Yet it still works extremely well, and without raping the MAF signal so you can still get a meaningful ECU datalog. Plus, it won't alter ignition timing as a result of said MAF signal hacking because there is no MAF signal hacking going on at all.

Besides, once you get over 2.1 grams per rev of airflow in a DSM ECU, you will be in the top load cell, anyway. You can hit that level of airflow with a 14B turbo. In other words, at WOT, or even at high boost/partial throttle, there really isn't any 3D mapping going on anyway. You are in the 2.1g/rev load cell, and you will remain there.

If the Xede really tries to match the load cells of the stock ECU, then it would act in the same manner. I usually don't fall below 2.7 g/rev at WOT and high boost. And at 5000RPM, I'm at 3.63 g/rev.

 

God I don't know why I'm posting. I REALLY don't want to.

"DSMlink appears to linearize between mapping points as a function of load"

It's not quite that simple. DSMLink uses the relationship established by the manufacturer in the enrichment tables to interpolate between what the user entered and what really gets used. Basically, what the user entered gets used straight up once you've hit the maximum load curve defined by Mitsu and the load adjustments become a static 2D curve.

Below that point, however, we adjust the user entered value by the relationship established in the 3D table between the highest load cell and the one the ECU is actually using at that time. We don't have to guess what that is, because we know. So in effect, we are providing the user with a 3D interpolation function, we just aren't requiring the user to establish it.

You might claim this is limiting to a user looking to tweak part throttle behavior. And there's some truth to that. But so far, we haven't had any requests for this because things work nicely as is; even for guys running 20% more displacement which, I'm sure you understand, really places them on different load cells than was intended by Mitsu because the ECU assumes 2 liter displacement in its estimation of load here... You knew that right? Besides, when we do get such a request, it's a simple matter to implement.

"I wouldn't even consider it as an engine control solution, to be honest."

Yeah, you're right. It's a waste. We'll discontinue the product now. If you wonder why people get pissed off at you and jump all over your ****, it's because of side comments like that one. Why even say something like that other than to stir up trouble?

I don't believe I've ever posted anything on any forum bad mouthing the XEDE product or even suggesting it to be an inferior product to anything else. In fact, I find it to be quite an interesting device and I'm sure it fills a market need nicely. In fact, it fits many markets nicely as it's not tailored to a particular vehicle. Great idea from a product viability standpoint!

There will still be an interest, however, in products tailored to the specific car it was designed for. There always will be. Such a product will never have a huge following, of course, because it's too customized. But so what? It's still a fine product doing what it was intended to do for those that want that type of thing. Did I mention there are folks pushing over 400whp on pumpgas with it? Doesn't mean it's the best at everything all the time and all other products should die off, though...

Thomas Dorris

 

Thank you for confirming this. Can we move on now? Part throttle performance is pretty important.

Shiv

 

Can you elaborate a little bit - what are these relationships between? Adjacent values in stock map cells?

In your opinion, under what modifications are these relationships invariant and under what modifications is interpolation not really appropriate anymore? Do the relationships really hold in situations like you mentioned (more displacement) or are they just "close enough?"

 

Actually, I've always felt that part throttle performance isn't actually all that important. If you don't feel that you are going fast enough at part throttle, what is there to keep you from pushing down the throttle a little more to get the performance you desire? That is the way we all drive, anyway, right?

 

Sorry, I'm not going to reply anymore. Shiv has confirmed my position on his arrogance and I refuse to contribute to the discussion any further.

Thomas Dorris

 

Partial throttle performance is a whole lot more than simply maximizing torque. In fact, it's usually not about maximizing torque at all. IMHO, it's about "clearing the way" for max load. By this, I mean, that it's not unusual to detune things a bit during partial load (those crazy load points between cruising and full boost) to keep things are knock resistant as possible.

In our Stage 2 car, for example, we had to be very careful in the 4000-4500rpm partial throttle range since this is where the car suddently gains 400hp. This sudden (almost immediate to be quite honest) increase in cylinder pressures tends to induce knock which, due to knock control hysterisis, will end up adversely effecting output when full load is acheived.

To fix this problem, I had to be more generous with fueling and conservative with timing in this particular partial load sites. With partial throttle knock avoided, I can be as aggressive as I want to be at full load where maximizing torque actually counts.

My 2c,
shiv

 

When you’re going in a straight line your right it's not important.

When you don't care the performance of you car is not factory like in smoothness again it's not important.

When you’re doing back to back 20min track events and want to run constant times and try to modulate the throttle for the best line pulling out of a corner well then it would be very Important to me but might not be for other people doing other types of motor sports venues.

But to each there own.
Eric

 

Sorry you feel that way. I wasn't trying to be smug. But you did make my point.

shiv

 

Throttle modulation is typically rather difficult with turbocharged cars. You try to get the turbo spooled up quickly by giving it a little extra throttle, but as soon as the turbo spools, you have to be careful to back off the throttle slowly and not too much so that you don't stall the compressor or open the BOV too much. And once you start running with larger turbos, it gets even more difficult.

What DSMLink does does not hamper the "factory smoothness" of the car, however. It is perfectly smooth. Smoothness at part throttle has more to do with your hardware setup (turbo size, blow off valve, etc...) and your right foot than it does tuning.

Incidentally, try left foot braking. It can work wonders to settle a car, and allows the turbo to spool under the braking load for explosive torque when you let off the brake.

 

I'm starting to race shifter carts and have always done left foot braking. I atended a riding school a few years back with my Zx750r to promote riding on the road safty by evoling my skill as a rider. The instructer Drilled in smooth ness to me to exit a coner for max control and lap times. I aslo pratice this in my car when taken corners and trying to get the most exit speed. I do see what your saying but even thow my car is turbo charged part throttle performance is smoother then glass.

Eric

 

"Sorry you feel that way. I wasn't trying to be smug. But you did make my point."

Dude, if that wasn't being smug above, WTH do you consider smug!? Regardless, I'll buy the retraction; once.

And no, I did not make your point. I clarified a misconception. I never said adjusting part throttle load cells was necessary; a precondition to your "point". I only said if one were to claim that was needed, then yeah, you could consider DSMLink slightly (ever so) limiting. But I don't feel that type of adjustment is necessary if you use the factory defined values as they were intended. And I further backed that up by stating that we've never had anyone request that feature, even the guys (myself included) that use DSMLink for road course events on a regular basis.

"To fix this problem, I had to be more generous with fueling and conservative with timing in this particular partial load sites."

I assert that you have to do this specifically because the airflow signal has been altered. Alter the airflow signal and you alter how the load tables are indexed by the ECU and that introduces the issues you're fighting with the fancy 3D table conversion. Leave the airflow alone, tune fuel with fuel adjustments and timing with timing adjustments and things play together much more nicely.

That's not to say the Exede sucks. And it's not to say DSMLink does everything right. It's simply stating different approaches have been taken and require different features to compensate.

Thomas Dorris

 

Thomas the DSMlink offers so many features, but what do you feel the DSMlink does not do right?

Eric

 

DSMLink makes a crap cup of coffee.

 

Well at least it makes coffee Throw enough sugar and cream in it and there is no such thing as a bad cup of coffee