[OpenFlash]

Hi guys,
As most of you guys know by know, we've made some pretty big advances in the last couple of weeks with respect to the new lagfix 2 beta maps that have been floating around. Boost response has been improved dramatically. As has boost stability and, as a result, general drivability.

But what has yet to get be released is something that we are testing right now in house. And that is a fix for the dreaded shift lag that many of you 6MT guys have experienced on the drag strip or when shifting quickly on the street. It doesn't matter if you are running a piggyback or a reflash as both are susceptible to it when running higher boost levels. It often happens just after the 3rd to 4th shift where the engine just falls flat on its face for what seems to be an eternity before picking back up. It kills both your excitement level as well as your 1/4 mile ETs

Here's a datalog of that dreaded shift lag during a 3rd to 4th shift. captured:

Take a close look at the 3rd graph were both applied throttle (black) and actual throttle (blue) are plotted on the same graph. You can see that the executed throttle lift (by my foot) is very quick, with the throttle going to full 100% just 0.2 seconds after it closes. However, the Actual throttle shows a very different picture, with the ACTUAL throttle opening lagging over 0.5 seconds behind the applied throttle opening. Not exactly an eternity but when you are trying to go fast, it sure does feel like it when you aren't accelerating for a few tenths of a second.

Now take a look at this log of a similar 3rd to 4th shift. But this time with our software fix:


See a difference?

I can't say how we are doing this for obvious reasons. But we are relying on CANbus data so this fix will only apply to Rev.2 users unfortunately. But can you imagine the benefits of picking up 0.4 seconds of acceleration time in one 1/4 mile pass?

Cheers,
shiv

[dkvby]

Awesome!!! What caused the delay?/Whats the fix?

[jpsimon]

now get the boost to hold like it did in that datalog I posted for every shift, and this will be the greatest thing since sliced bread.

[scottp999]

Please release some of these new beta maps to the general population soon. I'm jealous.

This new feature looks amazing.

[RambleJ]

Ghost edit.. Reread the OP.

[jpsimon]

Black line = shiv's foot
Blue line = actual throttle

[OpenFlash]

Originally Posted by jpsimon

now get the boost to hold like it did in that datalog I posted for every shift, and this will be the greatest thing since sliced bread.

I don't think that will be a problem. The throttle lag issue was a lot harder nut to crack than getting boost to hold during the shift. And it's benefits are also MUCH more evident on the road.

Shiv

[stressdoc]

MT ftw! Good stuff.

[mmmotornutz]

Why does 6MT suffer from this lag? What is the delay in place? What is it suppose to prevent?

[supradriver]

Sounds great. I can't wait to test it out.

[OpenFlash]

Originally Posted by mmmotornutz

Why does 6MT suffer from this lag? What is the delay in place? What is it suppose to prevent?

Can't say much about it other than it is caused by a condition that occurs when running more than ~12psi while shifting quickly at high rpm (above 6000rpm). There exists an aspect of the factory boost control logic that works well when running 5-8psi. But is inappropriate at higher boost pressures. We are modifying this control logic based upon CAN data inputs.

Shiv

[bren335i]

Shiv - why does this sound very familiar to the fuel boost resume hex box ive worked on in subarus before

[Ramos]

Shiv, Is there any safety factory feature that is being compromised by doing this?
I am just curious and not very knowledgable as to what this is doing to the car?
Maybe more abuse on the clutch?

And when can we expect a public release?

[citrixscu]

Shiv,

Will this fix be limited to the Rev II CANbus models only, much like the current beta lagfix parameters?

[Grendalf]

Great work!
Looking forward to test it out

[Grendalf]

Originally Posted by citrixscu

Shiv,

Will this fix be limited to the Rev II CANbus models only, much like the current beta lagfix parameters?

I am not Shiv but I know the answer since he wrote it in his post.

Yes it will only be available for Rev. II since it is relying on canbus data.

[citrixscu]

Originally Posted by Grendalf

I am not Shiv but I know the answer since he wrote it in his post.

Yes it will only be available for Rev. II since it is relying on canbus data.

Read that just as I pressed post. Cheers.

[bulldog_yyc]

does this lag have anything to do with the torque management system?

[Clap135]

Target boost during a hard shift in a 6mt is 0psi. As long as you tell the ecu you are at 0 psi, it will request full boost again. Sounds good.

[scalbert]

A software upgrade to gain about 0.5 seconds in ET, That is nice.

[OpenFlash]

Originally Posted by turkish335

does this lag have anything to do with the torque management system?

No. It's a "glitch" in the boost control system. By glitch I mean that it works perfectly fine in low boost applications. But in higher boost applications, it expects the turbos to do things which they physically cannot do. As a result, the DME closes the throttle and plays the painful waiting game until the turbos de-spool.

shiv

[dkvby]

Does the "glitch" u talking about the higher boost is only from Procede cuz i saw all procede datalog with boost spike to 20+. Thats when the glitch happens?