Common vacuum leak points?

After finally getting my Speed6 back to its previously working state (thanks big dill!), I’m back to another issue that I’d really like to solve (which I thought had been solved...).

When I cold start the car it’ll start but immediately die. This will happen repeatedly until I can catch it with the accelerator and hold the rpms up for maybe 10-15 seconds.

I had previously tried to do some basic vacuum leak troubleshooting—such as spraying carb cleaner at points, checking the tightness of various parts, etc.

I was wondering if there were either common failure points or an easy place to test? If I were to try a basic smoke test where would be a good place to put the smoke in?

Thanks!

 

Vacuum leaks are a pain, and I am pretty sure I have been putting up with one since I rebuilt my engine.

How did you diagnose your vacuum leak?

I know you did a head gasket: when reinstalling the intake manifold, it is critical to clean the surfaces thoroughly and carefully. In addition it is important to tighten in the correct pattern to seat the gasket well. IF you didn't do this carefully you can have a vacuum leak there.

A better way to test this is to use acetylene out of an oxy-acetylene torch and move it around suspect areas.

 

Thanks! When you said about the tightening order I realised that I absolutely did not do that I forget why, but I’m planning on doing a compression test tomorrow and while the intercooler is off I’ll retighten the intake manifold.

Thanks for the tip on the acetylene. Do you know a cheap source for that?

 

Re-tightening the intake may not be the solution, gasket replacement may be necessary. checking its fastened down well shouldn't hurt though. Don't get too hung up on this, a vacuum leak can come from many places.

My cheap source for acetylene is the shop torch where I work part time.

 

I apparently have one of the AP TIG gaskets, though I have reused it more than a few times (not sure how many). I had actually replaced it with an OEM gasket back in Dec because I had no idea what that white ‘plastic’ gasket was. I put back in the TIG gasket once I learned what it was. The OEM gasket looks in great shape still, but I was hoping to keep using the TIG one which also looks decent as far as I can tell.

 

Don't guess by the looks, there is a reason the manufacturer tells you to replace the gasket every time it is removed. You may get lucky and not have issues, but its not worth it.

Also I have heard mixed reviews on the TIG. I would remove it and install a fresh gasket with proper install procedure just to eliminate that from being a problem.

 

Thanks I went ahead and stopped at the dealership this morning to order another gasket. He gave me a discount so I ordered two (who knows when I’ll take off the intake manifold again chasing problems...).

I know the gasket may not be an issue, but at least I can have some peace of mind. I’ve had the intake manifold off many times over the past few months!

 

Reminds me of how good I am at doing steering racks on the 3s haha.

Take a picture of the gasket surfaces before you put it back together, it should be very clean and free of debris/imperfections.

Do you still have your VTCS?

 

I do pretty well cleaning the surfaces but I’ll take a photo when I do it next. I decided I’m going to wait to replace the gasket until I replace the turbo because there’s a reasonable amount of oil from the turbo and I’ll be tempted to take off the intake manifold to clean the oil up. I might start on the turbo tomorrow after I do the compression test.

The intake manifold that came with the car had the butterfly flaps taken out, but I replaced the manifold because the EGR threads were stripped and the replacement manifold has them.

 

Two things:

1. When I started it up today I noticed during the very beginning of the warm up that it was doing some light backfiring. Not sure if that is related to the difficulty starting.

2. I checked compression and it was:
Cylinder #1 - 148
Cylinder #2 - 165
Cylinder #3 - 150
Cylinder #4 - 0 (no movement on the gauge at all)

I’m pretty sure the poor compression doesn’t have anything to do with the starting issue because it was starting fine for a month or two when I bought it even though it had no compression on 4 (and low compression of 3, which might have been improved when I added the Engine Restore with two oil changes).

Here’s a photo of the bottom of the head for cylinder 4 from when I replaced the head gasket. The cylinder and piston actually looked ok in that they looked cleaner than 1-3. The #4 spark plug also shows signs of wet oil every time I take it out—except for today, strangely, it looked exactly like the others with no wet oil. Not sure why. But would the oil indicate the piston ring rather than a valve?

 

Even with a chipped valve, I was getting some compression reading. 0 sounds pretty severe and from what I am seeing it is from the bottom end. You would likely have some pretty unstable shaking from the engine running like this.

You may want to diagnose why you have low compression. You can put a small amount of oil into cylinder 4 and do compression test on that cylinder again. If you see numbers increase, it is pointing towards your rings not sealing. I did this to diagnose my valves successfully. IF you still see low numbers and you are certain your valves are whole and sealing, look into a cylinder leak down test.

Seeing as cleaning the plugs may have solved some of your issues, once it started firing again is when it stopped collecting fluids on the plug. I wouldn't draw any conclusions on the plugs yet.

 

Thanks big dill!

What would you describe as unstable shaking?

I wish I had thought of putting oil in the cylinder when I was doing the compression test. I have the intercooler back on and I have some work to do on my old Hyundai tomorrow.

I’m still vexed about the difficulty starting though. Today it started twice and stalled like normal, but then wouldn’t start anymore at all. I didn’t have time to keep at it or investigate.

 

You will feel the whole car shake, especially when on the gas. Your engine light should be telling you there is a miss-fire in that cylinder. Wondering if you have an intermittent problem.

A possible symptom of low compression is difficulty starting.

I wouldn't bother putting the intercooler on until I figure out whats wrong with that engine.

 

Ok, it doesn’t do that. If I keep the rpms above 2500 you wouldn’t really notice there’s anything wrong once it’s warmed up (unless you knew how it was suppose to perform). The misfire code is very intermittent, not even every time I drive it.

The only reason I was dismissing the compression as a reason for trouble starting is that it happened a couple of months after I bought it and it’s had no compression on #4 since I’ve had it.

Though I’m committed to investigate the compression more thoroughly. Thanks for motivating me!

 

Big dill, if you're interested I made a recording of the engine accelerating from start. It's clear enough to get a feel for what it sounds like. Maybe it'll mean something to you. I do a little acceleration at the start (I was turning left at a light), followed by a stronger acceleration in 1st and 2nd.

By the way, I said it doesn't shake, but maybe that's not totally correct. If I'm at around 1500 or so and if I press hard on the accelerator, the car does not like that at all. But above 2000-2500 rpms it seems to accelerate fine (though probably weaker than it should be).

http://files.rjsilva.com/mazda_acceleration.mp3

 

Something isn't right, with such a low compression in cylinder 4, it should be a pretty rough ride. Maybe boost pressures are sealing past 2500 rpm?

See if you can investigate with a leak down test or check your oil quality/coolant. Based on the information you provided there is a leak into the cooling jackets.

 

Have you done a cylinder leak down test? If your rings are bad you'll hear air inside the crankcase, if the intake valves are bad then you'll hear it in the intake, and the exhaust if the exhaust valves are bad.

 

Thanks big dill and l337turboz.

I’m pretty sure there is no coolant/oil mix. I’ve changed the oil several times and it was fine and the coolant is the normal green colour.

Is there a way to do a leak down test without an air compressor? I have a friend who has one and could borrow it but I don’t have one myself. I have bicycle pumps which could easily reach over 100psi...

 

I dont think a bike pump could work that well. There is cylinder leak down regardless of a failed component, so you may not be able to create the volume needed quick enough for a proper test.

You'll also want to have a gauge that'll show you the percentage of leak down and regulate the amount of psi inside the cylinder.

Like big dill stated it probably is a leak with the cooling jackets. By doing a leak down you'll know which way to go. Or you could just buy another engine and swap that in to save time. You may end up having to do that anyway.

What do you use to clean surfaces? I only use razor blades, no power tools or abrasive discs. You said you replaced the HG but did you send the head off to be pressure checked and have the deck surface redone? Did you measure the deck warp of the block and the head?

 

Thanks L337TurboZ. I did find a decently reviewed leak down tester for about $40 on Amazon. It’s no problem to borrow the friend’s compressor so I’ll plan on that. Hope that will turn something up.

As far as cleaning the block/head, I used a Mahle plastic gasket scraper followed by some purple scotchbrite. I didn’t send the head anywhere or do any measurements. In retrospect I wish I had done more when it was apart, but I was simply going on the symptoms and what the seller told me.

If there was a leak from the cylinder to one of the jackets wouldn’t there be another sign? Especially if the leak is so bad that there is no compression? Seems to me I’d either be burning coolant or possibly getting dirty coolant. I’m no expert though.