Morning to afternoon tune

All questions about MS2/Microsquirt/Microsquirt module. See also MS2/Extra manuals

Moderators: jsmcortina, muythaibxr

ol boy
Super MS/Extra'er
Posts: 1532
Joined: Mon Sep 10, 2007 3:06 am
Location: Tucson, Az

Morning to afternoon tune

Post by ol boy »

I've been chasing a weird tuning behavior over the past year or so. I can do an auto tune on the drive to work and end up with a fairly good tune with EGO correction +-2%. Then in the evening on the ride home EGO is stuck at 105% for most of the drive. The car drives noticeably leaner in the evenings. Total trip only has a few hundred feet in elevation change. Baro starts at(home) 94ish kpa and finishes at(work) 92 kpa.

The car is a 64 Ford Falcon with a 306, VS Racing 7668 single turbo blowing into a carb to EFI converted torquer2 single plane intake. Deka60's, and D581 coils. Using a LC2 WBO2 system.

Before everyone jumps on the air intake temp correction (idle gas law) band wagon, I've flattened it to 102% at -40* and it ends a 100% at 125* then goes up to 106% a 250*F.

Whats making this really hard to sort out is with a non-intercooled turbo car the air intake temp run between 140*F and 150*F all the time while not in boost. Ambient air temp could swing from 50 to 110 and the intake temp will get no higher than 160*F in mid summer.

In my logs I can see the PW is larger in the evening but still yields a leaner mixture.

I've began to wonder if there is a mechanical problem with the engine... or if the injector drivers get warmer in the evening which increases the RDsON value which increases the injector deadtime enough to move the fueling by 5%..? :roll: The WB controller is mounted in the car under the front bench seat. In car temps will follow ambient temps as the car is just old and lacking door seals, AC/heat, ect... During over run fuel cut the AFR shows 21.7 and not 20.0 like I'd expect. Any chance of a voltage drift on the WB analog out wire as the cab heats up from morning to evening?

Just tossing out ideas. I'll post msq and a link to a drop box account with the morning and evening logs. Each one is in the 10MB range.

Thanks Ryan
306 SBFord, Torquer II EFI intake, 60 lbs injectors, 8 LS2 coils, VS Racing 7668 turbo, 4R70W, MS3x fw1.4 w/built in trans controller.
benckj
Master MS/Extra'er
Posts: 464
Joined: Mon Dec 19, 2011 3:21 pm

Re: Morning to afternoon tune

Post by benckj »

My first thought was IAT response possibly lagging which could throw the tune off. What type of IAT sensor did you install and where did you put it?

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
ol boy
Super MS/Extra'er
Posts: 1532
Joined: Mon Sep 10, 2007 3:06 am
Location: Tucson, Az

Re: Morning to afternoon tune

Post by ol boy »

Stock GM IAT and it's plumbed into the intake runner, #4 if I recall right. It's not laggy at all. I can see when the water injection comes in and out. Like I said before, it reads between 140 and 150 after 10 minutes of run time no mater the outside air temp. I'm temped to unplug it so it reads 70F and do 100% across the IAT curve, but it won't change the math much from what it's all set to now.
306 SBFord, Torquer II EFI intake, 60 lbs injectors, 8 LS2 coils, VS Racing 7668 turbo, 4R70W, MS3x fw1.4 w/built in trans controller.
benckj
Master MS/Extra'er
Posts: 464
Joined: Mon Dec 19, 2011 3:21 pm

Re: Morning to afternoon tune

Post by benckj »

I know with various IAT sensors and the surrounding mass they can heat soak which leads to false readings. Missed the point about running WI and seeing this on logs including reaction time can discount reason for poor tune.

Several of us have had problems with the correct O2 readings especially when using an Innovate LC1. What cal values are you running as the factory setting seem to translate to a point off between gauge and TS.
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
2swe
Experienced MS/Extra'er
Posts: 330
Joined: Sun Dec 25, 2011 12:41 pm

Re: Morning to afternoon tune

Post by 2swe »

Baro correction ?

do you have datalogs , morning and afternoon trip both with the same msq , without auto tune...
1978 vw golf gti 16v itb cop full sequential MS2
1992 vw corrado 16v moonlight blue MS2
ol boy
Super MS/Extra'er
Posts: 1532
Joined: Mon Sep 10, 2007 3:06 am
Location: Tucson, Az

Re: Morning to afternoon tune

Post by ol boy »

I'll push the tune and logs into this folder.

https://www.dropbox.com/sh/pgu62dm08xfq ... 8s3Sa?dl=0

Baro correction is also 100% across the kpa range. I only have 2 kpa spread from home to work.
306 SBFord, Torquer II EFI intake, 60 lbs injectors, 8 LS2 coils, VS Racing 7668 turbo, 4R70W, MS3x fw1.4 w/built in trans controller.
racingmini_mtl
Super MS/Extra'er
Posts: 9129
Joined: Sun May 02, 2004 6:51 am
Location: Quebec, Canada
Contact:

Re: Morning to afternoon tune

Post by racingmini_mtl »

How good is your deadtime value? Did you actually measure it?

Jean
jbperf.com Main site . . . . . . . . . . . . . . . . . . . . . . jbperf.com Forum
Image
ol boy
Super MS/Extra'er
Posts: 1532
Joined: Mon Sep 10, 2007 3:06 am
Location: Tucson, Az

Re: Morning to afternoon tune

Post by ol boy »

added msq. for drive home tonight and return trip for tomorrow morning. added a notes.txt for file info on the drive file names and the msq they go with.

Jean: Thats another topic I wanted to bring up. From the info I've found on deka 60's, something around 0.4ms seems a bit short. I ran those values for a bit and the VE table had a big bowl shape to it. Any load under 40ish% the VE values started going back up quite steeply. I increased the deadtime to 0.8ms and the VE table flattened out quite a bit and the overrun and low % load values seemed more reasonable. A steady decline in values from 80% to 20%. I've had too large of a deadtime before and you end up with a steep fall off and the lower loads. I've settled in on a value thats working for the most part.

Others jumping in.. The logs have other info in them. gear, watermeth%, MPH, MPG(not right), odometer, trip meter, x/y accel data and a few others. I'm receiving CAN data from the ECU into a teensy3.2 and dumping that data into a CSV on a uSD card.

many thanks Ryan
306 SBFord, Torquer II EFI intake, 60 lbs injectors, 8 LS2 coils, VS Racing 7668 turbo, 4R70W, MS3x fw1.4 w/built in trans controller.
ol boy
Super MS/Extra'er
Posts: 1532
Joined: Mon Sep 10, 2007 3:06 am
Location: Tucson, Az

Re: Morning to afternoon tune

Post by ol boy »

Added to the drop box link the drive from last night and this morning.

DATA263 is the drive home and DATA264 is the drive back to work. The msq is the same for both trips.

https://www.dropbox.com/sh/pgu62dm08xfq ... 8s3Sa?dl=0

during the freeway cruising section you can see this mornings drive required an EGO correction of 96% and the same freeway cruise section required 103%.
306 SBFord, Torquer II EFI intake, 60 lbs injectors, 8 LS2 coils, VS Racing 7668 turbo, 4R70W, MS3x fw1.4 w/built in trans controller.
EspeNS
Experienced MS/Extra'er
Posts: 158
Joined: Sun Dec 13, 2009 11:10 am
Location: Norway

Re: Morning to afternoon tune

Post by EspeNS »

Having mesured 60lb/hr Siemens injectors, I know 0.8ms is likely about 0.2ms to high. Measured with MS3X injector drivers, so most likely a bit higher deadtime than you have.
I have a spreadsheet on my tuning laptop if it would help. But I would recommend measuring deadtime and flow with your fuelsystem and injector drivers.

Espen
-84 Ford Sierra 2,0ohc GL, soon to be MS'd, NA tuned with rallyecam and TBI.
-98 Mustang Cobra, track car.
ol boy
Super MS/Extra'er
Posts: 1532
Joined: Mon Sep 10, 2007 3:06 am
Location: Tucson, Az

Re: Morning to afternoon tune

Post by ol boy »

EspeNS wrote:Having mesured 60lb/hr Siemens injectors, I know 0.8ms is likely about 0.2ms to high. Measured with MS3X injector drivers, so most likely a bit higher deadtime than you have.
I have a spreadsheet on my tuning laptop if it would help. But I would recommend measuring deadtime and flow with your fuelsystem and injector drivers.

Espen
I'll take a look at your spreadsheet. What ms/v number did you come up with?

I've added the drive from this morning to the dropbox folder. DATA269.csv. Showed 96%ish correction on the freeway part again. Yesterdays drive home I took a different route because I got off late and just wanted to get home.
306 SBFord, Torquer II EFI intake, 60 lbs injectors, 8 LS2 coils, VS Racing 7668 turbo, 4R70W, MS3x fw1.4 w/built in trans controller.
2swe
Experienced MS/Extra'er
Posts: 330
Joined: Sun Dec 25, 2011 12:41 pm

Re: Morning to afternoon tune

Post by 2swe »

So , I found some differences between 263,264 and 269.

I used scatter plots with RPM for X axis, Egocor for Yaxis and BATv, MAT and Baro for Z
I also removed all records with Load lower than 46 and higher than 48, formula : [Load]<46 || [Load]>48
Schránka 01.gif
Schránka 01.gif (98.88 KiB) Viewed 1237 times
as you can see, higher BATv = leaner mixture, I know its only 0.1 volt difference, but maybe a little bit smaller BATvoltage correction would help.

don't ask me why, but higher MAT = leaner mixture..

and finaly, you need some Barocorrection, lower Baro = leaner mixture.
1978 vw golf gti 16v itb cop full sequential MS2
1992 vw corrado 16v moonlight blue MS2
ol boy
Super MS/Extra'er
Posts: 1532
Joined: Mon Sep 10, 2007 3:06 am
Location: Tucson, Az

Re: Morning to afternoon tune

Post by ol boy »

I'm wondering if comparing all morning datalogs, then do all evening datalogs and see if they show the same trends. What I see in the 3 scatter plots is EGO at both 95 and 105 correction at the no matter the battery volts, mat, and baro. Varies with RPM..

The battery volt plot holds true but only at 2300ish rpm. Not too conclusive in the 1600 rpm area. Same with the other 2 scatter plots. Maybe looking at a specific 100 rpm band and load then see the battery volt, baro and mat.
306 SBFord, Torquer II EFI intake, 60 lbs injectors, 8 LS2 coils, VS Racing 7668 turbo, 4R70W, MS3x fw1.4 w/built in trans controller.
2swe
Experienced MS/Extra'er
Posts: 330
Joined: Sun Dec 25, 2011 12:41 pm

Re: Morning to afternoon tune

Post by 2swe »

yeah, we need more datalogs.

I used 47% load, because there were a lot of hits in this range, but only 1900-2400rpm range contains both morning and afternoon data, 1300-1900 contains only morning (or afternoon, I'm not sure) :D

BATv : 1300-1900, there's mostly 14.2 volt so no Batvoltage correction used, except bottom left 1300rpm/95Egocor ,0.2 volt drop , same trend..

MAT : 1300-1900, mostly 135-150°F, no data for 90-130°F MAT.

Baro : I see the same trend between 1300-1900 and 1900-2400, lower baro - leaner mixture.
1978 vw golf gti 16v itb cop full sequential MS2
1992 vw corrado 16v moonlight blue MS2
EspeNS
Experienced MS/Extra'er
Posts: 158
Joined: Sun Dec 13, 2009 11:10 am
Location: Norway

Re: Morning to afternoon tune

Post by EspeNS »

Deadtime60lbhrSiemens.png
Deadtime60lbhrSiemens.png (60.83 KiB) Viewed 1221 times
Not very good match on this injectors.
0.5-0.6ms @ 13.3V.
This is dedtimes for Injector 1:
0.5ms @ 13.3V
0.65ms @ 12.1V
0.35ms @ 15.3V
Also, nonlinear is measured on injector 1. Not far from linear from 1.3ms and up.
And I belive Simens have 2 different 60lb/hr injectors. This was high ohm, 30deg cone injectors, standard EV1 dimensions and connectors.
injtest Siemens 60lbhr.rar
(35.92 KiB) Downloaded 43 times
Espen

Edit: Deadtime measured @ 4bar fuel pressure
-84 Ford Sierra 2,0ohc GL, soon to be MS'd, NA tuned with rallyecam and TBI.
-98 Mustang Cobra, track car.
ol boy
Super MS/Extra'er
Posts: 1532
Joined: Mon Sep 10, 2007 3:06 am
Location: Tucson, Az

Re: Morning to afternoon tune

Post by ol boy »

Playing with the sheet, does it produce the dead time for you? What is the period for the test? 10, 20, 40 ms? I think I see the goal of getting the same fuel volume at every PW from 2 to 10. If you think the dead time is 0.6ms, then for the test do you enter 2.6ms in test mode and record the volume for 2ms? It seem there is just one formula away from the spread sheet producing the deadtime for you. I can view the trendline formula.

did some tweaking to the sheet. added a line that subtracts the guessed DT value from the PW, 2,4,6,8,10 then divides into the cc/pulse. You should end up with 5 rows of the same value.

Just keep moving the DT value up or down to recalculate the cc/pulse with deadtime.
injtest Siemens 60lbhr.zip
try this
(17.73 KiB) Downloaded 32 times
306 SBFord, Torquer II EFI intake, 60 lbs injectors, 8 LS2 coils, VS Racing 7668 turbo, 4R70W, MS3x fw1.4 w/built in trans controller.
EspeNS
Experienced MS/Extra'er
Posts: 158
Joined: Sun Dec 13, 2009 11:10 am
Location: Norway

Re: Morning to afternoon tune

Post by EspeNS »

Yes if you enter your numner of pulses and cc in the upper left, the curv vil extrapolate the deadtime. I try to fill up the measuring vial as much as possible, to get consistent results, without spilling. Play with the pulses to do that. Without knowing deadtime it's hard to calculate accurately how many pulses needed.
I keep the output interval at least 2ms higher than the pulsewith, to ensure the injector fully closes, and all current is discharged.
I think this spreadsheet was made by one of the devlopers. I downloaded it from the old web msextramanual some years ago.
Dont account for deadtime in the pulsewidth when measuring. In testmode the raw pw entered is used by the MS. Enter 2ms for 2ms, not 2.6ms. 4ms for 4ms, so on.
MS2extra and MS3 work the same in injector testmode. I have measured deadime several times with both MS2 and MS3.
The first time I used output interval to close to the pw.

Espen

Edit: Just for fun. The first time I measured deadtime.
https://youtube.com/watch?v=SoKEbQtnuCo
-84 Ford Sierra 2,0ohc GL, soon to be MS'd, NA tuned with rallyecam and TBI.
-98 Mustang Cobra, track car.
Motoren
Helpful MS/Extra'er
Posts: 69
Joined: Wed Aug 02, 2017 3:24 pm

Re: Morning to afternoon tune

Post by Motoren »

How do you measure the dead time at different volts ? 12.2V engine stopped. 13.5V another car charging the battery? What about 10v ?

Kent
Kent
Supra mk3 2jz-gte
Ms3pro ultimate (installing)
Microsquirt trans control (installing)
Denmark
EspeNS
Experienced MS/Extra'er
Posts: 158
Joined: Sun Dec 13, 2009 11:10 am
Location: Norway

Re: Morning to afternoon tune

Post by EspeNS »

I use a 40A powersuply with adjustable voltage, from ebay.
The first times I used a battery charger and loads like heater fan and lights. One time I used jumper cables from another running car.
The powersupply of course is the easyest way to keep stable voltage.
Offtopic funfact: I use the same powersupply to power a fan when cold smoking bacon 8)
-84 Ford Sierra 2,0ohc GL, soon to be MS'd, NA tuned with rallyecam and TBI.
-98 Mustang Cobra, track car.
ol boy
Super MS/Extra'er
Posts: 1532
Joined: Mon Sep 10, 2007 3:06 am
Location: Tucson, Az

Re: Morning to afternoon tune

Post by ol boy »

https://www.dropbox.com/sh/pgu62dm08xfq ... 8s3Sa?dl=0

made some little VE table tweaks and added the current tune(4xInj_REVH) along with this mornings drive(DATA271) and last nights drive(DATA270).

The tune seems good until warmup ends and MAT becomes stableish, then it goes fat.

There was a question about the WBO2 settings. I set it up for 1-2 volts (10-20AFR) so I could read it from a DMM. I could see a small offset in the output voltage would cause this +-5% ego correction problem. I'll program it back to default LC2 output and redo the AFR table cal. Or maybe tomorrow before the drive home now that I've tweaked the tune.
306 SBFord, Torquer II EFI intake, 60 lbs injectors, 8 LS2 coils, VS Racing 7668 turbo, 4R70W, MS3x fw1.4 w/built in trans controller.
Post Reply