The small fix above may help a little below 5400, but it can't/doesn't level out the big spikes much.
You may find your AFRs are a bit more consistent, though.


The smoothing code which flattens it out, what I am making for my S4 chips, is up to 70 bytes and counting (which is a lot in hex coding).

I am also doing other things like disabling the IAC/ICV output at WOT, in hopes of giving more/steadier voltage to the injectors.
(IAC and injectors share the same 12V circuit.)


You kids and your high level languages.

 

Ken,

As I understand, these issues you are very cleverly weeding out are probably caused by the lack of computer power relative to what is possible today. Higher sampling frequencies to give more consistent results etc. If this is indeed the case, how difficult would it be to use the same wiring to the LH unit and replace the internals with something more powerful/easy to repair/replace as a bolt in replacement without the problems adapting other systems? Maybe even enable sequential injecton capabilities?

Trust the above might be relevant to the thread development?

Rgds

Fred

 

Just wondering... if this change only has positive implications and no negative ones, why wouldn't the factory ship the LH modules programmed this way?

Dan
'91 928GT S/C 475hp/460lb.ft

 

I wonder more why the factory made only a passing effort to tune the engine above ~200 CHP (324 g/sec)?

 

The "if then..goto " has been replaced by "if then Maybe" and the new "Come From" statements.

jp 83 Euro S, with the simple logic "if air is moving, Inject"

 

Tunerpro patch for MAF max rate added in the second post.


For each cycle through the code (many times a second), MAF values are incremented or decremented at a steady rate.

The hope is that sudden high or low values from the A/D chip won't last long enough for the additions or subtractions to catch up.

 

Great stuff Ken. Many of us wish to ditch the ECU all together, and will continue to but that is still a small fraction of the population, which are mostly S4s, and that are unwilling to do so. Your code changes can make a big difference.

 

Note whereas I have been using g/sec, grams/second, habitually throughout this thread, I should be saying something like MAF units or '-ish', as the MAF scaling numbers used in the code don't quite match actual (K)grams per second airflow.

(G/sec sounds cool. )

 

How about hay bales per fortnight as a working unit? Much more agricultural.

Rgds

Fred

 

Hi Ken,

Would you please be able to offer a comment to my post at #7?

Thanks

Dave

 

That's a tall order.

Yes. ('88 EPROM image or ST file. Different code addresses for '87 and '89-up.)

Maybe. (Likely no ill effect.)

No. (It's probably not going to feel any different without some minor tuning to take advantage of the smoothing.)


With some fuel changes and more ignition timing, it should be noticeably stronger, though.
So far, I have been able to dial in a lot more timing at WOT around max HP with a more predictable AFR.
And, when I advance the timing, the AFR goes lean, indicating it is using that timing. Before, I couldn't tell what was going on.

 

Has anyone with a STer tried this modification? Does the STer preserve this change in the code? or will it get reverted when a new image is produced?

John, Jim?

I'm really looking forward to trying this on the dyno.

 

The ST software reads the BIN file as-is and will write/read it to/from the disk or device w/o changes.

The patch code is in empty space, AFAIK.



The LH and EZ only reads the upper 16K of a 32K byte file.
The ST writes to both 0-3FFF and 4000-7FFF, 16K banks.
IIRC, it reads the lower bank into the ST software.
(The ST LH BIN is based on a '88 EPROM.)

 

FWIW, I would say the stock LH is way oversampling the MAF signal at WOT.

At least for my S4, the plus rate can go all the way down to 1, becoming more stable as it goes down. The minus rate may still be too aggressive at -1.
I theorize that the little lean spikes are being uncovered by smaller rich spikes, so I will be putting in a counter to only subtract every other time (or less)...on my chip.

I think there could be very small spikes (from reversion?) that are undetectable by the WBO2 which may contribute to individual cylinder knocking.


I don't think this will raise peak HP, but it should make a really consistent 310+ (@~6100).
On the dyno, the red line, injector duty, has fairly matched the hp curve in the past, if you turn your head the right way.

It certainly feels nice to drive. A constant thrust up to the limit in each gear.