Here's my problem. This is my setup from last year: https://honda-tech.com/forums/forced-induction-16/dyno-e85-gt2871r-84mm-gsr-b16head-2035882/

I added a video of my boost leak test I just did. it sounds like the air is leaking out into my turbo manifold. Could a bad valve lash cause this kind of leakage? http://www.youtube.com/watch?v=2gu4xC_j1qU

my first tune last year with a gt2871r .63ar, 20psi made more power than my tune last week with a gt3071r .82ar at 25psi? What gives?

Here is a the overlay of the 2. 2871 is the blue line


Here is a comparison of both showing boost levels. Look at what happened when we bumped it up to 25. Shouldn't I be making more power than my previous turbo at a higher boost pressure?


This graph is all of the 3071, at low pressures the boost creeped a little but holding. at 25psi, it drops like a rock.


For some oddball reason my afr's were going really rich towards the top. My tuner mentioned the way the map was setup, it shouldn't be doing that. What could be causing this?

 

seems to me its too rich up top, you see the afr's go almost to 10:1 and your power is dropping off like a bad habbit...

has your tuner tried leaning out the top end of the map at all to correct the richness of the boost increase...

if its only tuned for 15 psi, when you bump up the boost to 25 you have to still tune that part of the map it should be making a lot more power then that, not dropping off

 

Complete a leakdown & compression test to ensure the motor is in healthy condition still.
Also, complete a boost leak test, something is wrong...

 

i'll try doing a compression test and leak down. when doing a boost leak test, how long should my motor be able to hold pressure? if i pressurize my system to 20psi, how long should it take before it bleeds back down to 0? Also, is there suppose to be a gasket in between the compressor housing and the chra? it looks like its metal to metal with no o-ring or gasket?

 

Which GT2871R do you have? It is possible you have the same compressor wheel as the GT3071R. If so all you have done is open up the turbine side which explains why the boost builds 600RPM or so later.

The difference in total power could be explained by many reasons, some already stated. Those being weather conditions, richness, boost leak, poor compression and overall motor health etc etc.

What difference were you expecting to see power wise?

 

I believe the 3071 has a larger compressor wheel than the 2871. I just don't see how the total output or the powerband could be worse than before. I'll run a couple of checks later today to see if I can find out anything. I'm really hoping it was just a massive boost leak or something I was thinking this setup at that boost level should have been around 475hp and the powerband holding better.

 

Certain 2871's share the same comrpessor wheel as the 3071.

Same dyno, same day, same operator, same software? The list goes on.

 

Same dyno, similar conditions but different day, same tuner and software. Nothing really has changed. It looks like my compressor wheel is the same but the 3071 has a larger turbine wheel. Shodan? turbo geniuses please chime in.

3071
Turbine
-Wheel: 60mm w/ 84 trim
-Housing: .63 ar, .82 ar, 1.06 ar
-Maps

Compressor
-Wheel: 71mm w/ 56 trim
-Housing: .50 ar
-Maps

2871
Turbine
-Wheel: 53.85mm w/ 76 trim
-Housing: .86 or .64 ar
-Maps

Compressor
-Wheel: 71mm w/ 56 trim
-Housing: .60 ar

 

The fuel maps should taper down to correct the afr issue. 10:1 is to rich for my likings

 

No, nothing at all.... lol.

Good luck.

 

Its your AFR's. Look @ 6200 rpm...your 3071 is ahead of the 2871 curve. Then the AFR's dip down so rich and your losing power. The 3071 is making more power until you go pig rich on the topend. You need to lean it out and keep AFR's in the 11-11.5 range after 6200 and youll find your lost power.

 

I added a video of my boost leak test in my first post

 

Was that a leakdown test? What are you putting pressure to?

w/e you are showing us is definately not good...lol

 

this was a boost leak test. i was pressurizing the turbo inlet. I'm hoping that a valve lash will stop the leakage into my exhaust mani.

 

Unless the intake cam was removed, atleast one set of valves will be open, air will just escape into the motor and out the exhaust. The only true way to do a leak test for just the charge pipe is to cap off before TB or take the intake cam out, then you will be able to check leaks in the IM too.

 

Then wouldn't I have to take out both cams then? That is way too much work! hahahaha

 

cause if none of the intake valves are open, it doesnt matter if the exhaust valves are.

 

Just cap off the end of the charge pipe before the throttle body than test it again. If there is no leaks than it should hold the pressure steady.

 

I thought about something. my tial wg is recirc'd back into my exhaust. but its not very far from the turbine housing. like maybe 12" downstream. Do you think there is increased turbulence from that causing the turbo to lose steam? But I always thought that at higher boost levels, the wg actually bleeds off less boost though?

 

No. For the wastegate, the turbulence wouldn't affect higher end power. Being further downstream is best, but it must come in at a very SMOOTH angle. It seems you need to check that AFR first, and then check your leakdown.

I have updated specs of both turbos on another thread. One more recent. It can be easily found.

 

maybe this theory...I currently have a 10psi spring in my tial. I think that the exhaust pressure at that rpm and engine load may be too great to keep the wg closed causing me to lose boost and power up top.

 

get a 1 bar spring or greater

 

I would run atleast a 1 Bar spring

 

Or a boost controller to keep that valve down until needed. I forgoat TiAL hates boost controllers....

 

i'm controlling my boost with neptune using pwm. but i've been doing alot of research on issues like mine with people trying to use weak springs and run large amounts of boost. the backpressure is still too great forcing the valve open causing an "exhaust leak".