Street Build ITFLYS - 1Gina2G

[Justin DuBois]

Did something improve that I'm not seeing? It's still choppy af.
Switching to a 5bar sensor cut your resolution significantly.

Its much smoother now, both boost stability and AF.

 

[Justin DuBois]

The jump in your cells between 6000 and 6500 is to drasitic. That's causing your airflow to start looking jittery around that rpm and above.

The GM sensor seems to have an issue, and the VE map has bias tuned into it.

I need to retune the VEmap with the new sensor

 

[Stapl3]

It's smoother because half the resolution is missing.

 

[Justin DuBois]

It's smoother because half the resolution is missing.

This is a 50PSI sensor, not the 75PSI sensor

 

[Stapl3]

Ok. It is smoother because more resolution is missing. :]

It's still choppy. Nothing was fixed IMO.

 

[Justin DuBois]

Choppy? Ok. I'll take 1psi oscillation over 4psi oscillation, even if it turns out that the sensor is smoother due to improved filtering. The air turbulance caused by valves opening and closing is not helpful to capture in the data used to deliver fuel.

The voltage resolution you are referring to is: 0.03v less per psi - ~27% difference.
GM3Bar - 9.1 psi/volt or 0.11volt/psi
ZT-35bar - 12.5 psi/volt or 0.08volt/psi

In comparison
ZT-5bar - 18.75 psi/volt or 0.05volt/psi ~45% difference

Both resolution and sensor filtering will help to smooth out the data - and in this case, I say that is a good thing.

 

[Justin DuBois]

Going there:

 

[pauleyman]

Post the pics if you want but you should also post the actual log file. Makes it easier for us to see everything.

 

[Justin DuBois]

Post the pics if you want but you should also post the actual log file. Makes it easier for us to see everything.

I agree! Which post is missing a log you were looking for? - I went back a week and attached the pictured logs.

 

[pauleyman]

I was speaking generally.

 

[Stapl3]

Choppy? Ok. I'll take 1psi oscillation over 4psi oscillation, even if it turns out that the sensor is smoother due to improved filtering. The air turbulance caused by valves opening and closing is not helpful to capture in the data used to deliver fuel.

The voltage resolution you are referring to is: 0.03v less per psi - ~27% difference.
GM3Bar - 9.1 psi/volt or 0.11volt/psi
ZT-35bar - 12.5 psi/volt or 0.08volt/psi

In comparison
ZT-5bar - 18.75 psi/volt or 0.05volt/psi ~45% difference

Both resolution and sensor filtering will help to smooth out the data - and in this case, I say that is a good thing.

Here is a log of a car I was working on a few years ago, this is a pull before the clutch gave up. This is comparable to you as it's a 16g @ 17psi, very similar physical setup. I left boostest on for keks. But seriously, look at how stable it is.

ECMTuning User Support Forums

This is a discussion forum for registered ECMLink users. See http://www.ecmlink.com .

www.ecmtuning.com

It may be interesting to swap the brake booster and the MAP sources on the manifold.

 

[Justin DuBois]

Here is a log of a car I was working on a few years ago, this is a pull before the clutch gave up. This is comparable to you as it's a 16g @ 17psi, very similar physical setup. I left boostest on for keks. But seriously, look at how stable it is.

ECMTuning User Support Forums

This is a discussion forum for registered ECMLink users. See http://www.ecmlink.com .

www.ecmtuning.com

It may be interesting to swap the brake booster and the MAP sources on the manifold.

View attachment 715435

The blue Omni4Bar line on the screen is super hard to see - and that sensor looks like it is producing smooth output.

Omni has stretched its PSI scaling to the voltage limits of the sensor in order to market a 4 BAR rating.

The sensor has a negative voltage offset, so it is clipping and using the 0-0.5v range for vacuum - the accuracy there is likely less, or obscured by other voltage noise in the system. At 11.963 psi/volt, the voltage resolution appears in-between the GM 3BAR, and the AEM 3.5 BAR. 17psi is in a good range for the sensor. I would question any 34-45 PSI readings at the top of the scale between 4.5-5v

But - as you show, 17psi is smooth. Now, do you think it is smooth because the pressure in the manifold is smooth or do you think the PSI is smooth due to the way the sensor is designed?

Boost control on this pull is very good - What kind of setup was the turbine, o2, WG and downpipe.

 

[Justin DuBois]

Back on boost:

 

[BeatupVR4]

Have you tried mechanically smoothing out the response with a snubber? I've used them in hydraulic ram systems that are PID controlled as well as leak detection systems that have to read 0-2 inWC ( 0.072 psi) with 100psi dynamic operating pressure. You'll probably have to play with the orifice size but they make more expensive ones with changeable "pills" similar to a nitrous jet.

 

[Justin DuBois]

Have you tried mechanically smoothing out the response with a snubber? I've used them in hydraulic ram systems that are PID controlled as well as leak detection systems that have to read 0-2 inWC ( 0.072 psi) with 100psi dynamic operating pressure. You'll probably have to play with the orifice size but they make more expensive ones with changeable "pills" similar to a nitrous jet.

I have not - but I like the suggestion. No need now with a better sensor.

 

[Stapl3]

But - as you show, 17psi is smooth. Now, do you think it is smooth because the pressure in the manifold is smooth or do you think the PSI is smooth due to the way the sensor is designed?

Boost control on this pull is very good - What kind of setup was the turbine, o2, WG and downpipe.

MHI e16g, ebay o2 housing, unknown 3" downpipe ending under the driver's ass. Intercooler, BOV I hate, standard MBC ball/spring. It's a fairly stock car.

I believe your situation has not changed, rather your new sensor shows you toward what you wanted. In terms of a proper experiment, I think it would be interesting to throw another 3-bar on there...

 

[Justin DuBois]

I'm getting closer to MAX boost again. I saw 27.3 PSI in the initial boost spike and then 24psi tapering down to 22psi - but I don't have enough WOT data yet. A few more runs and I'll have the map filled out.

What I'm wondering is if anyone has SD Idle hunting like we see in the beginning of the log. This starts happening after everything is nice an warm/hot.

 

[delta448]

I've been away from tuning for awhile and I have limited experience using ECMLink, but I am curious, do you have closed loop fuel trimming enabled? If so, what's the minimum injector pulsewidth correction per cycle?

I'm assuming you're using much larger than stock 450s.

Might try locking into open loop to see if it still oscillates.

 

[Justin DuBois]

I've been away from tuning for awhile and I have limited experience using ECMLink, but I am curious, do you have closed loop fuel trimming enabled? If so, what's the minimum injector pulsewidth correction per cycle?

Yes - closed loop is working, Fuel trims are +- 5% and during idle, the minimum InjDuty cycle is 1.2% with these RC-1000cc injectors

I'm assuming you're using much larger than stock 450s.

Yup

Might try locking into open loop to see if it still oscillates.

It probably won't - that is a good data point to check.

 

[Justin DuBois]

Here is my latest attempt at MAX boost. - 29psi Spike - yeah.

 

[Stapl3]

I'm getting closer to MAX boost again. I saw 27.3 PSI in the initial boost spike and then 24psi tapering down to 22psi - but I don't have enough WOT data yet. A few more runs and I'll have the map filled out.

What I'm wondering is if anyone has SD Idle hunting like we see in the beginning of the log. This starts happening after everything is nice an warm/hot.
View attachment 715878

That is a BISS adjustment, as 101 as it sounds.

 

[Justin DuBois]

Kinda freaking me out how its staying true. I touch up a cell here or there, but It stays true to estimate on all the follow on logs.

About the only thing I'm fussing with is the boost control. For this Big16G, I'll see if I can get a consistent 23psi.

 

[Justin DuBois]

The period of the idle hunting matches the signal from the front 02 sensor cycling:

Hmm, I'll try NB simulator and see if that changes the result. I wonder if a new front 02 would resolve this.

 

[We're on Boost]

The period of the idle hunting matches the signal from the front 02 sensor cycling:

Hmm, I'll try NB simulator and see if that changes the result. I wonder if a new front 02 would resolve this.

I don't know Jus, mine does the same thing sometimes, I think. I have a hard time seeing it in my logs. But when I'm in the car and judging rpm by sound, when I hear an oscillating idle speed and look at the wideband gauge, it usually seems like the higher or stronger idle sound corresponds to slightly richer on the wideband gauge, which would be a voltage hump (high) on the front O2.
I figure it's a natural result of how closed loop works, so I don't worry about it.

I don't know if the ISC is supposed to work fast enough to track that and compensate for it.
On my car the ISC is usually mostly bottomed out (at 0) after 20 minutes or so of running. So at that point I essentially don't have an ISC for idle any more.
I'm fine with that because it gives the ISC more headroom in the other direction - which I need for fast idle after a cold start. (No FIAV you know.)
But I see on yours, the ISC still has some room to go, 15 steps or so remaining, and it's not doing much with it.
If it had more headroom, like the "standard" 30 steps at your designated target idle speed, maybe it would do more? I don't know.
What I've always thought about trying to run an engine at stoic during idle conditions is, no, it will always idle better (stronger, more stable) if the mixture is a little richer than stoic. Closed loop takes that all away from you, it's going to be stoic on the average or die trying.

What happens if you raise your target idle from 850 to say 950, or 1000, and just leave the BISS where it is?
That should give your ISC a little more headroom, and it might also make the O2 cycle a little faster which could make the trim up-down cycle a little shorter in time, quicker. Anyway it would be easy to try and that's always nice!
Or, if you really want the low idle, close the BISS a little bit so you are throwing more of it onto the ISC?
For me, I've had it with slow weak idles. Let that baby run is what I think.

Your simulated idle switch seems to be working fine so that's not it.

 

[Justin DuBois]

Switching to narrow band sim, the results are nearly the same.
Raising the RPM to 1050, similar result.
Open loop mode - no improvement.