VE "dip"

[Calan]

You guys will hate me for this, NO FUEL PRESSURE REGULATOR, just 255lph and whatever pressure it goes to.

So who knows where your global was at.

Calan -In regards to slider adjustments, cmon..its ~3%

Yeah I know. I was just wondering why they had been touched. No biggie once I thought about it a minute.

Is this on a 2.0l or 2.3l?

**********

You have to expect to be picked apart when you post up a VE table that could have come straight out of a Greg Banish book.

 

[mindset]

Huh?

For all the hours I've spent on here I can't recall what VE means.

 

[Calan]

VE = Volumetric Efficiency

In simplified terms, it is a way of quantifying how efficient your motor is at filling the cylinders with the air/fuel charge.

**********

So as for my original question (Sam?)...any idea why I and apparently Ravenous see this dip at 3500 rpms? I've seen it with both an Evo 16g and the current Holset, different UICP and intake configurations, and rebuilt head.

Relevant components are pretty basic: 2.3l, FP2 cams, stock IM and head, 62mm (otherwise stock) 1G TB, FP race exhaust.

 

[Sam@RRE]

2.0L It could be injector latency vs rpm. What injectors are you both using. My experience with PTE/FIC injectors is they have been picky with voltage. That entire area of the map is controlled by the fuel trims still, and if the basic injector settings are off, it will reflect in that area of the map because the fuel trims are trying to correct for it. So in essence, your car is going rich at 3500RPM and lean or perfect at 2500RPM and the CombinedFT method is pulling from the VE to compensate for the richness. And when you say you do the math manually, are you going off the fuel trims? If so, again if the basic injector settings are off, you'll get the same. And judging that your VE values at idle at are kind of low I think there is something up with your global/deadtime values and the fuel trims are using the VE table to compensate.

Post a log, what size injectors are you running?

 

[Calan]

...when you say you do the math manually, are you going off the fuel trims?

No... open loop and a fixed AFR.

And judging that your VE values at idle at are kind of low I think there is something up with your global/deadtime values and the fuel trims are using the VE table to compensate.

Anything is possible, but these injectors were flow tested and matched to within 1%, and came with a flow data sheet from FIC. I've had deadtime all over the place (like most people), but finally settled on something that gets everything where it should be and got rid of any lean misfires at low loads/idle. Right now it's at 715us I believe; before that I was at 525. Either way, it has never seemed to affect the dip.

Post a log, what size injectors are you running?

1250cc FIC Bluemax. I pulled some more logs last night and this morning....I'll see if I can find something useful to post up.

 

[Ludachris]

I wonder if we could get T. Dorris in here. Maybe he'd have some ideas.

 

[gofer]

I wonder if we could get T. Dorris in here. Maybe he'd have some ideas.

T. Dorris and D. Mertz are too busy giving Craig basic math lessons over on the ECMLink forums.

A Question for Tom or Dave... (VE Scaling)


I personally think that "VE dip" that your getting Craig is the awkward transition between closed/open loop, that seems like the place it would happen too. True engine VE won't have any weird dips like that so neither should the SD VE map, which needs to reflect true engine VE to be tuned properly. My SD VE map doesn't have a VE dip there either, below is a screen shot.

 

[twdorris]

Craig's a smart guy. We're just having a little fun.

The ripple you're seeing is probably real. Dave noticed something very similar with his own car. In fact, that's one of the main reasons our SD implementation went full 3D. It was originally two linear tables, one indexed by MAP and one by engine speed (RPM). But Dave noticed several areas of operation where this was not cutting it. You just could not dial in correct behavior using two linear tables.

Thomas Dorris

 

[knochgoon24]

I don't have that dip either.

 

[Sam@RRE]

T. Dorris and D. Mertz are too busy giving Craig basic math lessons over on the ECMLink forums.

A Question for Tom or Dave... (VE Scaling)


I personally think that "VE dip" that your getting Craig is the awkward transition between closed/open loop, that seems like the place it would happen too. True engine VE won't have any weird dips like that so neither should the SD VE map, which needs to reflect true engine VE to be tuned properly. My SD VE map doesn't have a VE dip there either, below is a screen shot.

Thats the base one I dropped in Gofers car and made some small tweaks. Any testimonials on how your car drives/idles/cruise/WOT with that base VE map?

 

[Sam@RRE]

Also to add to this, I just don't like FIC injectors, especially the bluemax 1250's. Tuned maybe 4~5 cars with those, and none of them worked as advertised, really picky and inconsistent. I've the standard FIC/PTE 1650's on 91 octane and it idled, drove, cruised fine, and done 950's and same where good and worked. 1250's for some odd reason just didn't like to play.

As far as thier "spec sheet" I really dont believe it, it looks like something they xerox and stick in every box. There was a thread on the Link forums about a guy fighting to dial in a set of 1550's or 1650's or something and turned out one injector was a 1250. When the box said they all were balanced.....I can't seem to find the thread though.

 

[knochgoon24]

I've had issues with getting my PTE 680's dialed in. I won't thread jack though. I just had to use really weird injector scaling and latency values to get them to work.

 

[Calan]

The ripple you're seeing is probably real. Dave noticed something very similar with his own car. In fact, that's one of the main reasons our SD implementation went full 3D. It was originally two linear tables, one indexed by MAP and one by engine speed (RPM). But Dave noticed several areas of operation where this was not cutting it. You just could not dial in correct behavior using two linear tables.

Sounds reasonable. I know I sure can't explain it, and it goes against everything I would expect as far as how the engine "should" be operating.

So what was causing it on Dave's car? Did the new implementation allow it to be dialed out, or do you still see it?

Also to add to this, I just don't like FIC injectors, especially the bluemax 1250's.

I really like them. They have worked better for me than several others I've tried (smooth idle, repeatbility, etc.).

Other than the airflow seeming a bit low across the board, I don't see any real-world negative effects from them...unless this VE issue is truly caused by them. I have a hard time believing that though, because the dip at 3500 rpms extends well into open loop territory where trims and latency shouldn't be a real player.

I'm going to do some more tweaking on it this evening, and will post up a new log when I have something solid to look at.

 

[Ravenous]

I'm running FIC 850's purely open loop, so the transition from closed to open wouldn't effect my table at all and I have a nearly identical dip. It is good to see that Tom and Dave have seen this quirk first hand.

 

[Calan]

So does anyone else have an ECMLink VE table they'd like to share that is NOT one of Sam's?

Preferably for a 2.3l.

 

[enos1010321]

Calan my friend, you are not alone...

You must be logged in to view this image or video.

Same spot and everything but I am on a 2.0 with 8.5:1 pistons. If I do the calculations by hand or have my tuning spreadsheet do it, I still end up with the same results.

Wait, do you actually have VE's in the 130's??? What are you running? That thing must scream! I checked your profile and I don't see anything like an aftermarket intake manifold? I am just wondering how are you getting that good of a fill rate!

what injector size are you using? global settings and dead times. that dip area is being effected by fuel trims and the VE wizard in link is based off fuel trims. If you got and incorrect global/deadtime/voltage vs latency tables it will mess with it. Best way to build a proper VE table is set entire fuel map off boost to 14.7 and under boost to 10.5 and disable any o2 feedback (lock in open loop) and adjust VE until reported AFR on the wideband matches closely to target AFR.

I would like to see an example of your fuel map for that. I have always tried to figure out a better way than using the wizard on ECMTuning. Also is there any issue with EGT while doing that?

Has anyone else noticed a dip in the VE at around 3500 rpm (specifically on a stroker, but on any setup in general)?

Does anyone have upgraded valve springs? I wonder if it is a half max rpm issue, 7000rpm. Or anyone have a different exhaust diameter then 3 inch, I know the turbo changes back pressure and so exhaust has a negligible effect. Anyone have a map with a ported head, 1g or 2g, is it a design issue with the heads?

 

[Calan]

Wait, do you actually have VE's in the 130's??? What are you running? That thing must scream! I checked your profile and I don't see anything like an aftermarket intake manifold? I am just wondering how are you getting that good of a fill rate!

I'm sure that table isn't referenced the same as the one in ECMLink.

Does anyone have upgraded valve springs? I wonder if it is a half max rpm issue, 7000rpm.

I now have Kiggly beehives on a completely reworked head (new valves, seats, guides, seals, retainers, and springs), but it was identical on my old stock head and springs.

 

[Ravenous]

Wait, do you actually have VE's in the 130's??? What are you running? That thing must scream! I checked your profile and I don't see anything like an aftermarket intake manifold? I am just wondering how are you getting that good of a fill rate!

I'm sure that table isn't referenced the same as the one in ECMLink.

That is exactly the case. The thing about VE in our tuning software is that it is an arbitrary number that is more of a reference point than anything. Also in real world measurements VE on a turbocharged car can be more than 100% so a lot of us could be running 130% VE or higher, it is just how the software we use is designed to express it. Since I am using Tunerpro and the Moates Ostrich to alter my pressure indexed maps mine will look a bit different than Calan's or Sam's because the software was designed with a different scale for VE although in reality they are both doing the same thing.

 

[enos1010321]

That is exactly the case. The thing about VE in our tuning software is that it is an arbitrary number that is more of a reference point than anything. Also in real world measurements VE on a turbocharged car can be more than 100% so a lot of us could be running 130% VE or higher, it is just how the software we use is designed to express it. Since I am using Tunerpro and the Moates Ostrich to alter my pressure indexed maps mine will look a bit different than Calan's or Sam's because the software was designed with a different scale for VE although in reality they are both doing the same thing.

Ok, that explains it. I knew you could have a ve of above 100% I just had never seen a dsm consistently in the 130%+ range. Well without radical modifications.
thanks

 

[turboglenn]

has to be something with the airflow dynamics of the engine i would guess. Even on the stand alones i've installed on DSM's i've noticed they like more fuel at 2-2.5k RPM and then tend to lean back out (or prefer less fuel at around 3k RPM) I don't think it's anything that's an "issue" but just a charachteristic of the engine and the way it breathes. For this reason i tune that area nice and crisp and choose the gear i'm in while cruising to keep me in that area for fuel economy, although it's firing the injectors another 500 times a minute compared to cruising at 2500 RPM, the pulsewidth is low enough that i think it's actually a litte easier on fuel

Either way i don't think it's anything to be concerned about. Every engine i've ever tuned seems to have areas where although the RPM is higher, the fuel requirements are lower than those at some of the lower RPM ranges in comparison.. I"m sure we've got some one on this board that understands the physics and principals as to why it happens, but i just consider it part of the engine's personallity

 

[BoostedBeaver]

So does anyone else have an ECMLink VE table they'd like to share that is NOT one of Sam's?

I have a HX35 with a BEP housing. Other mods are in my profile. This tuning is not yet complete but my VE suffers horribly up top. I have only taken the car to 7300 or so but VE had to be dropped to line up WB and AFREST. My downfall is up top which is a issue where your problem is in the lower rpm.

 

[twdorris]

but my VE suffers horribly up top

You should have seen mine with the 2.4L on stock 2.0L cams. THAT was bad up top. It got a LOT better with HKS 272s, but it was still down a good bit over the original 2.0L/272 combo. The end result (higher displacement with lower VE) still ends up making more power, though.

Thomas Dorris

 

[gofer]

I went back through my dyno graphs and found one of my AFR's, seems there might be a dip or am I not looking at it correctly?

You must be logged in to view this image or video.

 

[Calan]

After hours of driving, adjusting, and close scrutiny, my bigger problem is a crazy lean spike at 2800 rpm and low-medium loads (which was making the 3500 dip look even worse than it is).

Here's a log. This is a constant acceleration in open loop from 2k to 5k, at a steady 35% throttle.

You must be logged in to view this image or video.

Someone please explain that crazy spike at 13.9 seconds in, because I'm at a loss. And feel free to adjust the VE table to compensate, and notice where the numbers end up.

 

[gofer]

Not a problem, just change those two cells to 134.5 and 157 in your VE map. Lean spike eliminated.

I think this may be what Sam was talking about, FIC BlueMax 1250cc Injectors being picky and inconsistent.