1994 Nissan Sentra SE-R going VEMS, then turbo.

[rob@vems.co.uk]

Nice result.

I'll be very interested to see your ignition maps as my SR20DE is nearing completion (I'll update my thread accordingly)  they wont be any use to me as I'll be using AlphaN but the ignition advance at WOT should be valuable.

[BenFenner]

I get to use my brother's car for comparison, as he's got the same exact car as mine, but completely stock.
Here's a video of my car (with VEMS) versus my brother's car (stock).

http://www.youtube.com/watch?v=9b0_HRtBnpM


I've finished the WOT tuning and will be posting maps first, then the rest of the tune. Rob, look for that timing map very soon.

[BenFenner]

Here you go Rob:


Ignition map:


Target lambda map:


VE map:



Keep in mind the 70 kPa line on the lambda target table should read 0.94 across the entire map and the VE table should be changed accordingly (lower values). Right now the actual lambda values at 70 kPa are about 0.94 which is where they should be, but the maps are sort of mismatched if you know what I mean.

Edit: Looking at the 50 kPa line on that VE table makes me feel like I've left work to do. =/

[rob@vems.co.uk]

Thanks - what was your req fuel value?

[BenFenner]

Req fuel is 12.5. Injector size is... *shrug* I'll have to look it up. Stock USDM SR20DE size.

Edit: 259 cc injectors

[rob@vems.co.uk]

Cool, we have both got the same req_fuel then, my maps are going to be AlphaN, but even so - mine are waaaaay wrong at the moment.

[BenFenner]

New smoother VE map looks like this:

[BenFenner]

Okay it's been a long, long time since I've posted anything here. I've been doing a ton of mechanical upgrades and really haven't messed with the ECU side of the car in a while. I've messed with idle and cold start until I was blue in the face, but that's another story.

I still have my entire annotated N/A configuration that I'd like to post up at some point, and I still plan on that. I just need to finish annotating it.

For now, the car finally got the turbo installed. It got the EGT installed, and a new open element intake temp sensor to replace the closed element unit.

I put some safe timing and fueling values in the maps and tested the car out and hit the 110 kPa boost limit I had. Increased that to 160 kPa (wastegate is set at 150 kPa) and noticed that pulls in 3rd gear caused the computer to die. I lose tach signal, the LCD goes blank and the ECU disconnects from the laptop then restarts, reconnects to the laptop but I never get tach signal back. The car runs like crap until the ECU is reset by turning off the car, and turning it back on. This behavior started happening at idle shortly after that.

After more checking I found that I was getting boost creep over 160 kPa so I increased that limit to 180 kPa and then after seeing 178 kPa in the logs I brought it up to 186 kPa. I did this because while sometimes the ECU would die, other times the logs indicated what looked like a simple fuel cut.

I decided to replace my warm spark plugs with proper cold plugs, and gap them properly for the turbo setup.

After changing the spark plugs and adjusting the boost limit the situation has gotten better and better.
I took it out last night to do some actually tuning and managed to get a few runs in.

First pull in 3rd went well enough, some bogging but no ECU issues.
Increased timing quite a bit back to sane values trying to keep boost creep under control and EGTs down.
Did another pull in 3rd with less bogging and no ECU issues.
Confidence is high so we go for another pull. Before doing that the ECU crashes at idle.
We make the 3rd pull of the night and hit what feels to be a fuel cut, might have been ECU crash, can't remember. Check logs and highest pressure reached was 171kPa. One instant the logs look good and rpm is 6,000 something. Next log entry rpm is 3,000 something and everything else in the logs is still good. Next log entry everything is zero'd out.
We pack it up and drive home for the night.
The ECU dies/crashes maybe twice trying to park the car in the driveway.



Right now I'm on firmware 1.1.18 still. Is there something newer I should be on?
I'm tuning with the same version and configuration of MegaTune that I started with over a year ago. There is newer stuff I'd like to be using right? Where can I download that?

I'm going to disconnect the EGT probe from the ECU and continue diagnosing the problem.

Any other ideas? Sorry I haven't posted logs and such to go along with it. Let me know what you'd like to see or tests I should/could do and I'll get all over it.

I have a track session (my first!) coming up towards the end of March and I've been scrambling the last couple of months getting the car ready. I'd hate to not make it because of this. =(

[gunni]

Check that your dwell timing isn´t to much and isn´t becoming to much, and check your Vbatt.
As the dwell goes higher, vbatt may go down, that will cause more dwell, and then more the vbatt will go down.
Until you either fry a IGBT or the ecu/spark breaks up.

Happened to me when I was trying out 1.1.44 or something after upgrading from 1.0.79
There where new dwell comp calculations so that was messing with things for me. Changed it and revved easy from there on.

[rob@vems.co.uk]

1.1.18 has been in service since god was a boy and at the moment I'd not see any reason to change it.

When you say that the ECU crashes does that mean you're you getting trigger errors?  Or is there something else thats happening?

[BenFenner]

gunni, I currently have 3ms dwell at 6 volts and 2ms dwell at 14 volts. These settings have been fine for over a year N/A. I forgot to mention I tried 2.5ms and then 3ms dwell at 14 volts with no noticeable difference in this behavior.
From what I know (and that's not much) about these Honda CBR 600 coil packs, they are happy at 2ms dwell, at least for N/A, and they are capable to much more than twice the power I'm making.
I'm certainly up for suggestions. I haven't tried messing with the dwell with the new plugs I don't think. Should i try more dwell? Sounds like you want less dwell? I guess I should go look at my logs, but from what I remember dwell just stays rock steady at 2ms.


Rob, when I say the ECU dies, this is what happens.
The tach on my dash goes to 0.
The LCD screen goes dark.
The engine stops getting fuel and spark.
Every logged item goes to "0" if I'm logging.

This happens for a split second and then the ECU turns back on.
Sometimes the tach on my dash recovers, most times not.
Sometimes the wideband readings recover, most times not (have to wait 60 seconds for them to come back).
Sometimes the car will run well afterward without restarting the ECU, most times not.


I forgot to mention I went over my harness adapter and made sure no pins were loose, no wires were touching, everything is insulated very well and I tightened down the ECU plug very tight and put zip ties around the ends of it to hold it in as I've heard from Rob this can be a problem.

There is a possibility I harmed the engine harness while installing the oil feed line for the turbo and I'm going to look into this. But I'd love other ideas as I'm not hopeful on that.

[rob@vems.co.uk]

it does seem that the thing is resetting - please datalog and see if its a trigger error or a reset
reset might suggest voltage drop out or short circuit.

[BenFenner]

I'll post up a log ASAP.

[BenFenner]

No log, sorry.

I took time to work on the car last night. Painted the exhaust and while I was under there I looked to see if anything was melted from when I installed the turbo oil feed line (it shorted with the starter and melted the braided steel, etc. and I had to replace that). While holding the hot line away from the starter wire, I thought maybe I melted part of the engine harness. Nope. All was fine back there.

The car was cooperating, so I took it out to tune some. Car started acting up again. In a last ditch effort to track down the problem, I unplugged the EGT probe since it was really the only thing about the set-up that had changed as far as the ECU is concerned.

The car cooperated all night after that.

I'm going to keep driving it this weekend and see how it goes, but I have high hopes. I'll report back later.

[rob@vems.co.uk]

Hmmm, does the EGT still give proper readings?  Can you try a dvm voltage reading between the block and the probe?  Might need to run the engine and see if theres a problem at higher speeds.