Hilton

Get outta the cave!

PorKen

May have to stay in the grotto a bit longer...yesss...


Although the code which calculates the rpm scales for idle/knock/cruise tables still baffles me - IE. I can't pick my own EZ columns yet - I may be able to expand the WOT table to 64 cols.

The scales are calculated as 64 (OK 60, but who's counting) wide. When reading from a 16 column table, the scale is divided by 4.

I would then add the WOT value to a 'sloped' WOT cruise table so it could still auto adjust somewhat for load.


Since most of the knocks don't happen neatly at the available column rpm, having 4x the resolution should allow for a higher average advance.

PorKen

After retiming (knock logging) the entire WOT map, for the third time, I noticed I had divided (right-shifted) the load variable too many times.
I am going to have to redo it. Again. Another {tank of super} bites the dust.


10 bytes was all the code necessary to convert to a load-compensating 64 col WOT map.
Fit right in place of the original WOT read code with room to spare.

In theory, wherever I time the map, the load comp. will modify it ± for different gears, terrain, weight, etc.
Load can modify the values up to 10° right now (max could be 20) - higher rpms are affected more as the load is higher.
Note there is more advance towards the high rpm end of the map because of the load modifier.


After making this map, I have a better understanding of the scaling maps.
I hope to be able to dump the primary cols 960 & 1500 and make new ones where they will be more useful.
In the 2700-3400 range, peak TQ, for example.


Red dots are the old rpm columns. Most, if not all, are higher than before.
Green ones are much, much, higher because they no longer have to fix nearby knocks. 3600 used to be in the sub-basement!
(~3600 is where the flappy opens, IE. a transition phase between different harmonic pressures on either side.)

There are some new peak advance points which should reflect well on the dyno.
6100 is sweet - max HP is made in this area.


The low points are 'knocky'. (Map below will continue to change, so it's not gospel, yet.)
I suppose this could be seen as a pressure map - highest cylinder pressure at the valleys.
Some areas may just have bad swirl or something, though.

FredR

Ken,Interesting that you have that dip at around 5600 rpm- that was where I had to pull back timing albeit my motor seemed to have more data points in the 93% load range rather than the 100% array. Maybe that is the point it transitions into the top rpm range that covers the last 1k rpm range. That was with 24.8 degrees of advance. For the top rpm cell range I set 23.8 degrees and that was OK but I was expecting to see 27 degrees.

Having fitted my breather mods and having some octane booster available to me, I will go back to sharktuner and see if I can get a bit more advance- to stock levels at least.

Regards

Fred

simos

BTW, can someone explain me why the flappy is closed after 5600 rpm or so? I have always tought that large plenum works best with high rpm area..
Large plenum does have bigger volume and therefore it's able to store more air which is needed when engine is reaching it's highest hp point.

That may be the reason why spacers will help little at top hp area, they will increas the plenum volumes and therefore helps there a bit. The intake path from throttle plate to plenum does have two tight curves and the restriction may be enough to prevent air flowing when the demand is highest.

The little timing dip after 5500 could be caused by flappy closing point. The air flow and resonance waves may be distrubted there giving temporary high pressure peak on both plenums.

Ken, have you ever tried to run with flappy open from around 3500 to over 6100 rpm?

Theoretical nonsense, but sometimes theory meets practise

PorKen

If you compare the cat and non-cat WOT maps, you'll see the non has lower advance.
The cat keeps more exhaust in the cylinders at higher rpm so less to burn, requiring/allowing more advance.

Better airflow generally requires less advance.
The highest airflow regions, IE the peak of the TQ humps, will have the lowest amount of advance.

Maybe your fueling is not consistent?
Try experimenting with the injector open time. EG. increase the open time by 10% at a time.
This will richen the AFR so reduce the injector size to bring it back to your target.


Octane boost in the wee cans don't do much, if at all. Bulk Toulene or something, maybe.


Not sure why it is closed again.
Closed or open, it makes no difference in MAF voltage or knocks except for the first torque hump.

Changing the size of the plenums would only change the rpm range - which may take it out of the sweet spot for the cams.

The flappy does not change airflow. It reflects pressure waves back down the intake for a ram effect.
With the size of the S4 plenums and the cam profile, the resonance effect only does anything from ~2600 to ~3600 rpm.


I figure the GT intake cam was advanced (vs the S3) to work with the flappy. (Tighter LSA.)

PorKen

After a lot of fiddling...SLC-X, injector advance time, new ignition map, etc...

Fueling is so consistent/flat now that I would say the WOT map could a good indicator of what the torque graph may look like on the dyno.


Flappy closed peak at 3200.

Flappy open peaks at 4400, 5700.

HP peak around 6100?


300 rwtq sae is going to be somewhere between 49 and 64...hopefully closer to 49...56?

PorKen

Not enough advance, darn it. Wasn't quite sure about the adaptive load so I failed.
Good data, though. (Cold comfort.)


The blue line is the from the first two runs, the red is from the second.
The red one has a touch more fuel and +2° advance overall. +10 HP!

The green line is from last month. At ~5900 the red one should have at least matched it. Same story at 2700.
By the lack of knocks on the log, it could have handled another +2° advance overall.


I figure 5600 should rightly be just over 300. Which would put 6200...


(STD), (UNC)

Hilton

Very nice result for injectors/chips/x-pipe!

Which injectors are you using? (got part no's?)

I may as well order a set in prep for my '87

PorKen

It's OK, but it could have been sweet, if only I had looked at my logs - before I was unhooked from the roller.

One click more fuel overall, 2° more advance...

I have a theory that the current placement of my WBO2, in the X of the X-pipe, may make it show richer than it actually is farther up.
Fuel is afterburning on the way, using up O2, showing as rich?

Anyhu, I'm going to richen it up, start logging again from the advance map above.
I could tell from my logs that over 5600 was getting slower as I cranked back the advance, but I didn't 'listen' to myself.

At least from this dyno I know that the fuel chip is essentially done.


I'm using the metal body 4-hole Fords. Don't know the part number off hand.

PorKen

I'm an idiot. I have been (stubbornly) using the open airbox lid this whole time, even though I already logged no loss in airflow with the (S4) lid on.

Put the stock lid and air tubes back on and the logging runs are now consistent.
Fueling is back to back steady with no knocks at +2° over the last dyno.
+2.5° is probably optimum, with just the occasional light knock.

Hopefully redyno next week, but essentially the manual version 'beta/release candidate' is golden.

I'll start cleaning up the code in the mean time.



The air intake temp sensor is a real advantage for the S3s!
Tuning for knock sensor feedback is not nearly as effective as having a predictive tune based on air temp and vacuum.

BC

Knock sensors, even today are still used for when all other tunable parameters are not coming through properly or there is a fuel anomaly, and the engine must not destruct itself. It is simply too much of a "reaction" instead of a "planned point" type of event in the tune.

FredR

Ken,

On post No 232 what are the Y axis values- surely not advance degrees? Whatever it is you appear to be relating torque to those numbers- apologies if I am being a dumb *** here.

As a matter of interest what does your seat of the pants experience tell you with respect to S4 cams versus S3 cams working on the 32 valve S4 motor?

Regards

Fred

FredR

Ken,

Rather illuminating- I have viewed the fact that I was having to pull top end timing back a bit a flaw but perhaps what you are saying is that it may be because I have actually achieved better specific airflow with the various mods [EIS system, GTS inlet cams, Louie X-pipe, plenum spacers etc]? Put anaother way, if one is sharktuning from a stock position, you dial in as much advance as you can until the onset of knock for a given point, then as you modify things, if you achieve improved air flow [the immediate objective] all things being equal you can expect to see advance needing to be pulled back?

Keep up this great/fascinating work

Regards

Fred

PorKen

Those are LH WOT map numbers.

They may seem high compared to what you might see in the ST WOT map because I have made a separate injector constant (one of the inj size vars) for WOT.

I have set the WOT injector constant to a much lower number (IE huge injector) than cruise/idle so that '0' on the WOT map is extremely lean.

Having a very low floor is one way I can control the AFR (inshallah) over 6000 rpm. Sometimes it's a bit random, but it always goes upwards/lean.

As you've probably seen, with stock maps/code it gets uncontrollably rich at high rpms which makes HP drop off much faster than it should.


S3s have max torque at a low 2700 rpm, so they pull hard right away, then there's a bit of a lull through the 3500s, followed by a rush of power all the way to the rev limit (my chips).


Tuned, the S4 can feel a bit 'lumpy' with the flappy operation combined with the peakiness of the S4 cams (and knock retard).

In good air, the peaks are blended somewhat so it pulls strongly all the way through the rev range (my chips).