I caused/found/fixed various boost and exhaust leaks, but boost still settles around 0.6-0.65 once the engine is hot. So weird, since it’s initially bang on.


My IATs were only ~33 degrees C, about 5 C above ambient. My water and oil temps were both in the 70s and 80s. Only had a little blowby.

Assuming there isn’t one last intermittent exhaust or boost leak I can’t find, I think it’s the wastegate. Last resort is to replace the spring (cheap) or get a new more modern gate. Would require welding a new flange, though.

As it stands, my car is basically undriveable in the wet. Hahaha.

On another note, the Bando WTAC Soarer was in the next pit from me.

 

I finally got around to installing a new wastegate spring. The old one was a full 10 mm shorter than the new one. The paint was a couple shades darker than the new one, as well (the old one also had some of the paint scraped off, but I doubt this was indicative of anything).

I am looking forward to see what difference this makes, if any.

I imagine it will be the same symptoms, but at a slightly higher boost level.

 

Peak boost once the engine is hot is still about 0.1 bar (1.5 PSI) lower than the target.

My IATs on a cold day were only 15 C or so while the car was at speed. About 10 C above ambient.

Oil and water temps were perfect.

I can't think of anything else to do, other than run the car on a dyno.

I guess I will call Greddy and ask them directly.
And I think I might pull the manifold and inspect it over the winter, since I never actually got around to doing that, and replace the gaskets while I am at it.

 

There were indeed a few small cracks in the manifold.

I wish I had noticed them before I sent the wastegate to Greddy to get it checked, since they said there’s nothing wrong with it… That was a waste of 100 bucks.

There was also a bit of soot on the joint between the compressor and turbine housings, so I wonder if I didn't tighten the v-band clamp properly. I learned you need to tap it with a hammer after torqueing it 70 percent of the way before you torque it all the way.

Now I'm just waiting on new gaskets for my turbo and wastegate before everything can go back together.

 

Womp womp, it wasn't the manifold cracks...

I hit 0.76 bar max when the car was merely warm, and it settled at about 0.64 bar.

IATs were about 37 celcius.

This was in fairly hot weather.

I am starting to think this is just how this particular turbo/intercooler setup behaves, since the wastegate signal comes from the turbo, and the boost signal comes from the manifold.

I wonder how much HP a TD06-25G makes at 9 psi?

I don't want to just crank the boost up, since the car is still faster than I am....

 

"I wonder how much HP a TD06-25G makes at 9 psi?
"

here's the map from back in the day. you will need to convert cfm to pounds per minute

 

my sheet says, 9psi at 7000rpm is 40lbs/min, 313hp. we do not know VE, so my estimate is high.
with the book value pressure drops for the air filter (0.2psi) and IC (2psi), the Pressure ratio is 1.633
at 11.5:1 AFR and stock fuel pressure (38psi) you should be right around 85% injector duty with stock injectors. if its not, this would give us a clue to what the VE is (or if the fuel pump is not good enough)

you want an exhaust that is at least 70mm, preferably 80mm
fuel pump needs 200lbs hour or more

 

according to a web calculator 600 cfm is 45.9 pounds.... 344 rotary rwhp

 

I have a fairly large intercooler (it's an old front-mount that got chopped up into a v-mount), stock injectors, and a SARD pump and FPR, so fuel pressure is increased over stock.

Years ago I did one pull (without a cooling fan for the intercooler) on a VERY old dyno and the uncorrected numbers were in the mid to high 200s, I think. I was given a corrected number of 345 or so, but I'm not sure how much I trust that.
I don't know how much boost it actually made during that pull, though.


The car was originally tuned for 0.8 bar (spring pressure), and considering it's a fairly sizable turbo, I question whether running such low boost isn't just making the car harder to drive than it needs to be. You really have to work really hard to keep stay in the powerband.

 

have you dropped the exhaust to look for melted stuff?

 

I pulled the manifold and turbo for inspection during the off season. All I found was two tiny cracks right inside the turbo flange, which I had welded up.

Nothing seemed to have been melted. I did notice some soot on the mating flange between the compressor and turbine under the v-clamp, but I chalked this up to me not tightening it properly the last time I took it off. I figure if this was really leaking it would have left a huge mess inside my turbo blanket.

There still might be a small boost leak when the engine is fully hot, but it’s hard to test for that (without a heat gun, I suppose).

I suspect the turbo-intercooler elbow, which doesn’t fit right since it wasn’t designed for this intercooler. It’s sort of jerry-rigged on there using every spare mm of the silicone couplers, with the hose clamps at slightly weird angles.

I might go back to my old couplers since they are longer. I already stopped using my new clamps because they’re too slippery. This made it hard to position them just right and apply sufficient torque.


For the record, I just did an on-car bench test, and when the car is merely warm, the wastegate cracks at almost exactly 0.8 bar. I’m going to try the same test after doing a few laps next track day and see if it’s not just the spring relaxing.

 

Hmm. I just did another boost leak test, but this time I actually looked at how much pressure the manifold was seeing.
Turns out, I had previously been applying significantly less than the target boost pressure. Maybe 3 or 4 PSI.

This time, I was able to directly use my compressor instead of a portable air tank, so I ended up applying several times as much pressure as usual, hitting over 14 PSI.
I immediately noticed a difference because it caused the doubled-up shopping bags I had used before to pop. This had happened before.

Guess what.... one of the post-intercooler couplings started leaking after it hit about 6 psi. Well, that explains a lot...

After getting that one to seal, the other one started leaking instead.
I ended up having to tighten the hose clamps much more than I thought I would to get them to seal properly.
I had been wary of overtightening them, because I bent the pipe at one point.

I may end up having to get the pipes worked on so they fit better. I should probably get an entirely new turbo-intercooler pipe made.
It doesn't detectably leak, but it does produce a tiny amount of foam when I do a soapy water test. I actually was surprised that this wasn't the pipe that was actually leaking.
This is probably since it's a cast piece, so I never have to worry so much about deforming it with the hose clamps.

I'll be able to see whether this fixes it on Tuesday.

 

I replaced the troublesome u-bend with a 135-degree flexible silicone elbow from Alibaba (among other parts).

Don’t laugh, it was the only place I could find the exact right size on both sides (63 and 70 mm).

I tapped the boost signal with a quick tap.

So, at today’s track day, I was only getting 0.55 bar, which is even worse than usual.

I then hooked the compressor boost signal directly to the boost gauge, and the turbo was producing almost exactly 0.8 bar consistently.

I then tightened the cold side hose clamps, which I could have sworn were already very tight, and the boost jumped up to 0.65 bar (peaking at 0.7 before settling at 0.65).

Hmm.

So, the problem is definitely a combination of intercooler losses and/or boost leaks. I don’t get why the cold side pipe doesn’t want to seal unless tightened to an inch of its life.

 

What kind of hose clamps are on your charge piping? If they are just standard duty hose clamps you would want to upgrade to Tbolt style or similar to get proper and even clamping force.

 

I have t-bolt clamps on the pipe I am having problems with (the cold side pipe).

I wondered if maybe it was because I deformed one end of the pipe (by over-tightening a worm gear clamp on it) years ago, but that doesn't seem to be the side that's actually leaking.

I still wonder why I need to tighten the t-bolt clamps so damn much to get them to seal.

If these were regular clamps, I probably would have damaged the pipes.

 

Check to make sure none of the IC piping is deformed at the coupling ends - the fact that you have to tighten the T-bolt clamps so ridiculously tight is a symptom that the pipe ends might not be exactly round. Check the beads on each pipe end for deformities too, they need to be consistent & uniform as well for a good seal.

Also your comment "I tapped the boost signal with a quick tap." has me questioning the the kind of quick tap used as a potential leak source - if it's one of those taps that pierce into one of your silicone couplers to create a vacuum/boost pressure port, I wouldn't trust it - might be fine as a vacuum only tap, but they typically don't handle positive pressure well.

 

Yeah, I didn't want to have to grind and drill a hole in my compressor housing.

The quick tap doesn't leak at all during bench testing, even at twice my targeted boost level.
Again, the pressure signal was exactly where it should have been (0.8 bar), so the issue is likely further downstream.
A leaky boost signal would actually produce an overboost situation, not an underboost one.

It's possible that a 2 psi loss is simply what this intercooler setup is going to get.

The only thing that *sometimes* actually leaks for sure is the cold pipe, which like you said, I should probably get checked.
I used a piece of wood to straighten it years ago, but maybe running a bead former over it would be a good idea.

I have a 100 mm silicone coupler coming so I can build a proper boost leak tester. I will try to get it up to more like 30 PSI.

I might try pressurizing just the cold side pipe. The intercooler itself held pressure well.

One small thing I noticed is that one of the three brackets on the intercooler doesn't actually match where the bolt hole is, so there's no bolt holding it down there.
The piping itself probably keeps it sufficiently secure with just two bolts, so I never bothered to fix it so I could add a third one.

 

Why are you taking your boost reference from the charge piping instead of using a nipple on the manifold? If you are seeing a discrepancy of a couple psi but your taking readings pre and post intercooler the intercooler is certainly going to have a slight restriction. A few psi is typical.

 

The boost gauge is from the manifold.
The wastegate is from the charge pipe.

Yes, two psi may just be how much restriction there is.

I was told the car was set up for 0.8 bar and the boost controller was only used to improve response, but it’s possible they were taking about the nominal spring pressure and not the actual boost.

 

Until you take the pressure readings from the same source (IE both from the manifold) your chasing false numbers. You need to tap both sources from the manifold. It is most inportant to know the actual boost entering the engine rather than upstream on the charge piping/turbo system. A 2 psi drop is very typical for a common intercooler at lower boost levels, which would likely grow into a larger pressure drop as input pressure is increased.

 

Taking another look at the cold pipe, they're definitely a little messed up.

The one I bent badly and fixed has a flat spot, and the rim still isn't quite round.
The other one is only very slightly off-round if you look at it hard enough.

The beads seem to be fine, though.

I'm going to see if more flexible worm gear clamps if seal better than t-bolt clamps I bought to replace them.

Otherwise, I'm not sure how I can fix them without specialist tools or cutting off the ends and welding new pipe on.

The pipe isn't a great fit anyway, so it might be better to have a new one fabricated.

The Trust and HKS instructions manuals both say the signal should be from as close to the compressor housing as possible, if not on the compressor.
Having the source behind the throttle plate could cause issues.

If 9 PSI is all I get with spring pressure, that's fine, I can always increase the boost.

Longevity is more important than power.

 

I looked further into this and you are correct. Some manufacturers suggest pulling boost controller signal source pre throttle body. I don't think this will be contributing to your current problem, aside from the possibility that where you have tapped is a potential leak source since it is not a proper hard or threaded connection. After reading your most recent assessment of your charge pipes that seems like the most likely source of your issue. When I was still using silicone couplers and flared end hard lines I would pop couplers occasionally and that was with perfectly clean, round pipe ends. Early on I had a pipe that was too long, but could be muscled a little bit to get the couplers on. That was a reoccurring issue until I shortened the pipe by 1/4" so it was correctly aligned. I would try to get the charge piping as aligned as you can, keeping bends as far away from the coupler mounting position. If a bend is too close to the end of pipe, they will oval and not seal.

 

^And the reason why is if you tap from behind the throttle, your boost control solenoid / wastegate will see vacuum whenever you're NOT in positive boost. That's why lots of turbos have a port built right into the compressor housing (at least Borg Warner does)

 

This only matters on a small percentage of wastegates that are intolerant of seeing vacuum. Aside from that, you may be able to get slightly improved boost response if not using a boost controller. If you are using a controller, this improvement is less impactful with the boost controller configured properly.

 

The pipe seems to seal better with worm gear clamps. They’re only tightened normally and I didn’t get any leaks at 16 psi. Ironic, considering how expensive the t-bolt clamps were.

My old silicone couplers may also seal better than my newer ones (also expensive) because they’re more pliable.