Analyse my Dyno charts, help find my problem

[frankiman]

so i went on the dyno 2 days ago

LAST YEAR, i got 382whp @ 6600rpm with 13,7psi peak boost



with AFRs at 12





but this year, i changed my old boost controller for a brand new HKS EVC-S, changed to brand new Splitfire Coilpacks, swapped the old recirc valves for brand new HKS SSQV BOVs and removed the cat in favor of a resonated test pipe

all those mods were to help minimize the boost drop, and make the turbos spool a bit faster

the HKS EVC-S was a pain in the ass to program and it seemed we never got what was programed,

and it ended up beeing much worst, i was loosing tons of boost, and getting an erratic boost curve

to get the same 382whp as last year, we had to set it up to get a peak of 15,5psi then it dropped to 12psi to redline

so at the same boost settings as last year, i would loose power


help me analyse my problem with these charts

Boost VS Power


AFR VS Power


Torque VS Power


at first we thought the boost controller might have been setup wrong, but Mitch (Mitch32) and i checked it twice and it was perfect

then i was told the problem could have been my restrictive exhaust (i have an HKS Legal cat-back, but no cat and Nismo downpipe)
but it seemed odd to me that i upgraded the exhaust and the problem was worst
WTF

i checked my SSQV for leaks and they were fine

but i do have old school HKS race BOVs and the passenger side of the car,
because i didn't want those to mess up the ystem, i blocked the valves' exhausts and disconnected the vaccum port

now, could there valves be leaking and creating all this erratic boost issu?

i'm dropping the car off @ autoworx this week for other stuff, but i will ask them to check into this, and to take no chances, i'll ask them to remove those old BOVs and weld the flanges shut


anyone has ever had the same problem??

[frankiman]

holy crap, just examined my charts again,
and i just realized my turbos spooling much much quicker
at 4500rpm, i used to get ~260whp, and now ~325whp, yeah the 4psi increase is making quite a difference at that spot

[GTRADDICT]

Does is leak boost? I'm not sure how to read the BOOST vs. POWER but it seems you're loosing boost or gaining boost as your RPM rise. Would nice to have a BOOST vs. RPM graf... And the other BOV's are potentially something that could fault you (should check if the BOV moved to release pressure, or even do a pressure test... If it's a variable, it can change...
Nick

[frankiman]

the rpm is at the bottom of each graph

[Skym]

Higher boost for same hp = leak somewhere. Could be BOV's, intakepiping, etc. Pressure test the intercooler piping, BOV's together. It should be obvious where the leak is. But if it isn't obvious where the leaks coming from, spray soapy water on joints, etc.

How to pressure testing intake piping and mentions where to buy the plugs -

http://www.youtube.com/watch?v=atretf9u-Mc

Where to buy plugs -

http://www.siliconeintakes.com/index.php?cPath=8&osCsid=20adf99bdad10f9e61603c94324a449a

Also check BOV springs, as there's different springs for different vacuum from engine. There should be optional springs, at least there is with I think is Tial? BOV. For example, if you fit cams, change springs on BOV's to suit due to lower vacuum from engine at idle.

Also if you are using stock BOV gaskets, that's another place where it can leak, as they are use once only type of gasket. Same type of gasket is used on turbo to exhaust manifold joint. From what I understand, if you crush the ridge seal in middle of BOV gasket, it's stuffed and can't be used again. Found this out the hard way after trying to find airleak and it solved itself after changing to Synapse BOV that uses a O ring seal on BOV flange.

[Zacho]

Your boost curve is all over the place. I suspect something leaking air or the boost controller setup needs big adjustment. You should have done a run on wastegate pressure.

What management/tune do you have? Did you tune/adjust for the increase in boost?

Where was the af sensor placed?

[cobrAA]

by the way frank, I lost my dyno sheet lol!! I hope robin save them on his computer!

[frankiman]

yup robin has everything saved up, me and mitch even checked TW's chart, hehe

Zacho;
the AFR sensor is right at the end of the downpipe, just before the cat
if you look at the AFR chart, its perfectly stable at 12

we did runs with the boost controller off, and at stock boost pressure (~9,5psi) it was a bit more stable, loosing only ~,7psi


i traced both boost graphs on top of each other and matched the scaling, and its quite weird



somehow, i'm getting boost much much ealier and because of the high boost peak, i'm getting a maximum power increase of 60whp

but after that, i'm not getting much power increase, might be caused by boost leak

and after thinking about it, the high boost peak may be due to the reaction time of the controller, its setup maybe 2 feet further (of tubing) than it could be,
and theres a reaction setting on the HKS EVC-S that i think stayed at the default setting

and if you look after 5500rpm, both boost curves look exactly the same

[cobrAA]

maybe you should try to use a profec ?

[GTRADDICT]

Your boost curve is all over the place. I suspect something leaking air or the boost controller setup needs big adjustment. You should have done a run on wastegate pressure.

What management/tune do you have? Did you tune/adjust for the increase in boost?

Where was the af sensor placed?

+1 for boost leak

[cortexx]

I had similar situation with my 3000gt after decatting the dump pipes and installing a much better exhaust.

The problem was the volumetric efficiency increased because of a less restrictive exhaust and also some intake mods, this means that for the same PSI the turbos had to deliver a much higher volume of air to keep the same boost level througout the rev range , this caused intercooler heat soaking that caused detonation (knock) making the ECU pull advance and go into limp mode. It took AGES to figure this out because it was not obvious. Do you have anyway to measure the knock under boost ?

The fix was to replace intercoolers with larger units and decrease the gain boost controller very slightly so that the 9b turbos didnt quite spool so quickly.

You may also have to adjust the blowoff valves a bit tighter , they may be bleeding under the higher boost, had this problem with a greddy unit due to improper adjustment.

Also the vac tubes should be as close and as short as you dare mount them to the heat of the turbos. If you have to run them any distance do it with hard lines ( i have used metal brake lines for this before )

[RBdreamer]

Don't wanna ask the obvious, but, did you do a compression test on the motor ? Wet and dry ?

I noticed reading in your build thread that you got all house-wifey and ran a cleaner through the system that puffed white out the exhaust ? sometimes, old rings don't like that soft of treatment. You should really check out the basics.

[frankiman]

i did a compression test after running that cleaner twice and still got 158psi +-3 across all cylinders

i can't see the problem beeing with the intercooler, during the dyno runs, i touched the cold side of the Trust intercooler and it was scary cold
also, i'm running the same exhaust as before, but with a decat

i checked the SSQV and they don't seem to be leaking, and they aren't adjustable because of their piston design

tonight i'm going to try and relocate the boost controller selenoid closer to turbos, may be able to shorten the vaccum lines by ~2½ ft.

then the car is going to autoworx on thursday, i'll ask them to do a full boost leak test, then take the old HKS race BOVs off and weld the flanges shut

i'm also going to read into the HKS EVC-S selenoid release speed option

i'll probly go back to the dyno in the coming weeks to see if anything has changed

[gtrbubba]

Boost versus torque?

It would be easier to analyze if you had a graph of boost versus torque, as that is what is actually relevant and directly related to boost levels. (power is just a function of torque and RPM)

It's very unlikely to be a boost leak as you have an early boost peak, and dip, and then another peak soon thereafter. I'd put money on the boost controller going weird (or something telling the boost controller to back off). The dip and peak are too close together to have anything to do with "volumetric efficiency".

Something to try on the dyno would be quicker pulls in lower gears to see if the basic effect is the same, but on a tighter timeline. My guess is that it might disappear, showing a potential heat/detonation issue with a slow(er) 4th gear pull.


Also, at high boost levels, unless you have mammoth turbos and an incredibly efficient intake/exhaust you will invariably have the boost slowly decline throughout the rev range. This is completely normal, and simply an issue of not being able to move the mass of air through the engine required for a given RPM and boost.

Commonly, in a typical stock/mild turbo application the boost "curve" will be a slope up and a flat line all the way to redline, showing that both the turbo and the intake exhaust system are running below maximum potential efficiency. In a lot of cases with small single turbos, a boost tweak will only effect the low-end boost as the turbo "runs out of breath" as the RPM rises.


Number one reminder; torque is everything, forget about "power" it is of no relevance.

[Skym]

Dropping off at higher rpm can sometimes be traced back to A/F ratio or BOV's are leaking. As far as I know, R32GTR turbo's usually don't run out of puff / drop off until over 430ps at engine.

Also A/F ratio can affect torque, power, boost level if you compare them. Engine lean, boost level drops. Engine richer, boost level increases.

Sometimes a sharp change can be stock ECU switching between main, knock maps or switching to different cell on fuel, ignition map. Stock GTR ECU when compared to good knock device picks up knock at same time. If you are able to monitor knock count on ECU and know at what knock count level ECU switches over to knock maps, should be obvious. With Nistune tuning software, can datalog what cells ECU has used and with trailing on, see where it's been on fuel, ignition map. Also datalog knock count, etc.