Boost Leak Test Correctly?
 

Hi all,

So I am wanting to make sure that I am doing a boost leak test correctly, as I have been having a lack of power issue.

I have made a basic boost leak tester which goes into the stock intake (post MAF) and has a tyre valve and boost gauge. I then block the line going to the rocker cover and regulate the air compressor to 25psi. I have not blocked off the throttle body or anything else.

When I try to connect compressed air or even a smoke tester, there are no leaks to be found, but the pressure doesn't hold.

I even tried this method: http://www.mazdaspeedforum.org/forum/foru...ranteed-94843/

Am I missing something or doing this incorrectly?

Also to give some background, I have a stock intake, but a downpipe, Autotech HPFP, and a Cobb AP V2 and Corksport BPV. Other work done includes JBR TIG for TB and Intake, valves have been cleaned, 1 step colder Denso plugs, new NTK O2 sensors, fuel pressure regulator valve, Toyota injector seals, all clamps changed to solid band hose clamps, PCV Valve and Damond EGR Block off. The previous owner had changed the timing chain, VVT Acuator and removed and polished the crankshaft with new race bearings.

When I have started the car today and taken it for a drive, the FT was around -11 at idle on stock tune and if changed to a Stage 1+ MAP, FT changed to -15. So thought I would do a MAF calibration, but I was watching the AFR and at cruising, it went as high as 32, with a 28 constant, but no Knock. So to avoid ZZB it is on jacks in the garage, so I won't be able to get a datalog.

If your car is pulling fuel on both the stock tune and on the Stage 1+ map at idle, that means your car is running rich at idle. That means that you've probably got a substantial boost leak. That is, air is escaping from the system before passing into the engine, and the exhaust constituents are low in oxygen. Thus, the O2 sensors would report a rich running condition.

Given that you say you're running 28-32:1 AFR at cruise, I'm a little confused though.
As far as I'm aware, the leanest possible AFR is 29.3:1. I want to see the data supporting your statements.

Do you have any datalogs showing your car's symptoms? If so, please post them here.

Thanks for the response, for the datalog, my fiance was checking the AP, so I am going off her numbers, but I will get back later and take a datalog and post up.

In the mean time, is the way I am doing the boost leak test right? I can't even hear any air escaping.

http://www.mazdaspeedforum.org/forum/foru...-leaks-138722/

so i was reading this and i think it may apply... so post turbo leak to the throttle body can make it rich under vacuum, correct? @Vansquish;

also a exhaust leak throw off #'s as well, right?

Hey all, wanted to update, I watched the AP this time and the AFR stayed around 14.7 at cruising, didn't go higher than 29.3 under acceleration, the only is the LTFT's, even at idle it is around -18 (datalog attached). I also tried to get a 3rd gear WOT datalog, for some reason on the AP it shows as been 99% for throttle position and a max of 12PSI, but on the datalog it shows 70% throttle position :/ (also attached). I have also attached a 2nd gear MAF Calibration datalog, pretty much had to hit redline to get 100g/s. This is all on a Stage 1+ MAP

OK. I've taken a look at your logs, but before I get to them, here are a few comments based on your post only.

1. AFRs of 29.3 should only be occurring when you're off the throttle. That is, they should spike to 29.3 during shifts, but under acceleration your AFRs should be between about 10:1 and 14.7:1 depending on the conditions.

2. Accelerator pedal position (APP) and Throttle position (TP) are not the same thing. It sounds like you are describing APP which can go as high as 99% at WOT, but WOT is only 70%-ish for throttle position.

3. If you're doing a MAF calibration run properly, then you probably won't get to 100g/s until quite close to redline. It's easier to do a MAF calibration log in a higher gear, as loads increase substantially with gear ratios.

4. The OTS maps are almost universally awful. After COBBS V100-series OTS maps, the tunes have all been neutered and weak. I'm not surprised that you're only hitting 13psi.

Now to the datalogs themselves:

1. You are risking serious engine damage the way that you are doing your WOT runs. It is well known that going WOT at very low revs on an undersquare engine is a recipe for bent connecting rods and blown motors. On our platform specifically, there is almost universal agreement that you should NOT go WOT below 3000rpm. Period.

2. At WOT, your car is adding fuel at a substantial rate to try and make AFR targets. That is, you appear to have a pre-turbo, post MAF sensor leak. However, you are hitting your AFR targets at WOT.

Additionally, your extremely short idle log (I mean, really? You expect us to diagnose something in a 3 second-long log?) suggests that your car is running rich at idle the entire time, as the ECU is pulling fuel for the whole log. This suggests that your engine is receiving less air than expected for the amount of fuel being injected.

Since the WOT, MAF Cal and Idle logs all say slightly different things, I would guess that you may just not be using a map that is appropriate for your modifications.

3. Your MAF calibration log seems to support this, as a significant amount of the log, up to around 4200rpm, shows that your ECU is trying to pull fuel to meet AFR targets. That is, you are running rich to that point. Above about 4200 RPM, your car suddenly starts adding fuel to meet AFR targets. You're in vacuum, rather than boost for pretty much the entire time.

If you had a vacuum leak, your car's behavior would likely be a bit different. With a pre-turbo vacuum leak, when your car was in vacuum and under boost, you would be running lean, and the ECU would be adding a bunch of fuel to the mixture. With a post-turbo vacuum leak, when your car was in vacuum you would be running lean, but when your car was in boost, you would be running rich (air would be escaping the system).

Your car has the opposite symptoms. You are running rich in vacuum and lean at WOT. Therefore, it is not a typical pre or post-turbo leak. Accordingly, you probably are running an inappropriate tune.

Thank you for your detailed and insightful post, for the actual datalog it is especially great to know not to go WOT under 3k, I have been reading all kinds of conflicting information.

Just in regards to the Boost Leak Test itself, is the method described in the first post correct?

I have seen videos where people have said they can build steady pressure, usually around 15 Psi

15.5 psi (ish) is stock peak boost. With your downpipe (what are you doing for a racepipe/midpipe?) and racepipe it will want to build to 18 psi on the stock tune. But that is with dangerous spikes to 22 or more. The stock MAP sensor will max out at 22.5, so you could be higher and not know it. With a proper tune, you’ll see 17-17.5 psi or so peak with taper toward redline.

You need a good midpipe (catless - called a racepipe) as the stock one has a restrictive catalytic converter and is not 3”. It’s ok (actually good) to run the stock CBE because gasses have cooled enough by then that it flows well. But, coming out of the downpipe they are still hot enough that they benefit from a 3” diameter until they cool a bit more.

There is NO confusion about going wide open throttle below 3,000 rpm. Don’t do it if you value your block. Bent and broken connecting rods will help you ventilate it with nice big holes in unwanted locations.

Hi MSMS3,

Thanks for the input. For clarification, the circuitwerks is a racepipe by definition then, it has replaced the existing stock cats, I will correct this in my sig, thank you for the education lesson.

In regards to the PSI under driving conditions, I am actually peaking at 13PSI with mods and tune.

In regards to my previous post, I was more referring to the system holding 15PSI, like in this video: https://www.youtube.com/watch?v=bN8f4PWfvnA. I am not getting any pressure holding at all, so I am not sure if I am doing the boost leak correctly?

Are you sure about that Circuitwerks product? Do you have a photo? Our upper pre-CBE exhaust is two separate pieces with two separate cats. Most aftermarket downpipes only replace the downpipe and not the midpipe. The catted midpipe gets replaced by a separate straight open pipe (racepipe). Some companies sell matching dp/rp sets, but I’ve never seen a combined one piece dp/rp. I’m not familiar with Circuitwerkes. It must be a new product for our platform. Certainly not something commonly encountered here. A photo would answer the question.

I have attached a picture of the Circuitwerks dp/racepipe for reference, which is in place of the stock 2 cats. I have also tested it for leaks (smoke test and soapy water test) and the welds and flex are all good.


In regards to the boost test, are we meant to block off anywhere, other than the hose going from the inlet pipe to the VC?

I see that it is one piece. Thanks. Haven't seen one like that before. Where are the bungs for the primary and secondary O2 sensors? Looks like there might be one right on top of the downpipe arch, but I can't tell for sure. And the one for the secondary sensor downstream?

It's been a while since I've done a boost test and I sold my car and the tester with it back in January. But, I don't recall blocking off anything because my tester was usually connected directly to the compressor side of the turbo, bypassing the TIP for the test. My tester was made to fit to the turbo.

There is a bung before the flex and one after, I've actually attached another picture for clarification.

Ok cool, I have tried both methods for the boost leak test and I am getting the same result.

Do you remember if you regulated the amount of air via the air compressor (eg set it to 30psi)?

I do remember this much:

1. Yes, I used a regulated air supply from my air compressor, setting it at no more than 25 psi. My tester connected by silicone hose directly to the compressor side of the turbo. It also had its own gauge. I used both my air compressor gauge to make sure my incoming air was not too high to damage anything. 25 psi is fine. My tester had an NP 1/4 inch male air hose fitting so I could connect directly to it with the hose from my air compressor.

2. I then would pressurize the system with air I knew to be at 25 psi and watch the gauge on top of the tester to see if it was building and holding pressure. If there was a leak, I would usually hear it and the gauge on top would have trouble getting to 25 psi if at all. A bad leak will usually show up by the time you get to 10-12 psi. I could then must squirt water on the area where the air sound was coming from, see the bubbles and fix the problem. Always was a clamp. I never blew a boost tube the entire almost 11 years I owned the car, although others here have.

I forgot to ask @Vanquish, how do I know what is an appropriate tune then? I am trying the tunes that match the mods (even given the Cobb OTS maps are universally horrible), but the LTFT seems to be getting worse? I have attached the latest logs

to tag people, add ; at the end instead of , @BR1JZL;

Are you sure you don't have an aftermarket intake on the car?

Noods of your engine bay

Yeah stock intake, including the stock pancake TIH. And because you asked so nicely, fresh noodes.

So I also decided to check timing, and pretty sure the timing is out

Where's your intercooler shroud?

Took it off for the work and pics


Sent from my iPhone using Tapatalk