Speedtoys

Pass #3 on the ford 19lb Design2 injectors.

A lot smoother....some eye candy after a couple hours of the daily drive to work/customers/home again over the mountain.

Rob Edwards

Nice, looks like a happy map! Are you using the default AFR target map in SP?

Looks like you stomped on it once or twice to get those WOT off-idle points in the 196/228/260 cells...

Speedtoys

Yes, the built in SP map is in place. Tomorrow AM's LH load is only changing most cells, 1-2%..need to beat on the top end more..especially a # of dead stop Go runs from the shoulders of various highways tomorrow.


I added 3, and then 2 more for a total of 5d of timing advance everywhere over 3k, and ive had -1- knock so far. I added a total of 3 from 1400 to 3k, and zero knocks so far. Still wondering what my timing goals should be..for now, just exploring the system and learning.

Rob Edwards

I wish I knew what constituted an optimal timing map too. It's probably unique to each car, dependent on the fuel, weather conditions, and the way the driver romps on it. 1 knock in a lengthy datalog means there's more timing on the table, for the conditions under which the datalog was recorded. Might not be anything left on the table when it's 10 degrees warmer, though, or when you're blasting uphill for long periods of time at altitude.

Any issues with stable idle when cold?

Speedtoys

Idle...ehh..Ive shifting timing in the idle map a bit..smoothed it out some, but im just hunting for 'whats right' in the warmup tables.

Tomorrow sposed to be a hot one..I'll add 2 more, and leave the pullback at 9d/10secs, and see what i can beat out of it.

The initial 3 I added from 3k up, certainly _felt_ like a winner..


Not sure at all, W/O dyno, id have any idea what to do with the flappy..just for giggles.

blown 87

While I have done zero tuning on a 928, I feel that if you start messing with the ignition curve, you better be monitoring the knocks at all times.

Different fuels, temps, loads and who knows what else can cause knocks.
And from what GB says, knocks will kill the bottom end of a 928 just about faster than any thing I have ever dealt with due to the soft bearings.

We need a way to switch between maps with a simple change, and this may be possible with the Carputer with a full time wide band and sharktuner.
We will see.

Speedtoys

Even dead stock, I could pull -1- knock outta the car in a couple hours of beat down...so at this point, I dont believe ive gone TOO far.

But going too far, isnt on the list.

Bill Ball

Unless this was only 1st or 2nd gear, you were a very naughty boy!

Speedtoys

Unsure what you're talking about officer.


But actually Bill, in the new version of SPlotter that Jim is working on, that zone is shown in this manner in 3rd gear and up.





..apparently, I went plaid!

Landseer

Looks like some intense storm cells in the north. Be safe out there.

jcorenman

Oh Nooo, not PLAID!!

Sweet looking plot!!

If you are having any issues with idle, take a look at the idle map for the EZK (ignition maps page). The factory map has some pretty big changes in the idle advance, from 19 deg at 600-ish down to 5-deg at 940. The goal apparently was to help stabilize idle, but whoever came up with that map wasn't talking to the guy in the next cube working on the LH, which has a whole fancy ISV to stabilize idle. Two computers doing the same job tend to argue.

Change the ignition idle map to be fairly flat, e.g. 12-13 deg at the lowest RPM's to 10-deg to 1700 or so, then meet the rising advance curve from there up. (That part is only used on trailing throttle in any event).

And also make sure that your fuel map is pretty flat in the idle range, no big changes in fueling as the idle wanders up and down.

Regarding knock-retards, that's a whole topic by itself. The knock-sensors are pretty sensitive, picking things up long before audible detonation. The factory spec was 50 per 10,000 cycles (20-30 sec at 5000-6000 rpm), they figured some knock-retards were OK.

But Greg is right, there are a lot of variables. What works at 1000 or 2000 feet won't be acceptable where we are, at sea level. It is quite easy, however, to wire a switch to the coding plug and select from three different ignition maps-- normal (i.e. conservative), aggressive, and meek (the Aussie 91-octane map, in case we want to run low-octane fuel). It's a little toggle switch under the ashtray lid, along with the "countermeasures".

Cheers, Jim

SteveG

Someone please tell us newbys how to read that display. I just downloaded the manual and I'm playing catch up.

Rob Edwards

Sounds like a job for button assignments on Hans' console. Kind of like a Manettino for octane...

Mongo

I have seriously got to GET ONE OF THOSE!!!

jcorenman

That is a plot from Sharkplotter, my companion software for Sharktuner, for plotting data from Sharktuner log files and tweaking maps.

This is the main plot, showing all of the data that Jeff collected on his drives to/from work-- a couple hours worth of engine running. The beauty of having that much data is that it can cover a wide range of engine conditions, the problem is how to analyze a thousand pages of data.

The horizontal axis is RPM, from zero on the left to near-redline on the right. The vertical axis is load, the signal from the MAF that indicates engine load, based on mass air flow. No load (i.e. no air flow) is at the top, max load is at the bottom.

Each colored dot is one data-point from the WBO2 sensor, plotted in a color to indicate the air/fuel ratio. The scale in the upper-left shows the code: rich (AFR=12:1 or less) is yellow, stoichiometric (14.7:1) is blue/green, and lean is purple to red.

So the curved upper-left corner of the plotted area represent idling. As idle speed increases, the ISV closes and load drops; conversely, as idle speed drops then the ISV opens and load increases-- hence the curve.

As the throttle opens, the first thing that happens is increased load-- the points below the idle area-- and then quickly-increasing RPM's. Off-idle, because the engine is turning slowly, load (per cycle) can be very high.

The big middle-area is cruising, 1500-3000 RPM and part throttle. Small dots (which all run together) are closed-loop operation, the NBO2 sensor is regulating fuel by fiddling with O2-adjust (which is also logged, and used).

And the area to the lower-right- higher RPM and higher load-- is where things "go plaid" (Mell Brooks' "Spaceballs", if you haven't seen it). The throttle is wide open (large dots) and rev's are climbing quickly (dots are spaced apart) so it must be first gear, officer.

Looking at fueling, it looks pretty sweet. The indicated AFR (i.e. color) is the WBO2 reading, plus/minus whatever O2-adjsut the LH was using to get there. There is a label that follows the cursor, which you can see here, for the indicated map-cell the average AFR was 14.7, standard deviation of all the points was 0.7 AFR units (i.e. 2/3 of the points were within 0.7 AFR of 14.7), and there were 1051 samples in that cell.

In the higher-load areas, towards the bottom of the plot both left and right, the mixture is more green/yellow than blue/green, because the LH is open-loop and the map is set for 12.5 or 13:1 AFR at higher loads. That is SP's "target map", set it up how you want it.

And along the top is closed-throttle deceleration, where the LH normally cuts off fuel. The Sharktuner provides an option to disable the cutoff, but in any event you want that top edge to be pretty lean.

There are a few gray dots, those are data samples that Sharkplotter is ignoring because of rapid changes in RPM or load (data may not be valid).

And I don't see any knock-retards, those would be "x"'s on the plot, indicating the load and RPM where the EZK detected a small knock and took out some advance. The size of the "X" depends on how cylinders retarded, by how many degrees-- making it easy to spot problem cells with too much advance.

The Sharkplotter manual is can be downloaded from www.sharkplotter.com and you can follow links from there to JDSPorsche's website, or to my Sharktuner page, and download the Sharktuner manual.

Cheers, Jim