SOLD - 1600 TURBO + oval + downunder

[petew]

I realize that CHT sensor is not ideal, but I figured it was something to work with so I went with it. I was able to monitor any changes in head temps pretty fast. You could see the temp climb quickly enough as the load increased. I'll post something later.

[petew]

Hi all, back from the work thing. Filled up the car and did some maths just now. I used 57 litres of fuel. A little more than I first thought. I travelled about 650km. So that works out as 8.7 L/100km. That's not a long way off a modern Subaru or similar. Considering it was revving a lot and had lots of hills to climb, that's not bad. Our diesel Golf does about 5 L/100km on a long trip. I also checked the oil. It's dropped a bit, but not near as much as I'd thought. The problem is the oil ends up on the exhaust and STINKS.

canberra oil leaks.jpg

As promised here's just some of the log from the way home. I sat on 70mph most of the way. You can see the car climbing a hill on very slight boost in the middle of the log.

Climbinghillslog.png

0.7psi, the MAT climbs (31-43degC) as does the CHT (69.3 to 73.4degC). As I mentioned, the TPS says a lot. It seems to be waving all over the place. I think that's the reality. A light throttle and vibration. Duty cycle max was barely above 50%.

Love to hear some people's thoughts on that log. If anyone wants a copy of the full log to look at, just let me know.

[madmike]

Sorry I'm no help, but that looks like my latest 'Cardiogram'

[petew]

Been a "mandatory in-service training day" all day today. Spent the whole afternoon learning more about domestic violence. Oh man. Do me a favor and speak up if you know any woman (or man) trapped in an abusive relationship.

Autotune 23.6.17.jpg

Managed to datalog and autotune both ways. One of things I've been noticing as I've been driving is the car seems to throw weird AFRs (15-17) when rolling down hill or coasting to a stop on zero throttle. It should be dead lean but isn't. So I went through the log and found it...

nofuelcut23.6.17.jpg

So I went looking and the over run fuel cut was turned off. Which makes me think I could have saved a heap more fuel on the way home from Canberra.

The other weird thing that happened was the motor died... in traffic... on the freeway. Yay. Thankfully it restarted straight away. Can't seem to sort out why it happened. But the PWM idle duty just drops out. Hmm... It does seem to be cycling during normal driving. Not sure why.

[andy198712]

fuel cut can be a tricky one to setup smoothly in my experience, was jolty as it kicked back in... defiantly feel it cutting fuel after a preset delay time, a lot more engine braking! you can write the map so the extreme overrun section (that never gets used on cruise ect) is mega lean and do it that way.... i tried that way too, to try and smooth the transition.

[petew]

fuel cut can be a tricky one to setup smoothly in my experience, was jolty as it kicked back in... defiantly feel it cutting fuel after a preset delay time, a lot more engine braking! you can write the map so the extreme overrun section (that never gets used on cruise ect) is mega lean and do it that way.... i tried that way too, to try and smooth the transition.

I switched the fuel cut on a few days ago. Then I took the car out and gave it a proper thrashing trying to rev it out to the limiter. The highest I got was 5500rpm. What I figured out was it just wasn't worth revving it past there. You can see why.

5500rpmlog28.6.17.png

The gate starts removing boost pretty quick and I was down under 3psi by the end of it. So it will need more boost to really make use of the RPM range. Beyond that, the fuel cut made no noticable difference to driving. I think I've got it so well tuned for street driving, it drops off under 40kpa anyway. One thing I did have issues with the is the idle control. I suspect the stepper motor is old and will need replacing at some stage. No biggy. I'm still a little suspect about the 85% injector duty readings. I didn't lean out when I had the car running 14psi before, but I don't want to learn the hard way.

Yesterday I got a few other little things sorted. A friend gave me a plastic glovebox replacement for nix. He forgot to put it in his oval and then sold the car. I've also had a USB charging point lying around waiting to go in. My auto electrician friend put in the power feed for it, so it was just 2 connector and a big hole for the fitting.

absglovebox1.jpg

absglovebox2.jpg

The other issue I've mentioned before is dramas with the racing seat being too hard even for shortish drives. My butt literally goes numb. I've had my eye out for some foam. On tuesday morning I found an old seat schwab on the side of the road while dropping my daughter at school. I stripped off the old fabric and cut it down to fit.

seatpadding.jpg

This is HEAPS better. I'm sitting a little higher, but I'm not sliding around in the seat and it's not like sitting on a park bench anymore. Funny how little things like this can make a big difference.

[Chip Birks]

Going pig rich at peak rpm doesn't help things either. I bet if there was less fuel there, it would keep the boost up better.

[Clonebug]

As Chip says you are pig rich at full boost for quite long in the throttle time.
It looks to me you have plenty room for improvement yet.
Keep working at it.
Since it is at a solid 10.0:1 it could even be richer.
You can do the numbers with some simple math to take some of the tuning time out of it. Divide 10 by 12(actual afr and target afr) and take that percentage off the VE numbers to get it closer. It will at least show a little change and you can go from there. So 10/12=.83 therefore multiply each cell you want to change by .83 to get a new number.... input that number and then run a new log to see if the AFR changes.
If too much or too little then run the numbers again.
Either that or increase the boost.......

[petew]

I took my older son out to a Pokemon (card game) tournament this morning. Man, what a weird group the role playing crowd are. They make us volksphiles look well adjusted and normal. On the way home, I got the car warmed up and decided to try and get the car up near the redline to log AFRs better. This was following CB's advice as the last log was done while I was autotuning. I wanted to be sure about what I was working with before I fiddled with the fuel numbers on boost.

The first time I really wound it out in 1st and 2nd gears, my 11yo son exclaimed...

"Don't do that again! I felt like I was going to fall out of the car!"



I roared with laughter and did it again at the next set of lights. "Do you have to do that?!", was the next exclamation. So I explained what I was doing and told him to keep and idea on the tacho to try figure out how high I was revving it. He got over his fear (he's a bit anxious) and I managed 5400rpm. However, when I got home the datalog tells a slightly different story.

5500rpmlog1.7.17.png

Looking through the log, the motor spun out to 5500rpm, the last 100 after I'd lifted off. When the throttle is down, the AFR is a little lean if anything. Once I lift off it goes pig rich for moment. So I'm not entirely sure what to do with that. Maybe the lambda delay is still not quite right.

[Piledriver]

Your wideband absolutely cannot read a 7.5:1 AFR.
Very few can read >20:1 either.
What unit do you have?

The pulldown wideband model selections are convenient, but they are really a non-feature, as they are almost absolutely guaranteed to be wrong, and they are not tweakable.

You need to calibrate your widebands input to the MS to even know if all you have is an analog output.

The most practical way to do this is via the "custom linear wideband" tool, and find the rails---twiddle the voltages until 10:1 is the min reading, and 20:1 (fuel cut) is the max, typically.
Not quite as good as reference gasses or the 2 step cal voltages a newer 14point7.com unit puts out on initialization, but a whole lot closer to reality than 7.5:1.

A few WBO2 models can read 10:1>free air, but that capability is not common on the analog output.

If you have a digital output as well and can also run that software at the same time as TS you can use that to calibrate:; The widebands themselves are pretty accurate when working properly, but the analog signal handling will always need calibration.

You need to do this sooner than later so your logs and reality mesh.

[petew]

Innovate LC2

PD, I'm really confused as to how to do what you suggest. Don't even know where to begin really. Is there a how to thread or article somewhere?

[petew]

Your wideband absolutely cannot read a 7.5:1 AFR.
Very few can read >20:1 either.

Hang on a mo, the reason you're seeing 7.5:1 is coz the motor died a couple times when the idle control didn't kick in properly. Otherwise it's within 10-22.

[Piledriver]

...and the 22:1 reading? An LC1 analog output rails at 20:1.

It's actually very easy, pull down "custom linear wideband", punch 10:1 as low end, 20:1 as high end,and twiddle the two voltage entries until those rail at their respective afrs.
An LC1 also has a serial output, beg or borrow a second usb <>serial adapter and run the innovate software to read the digital data, use that for an accurate calibration while running TS.

The output voltages of an LC1 are documented so start there, you probably are only a dozen or so millivolts off on either end, but its really the only way unless you go full digital comms over CAN.

[petew]

...and the 22:1 reading? An LC1 analog output rails at 20:1.

I've watched the Innovate LC2 gauge show 22.4. So Either the gauge is lying or it gets close to it.

[Piledriver]

I apparently cannot read.

I stand corrected.
an LC2s range is 7.35>22.29, so those should be the high/low rails if the analog calibration is correct.
Sounds like its far closer than usual to correct, they must have stiffened up the analog output driver.

Its not unusual for LC1s (or many other analog output WBO2s) to be high or low a full AFR point on the analog out<>MS connection, the error is from a little of both.