worf928

This is the point I/we were trying to get across. Especially the bold part. You've got it except for, perhaps, one nuance: it is up to the person/organization responsible for the health of the engine to make the ultimate determination of what is "appropriate for [their] boosted motor". Thus, information and/or fine-grained control is required in addition to any 'help' the ST UI might provide. A solution along the lines of a checkbox on the UI for 'boosted' engines isn't sufficient.

I assume the need for fine-grained control is obvious? If not, let me know.

Only to the extent that it highlighted, very early in the tuning process, that the behavior of the LH was unknown with respect to 02 trim cut-off. I'm actually glad that I discovered this behavior early. If, hypothetically, the D.R. TS system had measured 100% of the air, I might not have found this 'unmapping-of-the-fuel-map' behavior until the few folks with the beta-2 kit tossed a piston.

We don't have a functional description of 02 trim cut-off. It's a black box (to us); We can only speculate on 'might be.'

Yes, back in 2008, I understood why I found this 'problem.' The logic behind the behavior was just about obvious. But, again, LH black box. Had to use a big hammer. Install pressure switch.

Not quite. It would require either a) two MAF's with the appropriate circuitry to combine their outputs such that they acted like a single SuperMAF or b) completely throwing away the twin setup and redesigning every bit of hardware that connects to the inlet of the twin-screw so as to allow the use of a single MAF whether Super or not Super.

In both cases I would still bet someone else's pistons, that the LH's default 02 cut-off algorithm is in the worst case not safe or in the best case suboptimal for boosted applications.

jcorenman

Dave, If your goal is to beat me up for not being involved in 2008, then go ahead. I am guilty of that.

I cannot discuss history, I wasn't there. What I would like to do is go forward, and try to help put problems and solutions together.

If you want to revisit the tuning for DR's twinscrew then I would welcome your involvement, and the chance to work with you and sort out the issues.
I am sure that with what we know now and didn't know then, that we can do a better job.

Cheers, Jim

John Speake

Wouldn't the Alpha-N be the solution for a twin screw where boost is well defined versus rpm and throttle plate position ?

FredR

Not sure where this thread is trying to go at the moment- ST2 is the saviour of all these problems and alpha -N an added bonus. Both Jim and John have been super at offering support and anyoen who has followed this thread knows the limitations of a single MAF. Those of us who have used [& abused] ST2 know how user friendly it really is and I am pretty sure my S4 currently runs as well as it ever has despite whatever may ail it in terms of wear.

Once you start "custom tuning" which is what ST2 is all about I fail to see the need for WOT switch, WOT maps, adaption or anything else for that matter- it is all right in front of you and you can see exactly what is happening in real time or after the fact with sharkplotter. Not saying what I have done is perfect or there is no room for improvement but compared to a shortwhile ago when folks were acting like "cavemen" using extra fuel pressure regulators to squeeze the crap out out the fuel system under boost we have moved light years ahead down this road.

Years ago Louie took someone's twin screw early [Andy Keele?] effort, sorted it out mechanically and tuned it using the original ST and it ran like a champ-and what a torque curve- [a table top to be precise].

Apologies if I am missing something here. A Great product that is still evolving and easy to get to grips with, great people and a great thread- hopefully we can keep it that way!

Regards

Fred

worf928

You now have me almost completely baffled, rather than mildly baffled. My goal was to reinforce Tuomo's speculation with actual experience and reinforce his point. More words on my part don't seem to result in anything other than further distraction and confusion.

jcorenman

I think it would work great for that, and would also remove the constraint of working around the MAF when designing the intake. The failure modes that I can think of-- belt slip, air leak, etc-- would cause Alpha to over-estimate airflow and cause a rich mixture, not lean. It has certainly been working well in our GT, three years now.

Cheers, Jim

ptuomov

Alpha-N may work for a belt-driven positive displacement supercharger, but I would caution that the hot-wire mass air flow sensor is a lot more consistent method for measuring steady state air flow than a table look up based on throttle position and rpm.

This relative inconsistency is especially the case if one uses a large plenum manifold with a single throttle body. Minor variations in combustion can cause a large variation in the low pressure pulse during the overlap and a correspondingly large variation in cylinder filling from the plenum manifold, even if the throttle angle is exactly the same. Higher the valve event overlap, larger this instability. The one way to minimize this is to use ITBs placed right next to the intake port, which reduces the volume between the cylinder and the throttle plate and dramatically reduces the instability. That's why by my casual observation almost all the cars that come as primarily Alpha-N from the factory have ITBs.

My understanding however is that Jim has been testing the Alpha-N system with the stock plenum manifold and larger cams. If the system works there without any instability, then it'll probably work just fine with a positive displacement supercharger and smaller cams.

Now, if we're only trying to solve the 02 loop and adaptation problems, just recalibrating when 02 loop runs (once that feature is in ST2) will be a lot simpler and, in my opinion, a better solution. If one wants a belt AND suspenders, one can then also ground the WOT switch pin based on both throttle position switch and a Hobbs sensor.

jcorenman

Except it's a bit of apples and oranges. I don't know if a twin-turbo can get into enough load at less than 3000 rpm to have the O2-loop and adaptation be of concern.
ST will show that, but I haven't seen the data. In any event I agree that making the O2-loop limits adjustable is a good idea. Last week I didn't know if that was possible, now we do and it does work.

A twin-screw generates a lot of boost at lower RPMs than a TT, and that's a different story. But that was also six years ago, and we know more now.
If you want to revisit the twin-screw tuning and test this in addition to or instead of a pressure switch, I would be happy to help.

Cheers, Jim

ptuomov

If someone has a twinscrew torque curve, one could post it next to these. It would be interesting to see at which rpm level the torque curves intersect. The same deal with normally aspirated engines. My guess is that the twin turbo load will be considered "too high for closed loop" pretty early...

Small turbos (gt2871rs):



Medium turbos (gt3071rs):



I'll post the large turbo (gt3576rs) graph when we run one at low rpms.

The punch line here is that the stock S4 engine makes 300 ft lbs of torque at the closed loop cutoff rpm (by my assumption 2800 rpm). This is only at full throttle, so it's a too high of a cut off but bear with me. The small turbo S4 makes that much torque at 2200 rpm and the medium turbo S4 makes that much torque at 2300 rpm. Therefore, it's pretty clear to me that we don't want to be running the turbo engine in the closed loop at all load levels all the way to 2800 rpm.

jcorenman

Tuomo,

This is interesting but to be useful I think it needs to be at 3/4 throttle (or wherever the throttle is just before the WOT switch closes). More than that, and the LH is running open-loop anyway by virtue of the WOT switch. And it also needs to relate back to MAF numbers, because that's what the LH is seeing.

But I think we are looking at the wrong end of this. Dyno sheets are WOT, and the topic is adaptation-- which has no place at WOT. For tuning, do what you are doing-- run the LH in open loop mode and then adjust each cell for the AFR that you want there-- individually or en masse with SP. Don't worry about the O2-loop at all.

Once the car is back on the street, what we really want is for the O2-loop to be active in the normal driving range, and not active elsewhere. I would submit that, long before you get to the worrisome parts of the map, we would prefer that the LH just go open-loop. The new ST code allows exactly that.

This allows the adaptation to "learn" what is needed from the middle of the map, and then use that elsewhere. Does that make sense?

Cheers, Jim

Hilton

The problem I see there (and this may be a gap in my understanding - if so, please correct me) is tuning the middle of the map for AFR 12.5 (say 3.5k rpms and 180 load) with the O2 loop disabled, which would seem entirely reasonable for a boosted car.

Once you plug in the O2 loop, the car will use O2 adjust to drive that part of the map at 14.7 (leaner), and as a result the adaptation function will shift towards leaner.

Once the adaptation function has compensated, the O2 adjust will be back towards the middle of the range in those cells. Then the adaptation function will effectively pull fuel out at all loads, as it's applied everywhere on the map (even though it only "learns" at low loads), including higher ones and WOT with potentially disastrous effects on a boosted car.

Even for a non-boosted car, I think the same thing will happen - just with a less disastrous effect which may just result in knocks and subsequent retard being seen post-tuning at higher loads/rpm?

PorKen

<snip>

Tony

I have a presure switch that goes WOT at appx 1psi...i can see it on my ARM 1 AFR. It hunts...then when it hits 1psi, it activates the WOT map and goes rich. Works great. It always comes on when i climb the hill on the hwy that i drive when i pick my kids up from school. The throttle is no where near the floor, but boost just touches 1psi as i go up the hill.

I dont give a crap about MPG

jcorenman

Sorry, this is not making sense. I've stared at it for an hour and can't figure out what you are trying to say, that hasn't already been covered.

The 3500/180 cell is not high load, boosted or not, why set it to 12.5:1 fuel?
In a higher-load part of the map, then yes you don't want the O2-loop to be active for the reasons stated.
And where the O2-loop is enabled then you want those cells tuned for 14.7, or close to that.
But at higher RPM, boosted or not, the LH always operates open-loop in the high-load cells.

Yes, we can probably construct a fuel map that will mess up adaptation (remember that it can only go +/- 20%). But I don't think that is your point?

PorKen

<snip>