Custom operating system idea

[wait4me]

Instead of having a added boosted ve table that is full of wasted cells for most of everyone,
how about a Super High resolution Ve table that they can use for speed density or really high resolution for 10lbs of boost down ect..
like on the regular ve table make it so it is in units of 5 kpa and 50 rpm ect...

That way when we are tuning auto ve, it would make it even that much more accurate. Not everyone needs 285kpa :) And we wont have to deal with spikes and large areas that need to be modified and blended on our own.. It is hard to tell what cells value should actually be when it has to INTERPOLATE between its surrounding cells.

:eek:

[QUICKSILVER2002]

Instead of having a added boosted ve table that is full of wasted cells for most of everyone,
how about a Super High resolution Ve table that they can use for speed density or really high resolution for 10lbs of boost down ect..
like on the regular ve table make it so it is in units of 5 kpa and 50 rpm ect...

That way when we are tuning auto ve, it would make it even that much more accurate. Not everyone needs 285kpa :) And we wont have to deal with spikes and large areas that need to be modified and blended on our own.. It is hard to tell what cells value should actually be when it has to INTERPOLATE between its surrounding cells.

:eek:

Nah, some people need 285kpa. The resolution is not that big of a deal if you ask me.

[Dirk Diggler]

Instead of having a added boosted ve table that is full of wasted cells for most of everyone,
how about a Super High resolution Ve table that they can use for speed density or really high resolution for 10lbs of boost down ect..
like on the regular ve table make it so it is in units of 5 kpa and 50 rpm ect...

That way when we are tuning auto ve, it would make it even that much more accurate. Not everyone needs 285kpa :) And we wont have to deal with spikes and large areas that need to be modified and blended on our own.. It is hard to tell what cells value should actually be when it has to INTERPOLATE between its surrounding cells.

:eek:

I wouldnt mind having a NA SD custom os like this.

[Black02SS]

I wouldnt mind having a NA SD custom os like this.
:iamwithstupid:

[wait4me]

I think with a sd tune with this high of resolution it would be completly accurate in 2 runs. Instead of it constantly fighting and "BLENDING" with its surrounding cells..

[Black02SS]

I think with a sd tune with this high of resolution it would be completly accurate in 2 runs. Instead of it constantly fighting and "BLENDING" with its surrounding cells..
I would really like to see something like this. I feel it would save some time.

[TAQuickness]

Hi-res SD would be pimp. :hihi:

[SS2win]

If you could set pressure scale of boost table that would be awesome. wondered why it wasnt done that way already.

[Tordne]

I agree too, this would totally rock!

[Blacky]

If you could set pressure scale of boost table that would be awesome. wondered why it wasnt done that way already.
The Stock LS2 PCM operating systems are set up to allow custom scales on tables... The stock operating systems on the LS1 PCM are not.

However, it may be something that can be done in EFILive's 03 Custom OS for LS1s.

Regards
Paul

[TAQuickness]

I love subtle hints

[Dirk Diggler]

If you could set pressure scale of boost table that would be awesome. wondered why it wasnt done that way already.

Can someone in laymens term explain this to me. Ive read it a few times but dont quite understand what its significance is sorry I am having a blonde moment. LOL

[GMPX]

Whoa, hang on guys, you need to consult with me first :eek:

Just kidding, it's funny as we go through this GM code how the LS1 actually missed out on some neat stuff. The Duramax (and as Paul said LS2) have alot of tables where you can define the axis breakpoints.
So as an example, instead of 400, 800, 1200, 1600, 2000 RPM, you can change it to 400, 600, 800, 1000, 1200 etc. It's a coding nightmare, so lets not go there :help2:

I like these custom O.S ideas, it's just time is the enemy, I was hoping that a few of the V4 O.S would be out by now, but with LLY Diesel about to be released in a few weeks it has been held up.
Plus a few of you seem to be having some drama's with the latest V3 O.S changes causing super rich running?, not sure what is happening there.

My opinion on high res VE tables, not worth it, 10kPa is plenty, anyone ever looked at FORD tables :eek: . I've asked Ben from EFI University to give he's opinion on the resolution of VE tables and 'real world' affects.

Cheers,
Ross

[bink]

I think with a sd tune with this high of resolution it would be completly accurate in 2 runs. Instead of it constantly fighting and "BLENDING" with its surrounding cells..

Hmmmmm....

wait4me - can you further explain or clarify this?? Why won't it then fight at higher resolution...of course smaller ripples.

Cheers,
joel

****ALL OF THESE EMOTICONS - and not a "Cheers"???************

[wait4me]

ok, you have 400 rpm at a cell, 400 rpms a lot can change with a higher hp car.

Currently, when the vehicle is being autotuned, the vehicle may only need 2% correction at 200 of the rpm, and 5% at the 400rpm but the blending would say that the cell is 3% off. but depending on the other surrounding cells, since every cell touching the cell in question, They all Start to OVER correct or UNDER correct the cells. But, with higher rpm resolution, it would be able to see those types of fueling needs and would make a less mountain range type of ve table. When you see a spike, that means the the surrounding cells around it are all off, and so is the spiked cell too.

[bink]

I get it! Thanks :D

Cheers,
joel

[EFIGUY]

Jesse,

Not to start one of those silly arguing games most foums suffer from, but I must disagree on this one.

When you tune an engine on a dyno (in steady state) you will rarely see dramatic differences in VE with 400 RPM increments. Air pumps just don't work.

Now, granted there might be one small section of RPM where the VE changes quickly due to several factors like camshaft, compression ratios, etc.. but overall it is rare to see an engine need such high resolution across the board.

In fact 9 out of 10 engines rarely need more than 1000 rpm breakpoints to achieve flat AFR's. Most will need finer resolution in one small section, but almost none will need it everywhere.

Part of the problem stems from doing Auto VE stuff on the street where the cell sampling can be hit or miss, unlike holding the engine steady on a dyno in the center of each cell. This tends to "smooth out" the table and eliminate the problems you have keenly observed of certain cells becomming blended or averaged to get an overall AFR quality without any one cell being exactly correct.

Having a higher resolution table will only magnify the problem using non-steady state tuning methods, and would yield almost exactly similar results by the hold and tune method, but with much greater time required to reach the same goals. What you would end up with is a whole lot of cells containing the same or very similar values.

As far as the MAP values needing higher resolution, the need doesn't really exist there either. Simple laws of physics dictate a relatively easily calculated rate of change in mass for a given rise or fall in MAP. Thus, if you have a value at 175 kpa, and another cell needed to be filled in at 200 Kpa, it is easy to out the correct number in the table to acheive the same AFR or any other AFR for that matter.

Let me try to simplify this all:

Each engine will be different and will tend to need higher resolution in one area or another, but rarely is the error so large that minsicule break points are needed across the board.

Proper steady state tuning on a load bearing dyno will yeild the results you are looking for without the need for our friends Paul and Ross to re-invent the wheel.

Try this for an experiement:

Tune an N/A engine on a dyno at WOT and (hopefully) 100 KPA to have the same AFR across the entire RPM range. (doesn't matter what AFR, so long as its all the same)

What you will notice is that the changes is the table values should reflect the same shape as the engine's torque curve.

Is the engine's torque curve jagged and mountainous? If not, then the table shouldn't be either.

Where ever there are peak's and valley's in the torque curve, there should also be the same in the numbers in the table.

If there are spots in the table where the correct AFR can't be reached because it is between two RPM breakpoints, then you could say with confidence that more resolution is needed in that area.

I think what you will find is that the stock PCM tables already have all the resolution you need for the majority of enignes.

I hope you won't take any of this offensively, as it was certainly not meant to be. I hate those kind of forums.

If what I said helps, then great...if not toss it out and forget I mentioned it!:hihi:

[ToplessTexan]

****ALL OF THESE EMOTICONS - and not a "Cheers"???************

Yes, but there is a pole dancer: :thankyou2:

:rockon: :rockon: :rockon:

[Delco]

I think with a sd tune with this high of resolution it would be completly accurate in 2 runs. Instead of it constantly fighting and "BLENDING" with its surrounding cells..

Jesse , you can already do it by tricking the map scaler , you can have a table with 3 times the resolution in 1 bar mode if you desire :)

[Scoota]

I hope you won't take any of this offensively, as it was certainly not meant to be. I hate those kind of forums.

If what I said helps, then great...if not toss it out and forget I mentioned it!:hihi:

Ben, I don’t think there would be anyone that would take offence to these comments and opinions.

We appreciate you’re input.:cheers:

Cheers Scotty.

[dc_justin]

I ran into this similar problem earlier this week as I was trying to nail down my one problem area, 2600rpms showing 1.1 Ben factor while the two parent cells, 2400 and 2800 are both right near 1.0, as Ben described.

This got me to thinking about a way to show how much interpolation you can expect from whatever conditions you are in. Got me to thinking about the way the Motec system did it in the car that Ben has for his class and how EFI Live could do something like this.

What would it involve to come up with another gauge type, one that has properties similar to that of a Map, but with only two inputs intead of 3.
What I'm thinking here, would be to define the two Axis used (ex, RPM, MAP), define the target values (similar to Col. and Row labels in Map), and define the precision level (size of the box shown attached).

I don't know, maybe I'm way off base here, but seems like something like this could help to nail down problematic cells...

[caver]

I think you guys are being way to picky.
Remember that the engine rpm and manifold pressure are continuously changing when you drive, the slightest change of throttle position changes the manifold pressure so the ecu is continously intoperlating from different points in the map. Trying to get every single cell absolutely perfect is near impossible.
The other factor is when you are doing your autotune the values you are working with are averages so at any time that 1 cell is varying from slightly leaner to slightly richer depending on how close the values are to adjacent cells..