When the SMART system is enabled based upon the SMART threshold map, invariably I have found that there is a transient rich AFR condition that results in a noticeable fluctuation in power delivery. In the attached plots you can see at 41% load point (SMART switch map threshold) the AFR rapidly goes one point richer.

From my reading of the SMART templates and maps, it doesn't seem that this condition is a function of the map data. Why does this occur?

The transient rich condition makes the car feel like it is encountering detonation, thus it is somewhat annoying, especially given that the old Xede was so smooth in power delivery. (Note: The data proves that there is no knock.)

 

Bump

 

Vishnu guys,

These are the questions we REALLY need answers to.

I see the same thing... and it is not pleasing.

 

Zeus, Its the weekend. I realize that you guys want answers, and I want to find the answers to your question. However, every weekday I wake up at 8, and work until midnight, whether it be on car stuff or school stuff. Its nice to have just two days off where I am not trying to trouble shoot issues that we have never seen on any of our test cars. Come monday, we will do our best to get the answer to the question figured out.

 

I understand being hammered schedule wise... Sometimes I think I need to just up and "leave" for a while to get my head on straight... hence the mini vacation I recently took.

So Monday, or Friday, it matters not. What does matter is that there have been some very specific support technical questions that have gone almost ignored concerning the SMART system (of course I don't think they are being ignored on purpose... well, I'd hope not).

Freedom among a couple others seems to keep stumbling onto the same issues I do. The ultra rich SMART fuel settings for example. These issues leave a few of us thinking that the reasons for the tuning being the way it is, is to "cover" an uncontrollable fluctuation. There is a huge customer base that wants to see a leaner, but "tighter" tune with the SMART system. Many of us were pretty happy with flashes running low 11 AFR's. We invested in the SMART system due to the purported advantages of the "real time" attributes it is advertised to have. I see an extreme amount of potential in this system, and myself and others are just looking to maximize the benefits of SMART. However, first we have to get it running in an acceptable manner in the first place.

So do rest up Dustin, the customers need you...

 

how does the car feel when this happens ?

 

Have you felt the stock ECU get involved on knock event? Like that. Basically a momentary hesitation in power delivery.

 

I also have those momentary drops in power delivery. However, all my datalogs still contain at least 2 knock events per pull, so i figured thats what it is. I've been slowly tuning AFRs and pulling timing, but have yet to get rid of the knock or that momentary loss of power. i have not seen momentary rich condition though. i will look for it now.

In 3 days my turboback will be finished, and i will load up the stage 1 maps and start working on them. we'll see how it goes.

 

I have that also
sort of like when I punch the gas it take a moment or two for the power to come up

 

So is the rich spot on the onset of boost?

 

Mine is at WOT going up through the RPM range. It is just a momenatry loss in power, almost like a misfire or something, but no actual misfire or code or anything. You can just feel it stop pulling, then regain power again.

 

Good question. I'm not sure as boost isn't directly logged. Clearly this is the region that the engine is going back on boost.

That said the logged data for Freq0_In shows increasing flow with time. On the other hand the Freq0_Out log shows a sudden decrease right at where the point Freq0_In bounces at or above 41%. Subsequently as Freq0_In continues to increase the Freq0_Out recovers back to where it was prior to Freq0_In crossing 41%.

Note that all of my logs show this behavior independent of what map I'm running. In fact I have tried running larger corrections in the SMART fuel map at the appropriate RPMS to force the Xede back to the goal (zero line). But even with larger correction values no difference (beyond run-to-run variations) was exhibited.

Is there processing that the Xede does such that when the map cuts in there is a non-obvious result on the Xede's output? For example I think your web site states that timing is only allowed to change at a limited rate to prevent the ECU from throwing a code (e.g. when cells next to each other have a large difference).

 

Hey guys, This is something we are going to do extensive testing on when lachlan from chiptorque comes out. After looking at the graphs, Shiv and I concluded that it may be a glitch with the turn on. I am going to do a bunch of data gathering in the week to come, and then we will work with lachlan on this when he gets here. I may need your guys' maps to test, so I can try to replicate the issue.

 

This is why I was initially trying to reduce the point at which the SMART started to work, like at 30% instead of 41. But I do get the same thing, any chance of this being fixed by the Texas dyno day Nov 12-13?

What is more disturbing is that the SMART see's the transient rich condition and just cuts the fuel delivery for a second thus going pretty lean. Is any damage going to come from this lean condition? Iknow that it is a really quick hit, but it goes pretty lean...

I go rich right at 4200, every time..

 

Does this happens to ALL the SMART users or just the people that tune their own cars???