AE%20Testing.png
On this install the AFR will be ignored for about .15 sec.
Andy
I use use three views to verify the tune
http://www.nbs-stl.com/tuning/Tuned%20T ... 20Fuel.png
I normally look at lots of data and create filters to toss out all data where the [Engine] >1 I am looking for no clear pattern of lean or rich. I tend to hand tune. Clear rich areas get adjusted leaner by about 5% Clear lean areas get adjusted richer by about 5% I like to have SD primary table 1 with a multiply by AN table 2.
The way I verify the lambda delay is I hold the motor at light accel and then stab the throttle. I look in the logs for how long it took for the AFR to notice the sudden change in airflow and then recover.
http://www.nbs-stl.com/CRX/AE%20Testing.png
On this install the AFR will be ignored for about .15 sec.
Andy
-
- Helpful MS/Extra'er
- Posts: 37
- Joined: Thu Feb 23, 2017 3:33 pm
- Location: 808 HI \[]]]/ ¤shaka¤
Re: Reading MLV Data Logs
So what are other useful mlv datalog setups do you have(ex:rpm,map,tps), and what are you looking for and why? Thanks a bunch for all of these useful info, to me these are very valuable. I would want to only have this seen for registered/verified ms purchasers
Efi Tuning Noob, so please go easy on me.
AE92 GZE'd 85 AE86 3dr(blue "cancer" pill)
ATS Racing 2.1 stroker GT3071R 92 MR2t t-top (red pill) DIYPNP w/5sgte in the works
"Nobody listens to Turtle"
AE92 GZE'd 85 AE86 3dr(blue "cancer" pill)
ATS Racing 2.1 stroker GT3071R 92 MR2t t-top (red pill) DIYPNP w/5sgte in the works
"Nobody listens to Turtle"
-
- Helpful MS/Extra'er
- Posts: 37
- Joined: Thu Feb 23, 2017 3:33 pm
- Location: 808 HI \[]]]/ ¤shaka¤
Re: Reading MLV Data Logs
Also, whats Geez GB?
Efi Tuning Noob, so please go easy on me.
AE92 GZE'd 85 AE86 3dr(blue "cancer" pill)
ATS Racing 2.1 stroker GT3071R 92 MR2t t-top (red pill) DIYPNP w/5sgte in the works
"Nobody listens to Turtle"
AE92 GZE'd 85 AE86 3dr(blue "cancer" pill)
ATS Racing 2.1 stroker GT3071R 92 MR2t t-top (red pill) DIYPNP w/5sgte in the works
"Nobody listens to Turtle"
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
I have 3 axis accelerometers mounted in the car. They are dirt cheep now. They come out of the Drone and Robot world.
One G is the acceleration of gravity. .911 G works out to 20 MPH per sec. In the log, Geez GB is Acceleration Gas/Brake Geez LR is Left/Right.
Andy
One G is the acceleration of gravity. .911 G works out to 20 MPH per sec. In the log, Geez GB is Acceleration Gas/Brake Geez LR is Left/Right.
Andy
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
Here is what is known as the traction circle. Geez GB on the vertical axis, Geez LR on the Horizontal axis. Total Geez is in the field in color. The goal on any road race car is to stay around the perimeter or transitioning thru the center very quickly. You need the HD version of MLV to do this type graph.
http://www.nbs-stl.com/CRX/CRX%20Traction%20Circle.png
This is the same car but zoomed way in on a normal trace. I am looking for a clean pull of power when we nail the motor coming off a corner. A rise in engine RPM does not always translate to acceleration. Controllable acceleration gets you from point A to point B the fastest.
http://www.nbs-stl.com/CRX/CRX%20On%20T ... sponse.png
The accelerator board is in the black box at the base of the windshield. The actual board is about the size of a quarter. Most of the volume of the box is taken up with memory foam to kill motor and chassis vibration.
http://www.nbs-stl.com/CrxTypeR/B18C.jpg
Andy
http://www.nbs-stl.com/CRX/CRX%20Traction%20Circle.png
This is the same car but zoomed way in on a normal trace. I am looking for a clean pull of power when we nail the motor coming off a corner. A rise in engine RPM does not always translate to acceleration. Controllable acceleration gets you from point A to point B the fastest.
http://www.nbs-stl.com/CRX/CRX%20On%20T ... sponse.png
The accelerator board is in the black box at the base of the windshield. The actual board is about the size of a quarter. Most of the volume of the box is taken up with memory foam to kill motor and chassis vibration.
http://www.nbs-stl.com/CrxTypeR/B18C.jpg
Andy
-
- Super MS/Extra'er
- Posts: 1978
- Joined: Wed Apr 15, 2009 11:31 am
- Location: Oxford, PA
- Contact:
Re: Reading MLV Data Logs
Pics of the inside of the box? Very cool.
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
Google ADXL335
I also used this to power the ADXL335
DROK Waterproof Auto Vehicle Voltage Reducer DC-DC Buck Converter 12V to 3.3V Step-down Voltage Regulator Transformer 3A/10W Power Supply
Andy
I also used this to power the ADXL335
DROK Waterproof Auto Vehicle Voltage Reducer DC-DC Buck Converter 12V to 3.3V Step-down Voltage Regulator Transformer 3A/10W Power Supply
Andy
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
Here is a link to data coming off my ITB Honda. All runs are using the tune that is included. This is the first event on a tune that was roughed in on the dyno. The is plenty of room for tweaking on this tune.
http://www.nbs-stl.com/CrxTypeR/B18C%20 ... t%2001.zip
I find it best to open all of the data logs at the same time to view what the motor is really doing in the big picture.
Andy
http://www.nbs-stl.com/CrxTypeR/B18C%20 ... t%2001.zip
I find it best to open all of the data logs at the same time to view what the motor is really doing in the big picture.
Andy
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
This is what the traction circle looks like for that car. I have filtered out all data below non racing speeds like when driving up to the start line. This is all done for less than $10 spent on Amazon.
http://www.nbs-stl.com/CrxTypeR/B18C%20 ... Circle.png
Megasquirt data logging is not just for motors any more....
Andy
http://www.nbs-stl.com/CrxTypeR/B18C%20 ... Circle.png
Megasquirt data logging is not just for motors any more....
Andy
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
Here is how I start looking at the data. I filter out Transients and startups. I am looking at clearly red or blue areas. Some show up more clearly on the Speed Density view and others show up better on the Alpha-N view.
http://www.nbs-stl.com/CrxTypeR/B18C%20 ... %20One.png
Andy
http://www.nbs-stl.com/CrxTypeR/B18C%20 ... %20One.png
Andy
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
I sure never expected to find this....
I sent the driver out to test running in second gear on an autocross course and then told him to run the entire course in third to see if there was a difference in times. I compared the corner times based on corner entry. What I found is it really does not matter what gear you choose.
http://www.nbs-stl.com/CrxTypeR/2ed%20g ... %203rd.png
Lesson.... Be willing to look at the data.
Andy
I sent the driver out to test running in second gear on an autocross course and then told him to run the entire course in third to see if there was a difference in times. I compared the corner times based on corner entry. What I found is it really does not matter what gear you choose.
http://www.nbs-stl.com/CrxTypeR/2ed%20g ... %203rd.png
Lesson.... Be willing to look at the data.
Andy
Last edited by whittlebeast on Tue Mar 21, 2017 3:44 am, edited 1 time in total.
-
- Helpful MS/Extra'er
- Posts: 37
- Joined: Thu Feb 23, 2017 3:33 pm
- Location: 808 HI \[]]]/ ¤shaka¤
Re: Reading MLV Data Logs
Keep talking! Loving the amount of info.
Efi Tuning Noob, so please go easy on me.
AE92 GZE'd 85 AE86 3dr(blue "cancer" pill)
ATS Racing 2.1 stroker GT3071R 92 MR2t t-top (red pill) DIYPNP w/5sgte in the works
"Nobody listens to Turtle"
AE92 GZE'd 85 AE86 3dr(blue "cancer" pill)
ATS Racing 2.1 stroker GT3071R 92 MR2t t-top (red pill) DIYPNP w/5sgte in the works
"Nobody listens to Turtle"
Re: Reading MLV Data Logs
Thanks for all the good info Andy. I need to study your logs and get my head around the explanations. Might be benificial to set my logs up as you suggest to compare
Thanks again,
Jim
Thanks again,
Jim
Toyota MR2 98 3sgte
MSPNP2 with 3.57 mainboard
Innovate WB O2 & soon to be external GM 3 bar MAP
FTDI USB interface on SD Android 5.1 head unit
TD05SL2-18g with external WG
HKS EBC running 17psi boost.
A2W IC with WI 50/50 meth/water
MSPNP2 with 3.57 mainboard
Innovate WB O2 & soon to be external GM 3 bar MAP
FTDI USB interface on SD Android 5.1 head unit
TD05SL2-18g with external WG
HKS EBC running 17psi boost.
A2W IC with WI 50/50 meth/water
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
I happen to own one of the red Qstarz 10 hz devices with data logging. I export the data as a CSV and load the data into MLVHD.
Here is what I get to document what the course looked like for the day. The MLVHD displays the average speed in each sector (box) It is shocking how slow autocross really is.
http://www.nbs-stl.com/qstarz/QStarz%20Track%20Plot.png
Andy
Here is what I get to document what the course looked like for the day. The MLVHD displays the average speed in each sector (box) It is shocking how slow autocross really is.
http://www.nbs-stl.com/qstarz/QStarz%20Track%20Plot.png
Andy
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
I was curious how the throttle position related the acceleration of the chassis. The goal is to have what the driver has control of, the throttle position relates to the acceleration. The more linear, the better. What I found on the new motor shocked me.
http://www.nbs-stl.com/CrxTypeR/Throttl ... ration.png
Check out the two traces on the bottom.
Andy
http://www.nbs-stl.com/CrxTypeR/Throttl ... ration.png
Check out the two traces on the bottom.
Andy
Re: Reading MLV Data Logs
Andy,
Attempting to add just a bit. Hope it helps. I really want to digest this thread more but also want to share. You did a great job on it.
You had made a comment that Oxygen sensor delay seemed about 0.25s for your engine. Coming from a long ago background in fuel systems, when only narrow band sensors were available, we determined that the delay of the sensor was related to RPM by something we called transport delay time. Basically if you think about it, it makes sense. If you change the A/F ratio at the injector, that volume of mixture has to go through the engine and manifold/header until it can be read by the sensor. There is a delay in the sensor but it was significantly shorter than the delay of the engine and could be largely ignored. I haven't looked in to the differences in response time between narrow band and wide band sensors, so that may be the difference, but 0.25s seems like much longer than we measured then.
I wish I could remember the numbers but it worked out to be tied to a number of crankshaft revolutions for any particular engine setup. Since most packages are relatively similar as far as sensor and injector placement relative to the piston, the relationship didn't change much from engine to engine in most cases. If someone wants to criticize this as being overly simplistic, I know it is. There are many other factors. But our data showed that was the dominant function and you were left with a rather simple relationship that worked quite well for trying to meet federal emissions in the 80's.
Attempting to add just a bit. Hope it helps. I really want to digest this thread more but also want to share. You did a great job on it.
You had made a comment that Oxygen sensor delay seemed about 0.25s for your engine. Coming from a long ago background in fuel systems, when only narrow band sensors were available, we determined that the delay of the sensor was related to RPM by something we called transport delay time. Basically if you think about it, it makes sense. If you change the A/F ratio at the injector, that volume of mixture has to go through the engine and manifold/header until it can be read by the sensor. There is a delay in the sensor but it was significantly shorter than the delay of the engine and could be largely ignored. I haven't looked in to the differences in response time between narrow band and wide band sensors, so that may be the difference, but 0.25s seems like much longer than we measured then.
I wish I could remember the numbers but it worked out to be tied to a number of crankshaft revolutions for any particular engine setup. Since most packages are relatively similar as far as sensor and injector placement relative to the piston, the relationship didn't change much from engine to engine in most cases. If someone wants to criticize this as being overly simplistic, I know it is. There are many other factors. But our data showed that was the dominant function and you were left with a rather simple relationship that worked quite well for trying to meet federal emissions in the 80's.
Steve B.
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
Steve B
The funny thing is when I apply a delay with filters and look at a big data log and then compare that to a log without the delay and filters, you will find very little difference in colors and conclusions.
Andy
The funny thing is when I apply a delay with filters and look at a big data log and then compare that to a log without the delay and filters, you will find very little difference in colors and conclusions.
Andy
-
- Super MS/Extra'er
- Posts: 2221
- Joined: Tue May 04, 2004 8:20 pm
- Location: St Louis
- Contact:
Re: Reading MLV Data Logs
I got a hour long data log from a friend in England. I helped him learn to tune about a year ago. His tune is really coming along. This happens to be a Miata.
http://www.nbs-stl.com/tuning/Fuel%20Ge ... 0Close.png
Andy
http://www.nbs-stl.com/tuning/Fuel%20Ge ... 0Close.png
Andy
Re: Reading MLV Data Logs
Andy,
In that case I expect it is a difference in the sensor response time. We were using what ever we could get that was fast as possible. My comments on transport delay are probably relating to an insignificant percentage. Technology is great.
In that case I expect it is a difference in the sensor response time. We were using what ever we could get that was fast as possible. My comments on transport delay are probably relating to an insignificant percentage. Technology is great.
Steve B.