3SGTE Tuning - Tweaking start conditions

[Zmit]

Hi all,
I've just started looking into tuning with my G4+. This is my first shot at tuning and the car will almost exclusively be driven on the street. I've been slowly working my way around various functions and getting a feel for what they do and how my car responds. I've been gradually going through some of the HP Academy webinars but still feel like quite a noob!

A few details about the car.
- 98RON Petrol exclusively.
- Walbro 255 fuel pump
- CT26/GT35 hybrid (yes I know....)
- 540cc Toyota injectors
- 3SGE (NA) cams
- 1ZZ COPs via an external converter box (Was operating this way on the factory ECU. Will be wired directly eventually)
- High flow CAT

I'm after any advice on the particular setup, but in particular I'm looking at some help trying to narrow down a low rpm hesitation. It feels like I have very little torque around 1000-2500 rpm. This has pretty much always been the case with the factory ECU and the  stupidly oversized/restricted turbo in it, but this I think still comes down to tuning.

There are probably a whole heap of things that need to be changed on the attached map. Some may have been tweaked in isolation.

Also, does anyone know if the 3SGTE factory knock sensor is able to be setup in the PnP G4+?

 

 

low_rpm_stutter_3.llg

MR2_Modified_Tune_v3.pclr

[Adamw]

The first thing I notice in your log is there is significant relationship between lambda and battery voltage.  Every time you have a bump in battery voltage your lambda spikes rich (without a change in inj PW), this could possibly be your dead times are very wrong but I suspect since your battery voltage is also very low you may a have a wiring/ground problem. So I wouldnt trust the Lambda measurement until you get that fixed.  If we could trust your lambda then I would say the fuel table needs some work - CLL is bottomed out at -10% for nearly the whole log.

I notice your trigger offset hasnt been changed from our base map so have you set your base timing?

Your ignition table is very conservative at light loads/cruise so that is going to make it fairly lethargic.  

Your Lambda overlay table should be turned off and the open loop lambda table should be turned on.

Also turn off torque management (I dont think it will be doing anything but turn it off anyhow).

 

[Zmit]

Well that’s definitely a number of things to check! Thanks Adam. I expected there would be quite a few corrections to be made but didn’t even notice the WB/Battery correlation. 

Ill get onto the ground issue first and then revisit the other changes. 

Base timing has been set from the dizzy cam sensor.

[Zmit]

Just updating this slightly.
The SW20 is notorious for dodgy grounds, and the ECU ground on the back of the intake manifold is one that causes a lot of people grief. Turns out, that was loose.
I also took the opportunity to replace the main battery -ve cable as well. The battery level is back to it's usual 13.8V.

I've got a few other things to sort out such as installing the FPR and fuel pressure sensor.

 

[Zmit]

Well i've now sorted out the low rpm hesitation. I've definitely improved the ignition table a bit now.

I'm still having two main issues.

The first is that my starting seems to be rather laborious. it will take at least 5 seconds, sometimes more to actually fire up. It coughs intermittently but will eventually kick enough to run up to idle. I've noticed that trigger issues seem to be common, so i'll be checking that when I get a chance.

The second is the ocassional stall under decel to idle. This happens only when it's been running for 20 minutes+.

Any suggestions on where to look first (and second) for each of these would be welcome.

I'll post the map tonight.

[Adamw]

7 hours ago, Zmit said:

The first is that my starting seems to be rather laborious. it will take at least 5 seconds, sometimes more to actually fire up. It coughs intermittently but will eventually kick enough to run up to idle. I've noticed that trigger issues seem to be common, so i'll be checking that when I get a chance.

Many Toyotas have very low voltage coming out of the cam sensor especially when cold cranking is slow.  Start by doing a trigger scope on a cold cranking engine so we can see if this may be the problem.

 

7 hours ago, Zmit said:

The second is the ocassional stall under decel to idle. This happens only when it's been running for 20 minutes+.

Try to capture this on a log will be helpful.

[Zmit]

4 hours ago, Adamw said:

Many Toyotas have very low voltage coming out of the cam sensor especially when cold cranking is slow.  Start by doing a trigger scope on a cold cranking engine so we can see if this may be the problem.

 

Try to capture this on a log will be helpful.

I heard this exact statement on the HPA seminar on triggers that I was watching this afternoon!
I quickly ran a trigger scope capture while cranking and noticed 0.5v peak on trigger 2 (distributor reluctor signal). There was also only one pulse in the capture time. Seems to indicate either excessive filtering(?) or an intermittent reluctor fault. I have clocked about 4 hours of drive time with this setup so far though..

Is it possible to get a partial log file from a larger log file?
The stalls are in a 16mb file....

The cranking log may not be overly helpful....

cranking.llg

[Adamw]

12 hours ago, Zmit said:

I quickly ran a trigger scope capture while cranking and noticed 0.5v peak on trigger 2

Then your arming threshold table is set too high.  Change it to match my screenshot below.  It would have been nice to post that trigger scope here so we can have a quick look for other problems.

 

12 hours ago, Zmit said:

There was also only one pulse in the capture time.

The trigger scope attempts to capture something close to one engine cycle, the trigger 2 on your engine only outputs one pulse per cycle so that is expected.

 

12 hours ago, Zmit said:

Is it possible to get a partial log file from a larger log file?

Not easily in the native .llg format.  Just share the file via google drive or onedrive etc.

 

 

[Zmit]

52 minutes ago, Adamw said:

Then your arming threshold table is set too high.  Change it to match my screenshot below.  It would have been nice to post that trigger scope here so we can have a quick look for other problems.

Yes... It would have been pretty easy to hit the save button. I dropped the first cell down to 0.3 but didn't change the second cell. I only had a couple of minutes to work with it and didn't even consider changing the second cell. 
Seemed to make no qualitative difference when I tried starting it again, but it's possible that the interpolation at 300rpm pushed it over the threshold.

52 minutes ago, Adamw said:

The trigger scope attempts to capture something close to one engine cycle, the trigger 2 on your engine only outputs one pulse per cycle so that is expected.

That makes sense. I saw one capture with two pulses at the same peak voltage, so that must have been the extremities of the cycle.

 

52 minutes ago, Adamw said:

Not easily in the native .llg format.  Just share the file via google drive or onedrive etc.

Will share the log tonight.

 

[Zmit]

It turns out that the 'save' button for the trigger scope didn't work correctly with the 4K screen....
The second image is just the screen grab. It started part way through.

Link to 3+ hours of logging. The last 1 hour is where I had the most problems. Stalled about four times.
https://drive.google.com/file/d/1mdowJIhanfdc3iK80HFzD2UJ5SpsOupw/view?usp=sharing

[CamB]

If stalling I would try:

- using "idle ignition", so you've got an increase in timing when idle falls below target (there is an HPA webinar on this)

- turning off transient ignition retard unless you are convinced you need it, and at least zero-ing it anywhere near idle

There's a bunch of other stuff in the map (assuming the first one you posted is close enough to what you are using) that I am unfamiliar with but wouldn't necessarily expect (like quite significant ECT ignition trims, etc). These may not actually be affecting anything - its not possible to tell as your log doesnt have enough parameters. I can, however, see that your first map's ignition table doesn't match the logs (although, without the data, can't say why).