eng92

iTrader: (3)

We had some nice dry, ice-free roads this weekend, so I had a chance to do a few WOT runs to get some timing and air flow data.
The first graph below compares the WOT ignition timing advance from my stock 2k2 ecu against some 4th gen data I logged last winter.



My base timing is 15 degrees. As you can see, the fourth gen timing is actually higher except in the midrange. You 3.5 swappers with 3.0 ecus are not as hard up as you may think. In case you are wondering, my knock sensor is working fine.

The second graph below would be of more interest to 5.5 gen people without the L-spec ecu.



It quantifies the throttle plate closure that occurs gradually throughout the rev range. It closes approximately 10% (from 91 to 81%) starting at whatever rpm you first mash the gas (1500 rpm in my case)

The airflow data in black shows how the engine with stock IM chokes after 5.5K. This is really noticable in the higher gears. I had to cut a few of my fourth gear runs short, at around 6K, because I was running out of road.

NmexMAX

iTrader: (22)

Very good stuff.


Between the throttle closing and the stock IM flow after 5.5k.. We can definately get more high rpm power with a few tweaks. The good thing is tht the potential is there, but now we need to figure out how to free it.


It be nice to see if the TS ECU (L-spec) actually does help with the throttle closing.

DandyMax

iTrader: (7)

Awesome info Dave. Someone should have done this a long time ago.

Nealoc187

iTrader: (23)

OK I have not kept up much (at all) with the drive-by-wire threads. Are you saying that the e-gas drive-by-wire system physically closes the throttle plate a bit over time even though you have your foot on the floor the entire time?

Jime

iTrader: (10)

eng92 what are you using to get the data, OBDII?

I had the L-Spec and did before and after comparisons on my OBDII reader and there was no change.

I have read that the OBDII does not give a true reading on the throttle opening for the DBW in a thread about the L-Spec on the Z forum.

Neal, yes the DBW totally controls the throttle plate and does it own thing. That why I am glad that I will have a cable driven 3.5 this year. At least when I put the pedal down I know it will be at 100% vs what the computer wants to set it to.

eng92

iTrader: (3)

The info in the graphs above did come from OBD data logging.

I will try putting my voltmeter directly on the TPS terminals to see how that changes with RPM at WOT.

Jime - Did the thread in the Z forum say how they verified throttle position? It is a little difficult to visually verify. I suppose if you could remove the cover off the throttle motor and run the car on a dyno.

I will be picking up an EU soon so I will be able to log a lot of these parameters directly rather that rely upon the "not so speedy" OBD port.

I discovered this weekend that although you can enable the fast sampling on the Auterra unit with the 5.5 gen ecu, the data you get is very difficult to work with, especially if you are logging more than 2 quantities. The logged data would have maybe 2 or 3 consecutive rpm readings, followed by a maf value and then a TPS opening.
It makes it very difficult to interpolate the rpm that your other logged parameters are occuring at
The OBD port is good for what its intended use is, diagnostics, but I certainly would not want to tune my engine with such a slow datastream.

Jime

iTrader: (10)

I can remember how they verified it, like you say it would be difficult without a dyno for sure.

The Conzult interface is much faster at data logging and for $100 is a great deal as opposed to the OBDII and it allows you to read the eprom, change timing, turn off fans etc plus the functionality of the OBDII.

This is a datalog I just did a day or so ago which shows the number of samples per sec.

Time, RPM, Water Temp F, Speed mph, Throttle, Air Temp F, Timing, EGR sol
11:15:19 AM, 4938, 172.4, 28.8, 4.14, 30.2, 14, On
11:15:20 AM, 5025, 172.4, 28.8, 4.14, 30.2, 15, On
11:15:20 AM, 5062, 172.4, 28.8, 4.14, 30.2, 16, On
11:15:20 AM, 5100, 172.4, 28.8, 4.14, 30.2, 16, On
11:15:20 AM, 5125, 172.4, 28.8, 4.14, 30.2, 16, On
11:15:20 AM, 5162, 172.4, 28.8, 4.14, 30.2, 16, On
11:15:20 AM, 5188, 172.4, 28.8, 4.14, 30.2, 17, On
11:15:20 AM, 5238, 172.4, 28.8, 4.14, 30.2, 17, On
11:15:20 AM, 5288, 172.4, 28.8, 4.14, 30.2, 17, On
11:15:20 AM, 5338, 172.4, 28.8, 4.14, 30.2, 17, On
11:15:20 AM, 5375, 172.4, 35, 4.14, 30.2, 17, On
11:15:20 AM, 5400, 172.4, 35, 4.14, 30.2, 17, On
11:15:20 AM, 5475, 172.4, 35, 4.14, 30.2, 17, On
11:15:20 AM, 5500, 172.4, 35, 4.14, 30.2, 17, On
11:15:20 AM, 5538, 172.4, 35, 4.14, 30.2, 17, On
11:15:20 AM, 5600, 172.4, 35, 4.14, 30.2, 17, On
11:15:20 AM, 5638, 172.4, 35, 4.14, 30.2, 17, On
11:15:20 AM, 5675, 172.4, 35, 4.14, 30.2, 17, On
11:15:20 AM, 5700, 172.4, 35, 4.14, 30.2, 17, On
11:15:20 AM, 5750, 170.6, 35, 4.14, 30.2, 17, On
11:15:20 AM, 5762, 170.6, 35, 4.14, 30.2, 17, On
11:15:20 AM, 5800, 170.6, 35, 4.14, 30.2, 18, On
11:15:20 AM, 5862, 170.6, 35, 4.14, 28.4, 18, On
11:15:21 AM, 5875, 170.6, 35, 4.14, 28.4, 18, On

eng92

iTrader: (3)

Jime - I count 22 samples of each data set per second.
However if you look at the data, only the rpms are changing with each data set. Your rpms increase by about 900 rpms and there are only 2 different speed measurements.

I see the same thing on the Auterra. Your interface can sample as fast as it wants, but if the ecu is not providing the data at that rate, then the values do not change. So really you may have 22 samples taken a second, but only 2 of them provide useful data.

EDIT:
I do not know about the 4th gen Consult interface, but the Consult II on the 5.5 gen uses the same connector, and data line, as the OBD-II port. Because of this, I would expect the ecu data rate to be the same regardless of what interface device you plug into the connector.

DandyMax

iTrader: (7)

His timing data changes more frequently than the others though, which is a good thing. Perhaps the ECU has specific (and different) sample rates per parameter. It would make sense as some things don't need to be sampled so quickly... coolant temp for example wouldn't change much in one second.

But yeah unfortunately Jim I don't think the Consult interface we're using on the 4th gen will work on the 5.5 gen. Nissan changed the protocol (Consult I to Consult II) in 2000 if IIRC.

Jime

iTrader: (10)

Ya I noticed that after I posted it, RPM is updated but not the rest although timing did seem to be ok after looking at the rest of the log. The MPH sure is slow.

Dan thats why I like the 3.5 with the 4th Gen ECU. Hopefully get all the power with the simpler electronics. I do understand there is a consult2 gizmo in the works though so that would be a bonus for the 5 Gener's.

Do either of you guys know what the sample rate is like on the EU? From a couple of the graphs I have seen it appears as though we are going to be much better off with it. I just wish it had more inputs for temp, press etc I would like to use it for the auto trans monitoring as well.

Dan with the ECU dump do you think its possible to do the EPROM swap and program our own?

DandyMax

iTrader: (7)

Yes I was told they are working on the Consult 2 version which is great. I'm supposed to be on the contact list for when it comes out...

Jim the sampling on the EU is dependent on whether you have the laptop plugged in. Without the laptop it'll do 8 parameters at once max, but with the laptop it's unlimited. The smallest sample interval on the datalogger seems to be 20ms from what I can gather (ie 50 samples per second).

As far as inputs, yeah it would be nice to have more but for a piggyback it is actually pretty good. To that end one thing I've thought of doing though is putting a simple switch on a couple of sensors to flip back and forth effectively increasing the number of loggable inputs (although not all at the same time, it would at least give me more possibilities).

The EPROM stuff... yeah in theory it would be possible, but I think you'd have to get someone who's really capable with hexidecimal computer programming and that could locate the relevant fuel and timing maps etc out of the downloaded BIN file using a hex editor etc. It's not that easy unless you have the knowledge to do it and can get quite tedious. Beyond those 2 maps a lot of the problem is trying to figure out what the various memory addresses are in terms of the real world values and which parameters they correspond to also. Many of the hex numbers in the ECU are scaled somehow, so it's not as simple as just looking for the number you are reading on the scanner. It would also take some on-board modifications similar to what JWT and TS do with their daughterboards. But in theory it is possible if you had the know how (not many do).

I30tMikeD

iTrader: (8)

good info!

I am new to data logging. Just did my first logs through the OBDII port with my pocketlogger software and PDA.

The best I am able to get is 6 s/sec, and the way I understand it is that you only get the full 6 s/sec if your only logging one parameter. Each additional parameter cuts the s/sec in half.

One question. How much do your logs vary from run to run? Specifically in regards to timing. I did some timing logs using two different variables, one with SAFC correctoins and one without. But only was able to get one log per variable. I know you would want to take an average of a few differnt logs, but I was just curious if my info with just one log per variable was valid at all. Will go out this weekend and to 3-4 per variable to get more accurate data but was just wondering if the data I got yesterday is even worth looking at without taking an average of a few logs.

Also, when I logged Air Flow I thought I would get a voltage parameter or something that correlated to our 5v MAF sensor but what I got was a number that started at 5 and went to 22

eng92

iTrader: (3)

Evidently your PDA-based pocketlogger is very similar to the Auterra in the way it obtains the data.

Because of the slow sampling, you will always want to log the minimum number of parameters during each run. Say rpm and timing one run and rpm and maf the next. I would consider 3 runs an absolute minimum and then average the data at fixed rpm intervals. You really need to look at the data to see how much scatter there is though. Typically, the more scatter you have, the more samples you need to come up with a meaningful average.

Runs for timing should all be done under similar conditions to yield comparable results. Significant changes in air intake temperature will change your maf which will affect timing to a small degree. For back to back WOT runs, I generally do not see more than a 1 degree variation in timing for a given rpm.

From your maf numbers I would say the units are [lb./min]

I30tMikeD

iTrader: (8)

Thanks.

What I am trying to compare is Timing with SAFC corrections (pulling fuel) and timing without SAFC corrections. From what I know pulling fuel via a MAF hack like an SAFC can indirectly add timing. The % of fuel being pulled and how it effects timing is unknown and that was what I was trying to figure out.

I run 550cc injectors, Z32 MAF, SAFC-II, and stock ECU. I have to pull up to 15% fuel in much of the rpm band inorder to get even a high 10:1 AFR. What I have always been concerned about was that I was indirectly adding timing where I was pulling alot of fuel. Not good on 10lbs of boost on stock timing.

But my two timing logs tell me different. I did one log with SAFC corrections and one without, expecting to see added timing with the SAFC corrections log. But it was the exact opposite. I got 2-3 degrees less advance with the SAFC corrections. Which is actually a great thing, just don't know if it was accurate with only one log on each variable and why I was seeing less timing not more. Both runs were back to back in 4th gear.