I'll try that out as well, just need to either acquire a scan tool or take to place than can help me with this. I have seen all of the discussion about the problems with the fuel pumps with these platforms, and I do have plans of upgrading.

Real quick, whats a good scan tool that can display fuel rail pressure for less than $100? I dont know a whole lot about scan tools...

@joshhuck
The OBDLink wireless/bluetooth dongles ($50-100) are great and come with their own app. I also use the VeePeak bluetooth dongle ($35) with the OBD Fusion app ($6). The OBD Fusion app and OBDLink app are literally the same...but with these dongles you can use any app you like. I personally like these apps because you can data log and also look at live data "dashboards" which you can customize.

@Razmig Bartassian
So, update. It finally threw a code while giving it moderate throttle. The drive was very rough after throwing this code. Shaking not only in acceleration, but while at cruising speed as well. Hooked the scan tool up to it today and it says P0303, misfire in cylinder 3. I scanned it on a warm start, the cold start prior was very rough, had a little trouble starting right up, followed by surging RPMs. Looking at some specific details on the scan tool, a few numbers were all over the place. Fuel pressure varied below 600psi while at idle, and went above 1600psi when revving up to 3000rpm. Fuel trim wasn't constant at idle. On top of all of this, there was a fuel smell from the exhaust. Usually the exhaust was fairly clean. I wasn't comfortable taking it out for a drive to get live data while driving, since it threw the code and felt pretty significant. I am beginning to think that this could either by a bad ignition coil, or a fuel injector. Again, the spark plugs were changed two months ago. Any thoughts? I know this is a lot...

@joshhuck
Alright, so if your fuel pressure went up to 1600 psi then the hpfp is good. I think 600 psi at low idle is normal. To test whether the misfire is caused by a bad ignition coil, swap it with another ignition coil and see if the misfire moves. If the misfire moves, then the ignition coil is bad. If it doesn't move, then it could be the spark plug itself or compression loss. Perhaps the spark plug got "fouled" up from too much fuel, which could be from leaky fuel injectors. The shaking was definitely from misfiring. What are your fuel trims? If your fuel trims are positive, then the ECU is adding fuel to compensate for a lean mixture. If it is negative, then the ECU is subtracting fuel to compensate for a rich mixture. Maybe check for any vacuum/air leaks, especially on the charged-air side of the turbo as well as on the inlet side. You may also want to take a look at your boost levels during acceleration and also check whether your turbo wastegate and bypass valve are working properly.

@Razmig Bartassian
Should I clear the CEL code before swapping the ignition coils? Also, can any of this be done without driving it? Because it did feel pretty sketchy being on the road. The fuel trim that caught my eye was STFT B1, which fluctuated between -3.9 and -3.1, which was during the entire time it was idling. And whats the best way to check for a vacuum/air leak?

@joshhuck
Yes clear the code after swapping the ignition coils so that you can see if a new code will be thrown. You could probably just rev it while it's parked and see if it will throw a code again, but you most likely will need to take the car driving for the misfire to occur again (engine needs load). Driving your car with a slightly rich mixture will not damage the engine. You're just gonna get lower gas mileage. But you definitely don't want to drive when it's running super lean. You're STFT is not abnormal though (normal between -10% to 10%; 0% perfect). But again, you want to see your fuel trims while driving as well to get a better idea what's happening under engine load. As for air leaks, you just need to make sure all the intake/charged-air hose clamps are tight. You could maybe try spraying some carb cleaner around the piping and see if the engine revs slightly higher. When were the spark plugs changed last?

Spark plugs were changed back in November, and I remember triple-checking the gaps. I plan on pulling them out just to see if they look okay, which will be a good time to re-tighten the hose clamps since I need to take the intercooler off anyway. Viewing the data while under load makes sense

@joshhuck
Yes that's a good idea. There could be air leaks on the charged-air side of the turbo (intercooler and it's piping to turbo outlet and throttle body). View data under load is especially crucial on a turbocharged engine. Turbocharged engines operate completely different at idle and under load. Let us know what happens.

@Razmig Bartassian
Will do. Thanks for helping me out!

@Razmig Bartassian
Tomorrow ill be checking into the ignition coils, and I just want to ask real quick how do I know which cylinder is cylinder 3?

@joshhuck
It goes from left to right. So cylinder 1 is the last one on the left side of the engine (the side of the engine that has oil cap), and cylinder 4 is the last one on the right side of the engine (the side of the engine with the air box). So remember to swap the ignition coils by actually UNPLUGGING them and connecting them to each cylinder's ignition plug. I've seen a few people do the swap without unplugging them and it caused even more misfires lol. Good luck and hopefully its just an ignition coil.

@Razmig Bartassian
So I swapped coil 3 with 4 this morning, and took it out for a test drive. Got it to throw a misfire code again, and it was still in cylinder 3... The least expensive thing I can think of doing right now is just changing out the spark plugs even though they are brand new. I did check spark plug 3, and it looked fine. I have pictures of the live data captured at the time of the misfire this second time around if you want to see those.

@joshhuck
Ok, I'd like to see the freeze frame data from when that misfire occurred. I don't think changing the spark plug will help since they are brand new. Let's look at the data and see if there is any clue as to what's causing it. But in the mean time, I would conduct a compression test on all cylinders to make sure you at least have proper compression. From what I remember, Mazda says those DISI engines should have 180-120 psi of compression (180 is new engine). And each cylinder should be fairly the same (+/- 5%).

@Razmig
I can send you a google drive link of the freeze frames if that works for you?

@joshhuck
That'll work

@Razmig
https://drive.google.com/drive/folders/1Rsdnw4MZaHwwx1YaWl8p-V2FCaLF3QJF?usp=sharing

@joshhuck
Ok so here is what I gathered from the freeze frame data: Everything looks pretty normal, but there are a few issues I noticed. From the STFT being -5.5%, it looks like the engine is running slightly rich but it's within the normal range. The spark advance and manifold absolute pressure seem a bit off. This data was captured while the engine was under load, so a spark advance of 12 degrees seems too little for when the engine is revving at 3664 rpm. The spark advance should get higher the faster the engine revs to accommodate for the faster combustion process. So if the spark plug fires too late in the combustion cycle, it can result in really low power or even cause a misfire depending on how close the piston gets to bottom dead center (the lower the piston gets -> the less compression -> less chance of air-fuel ignition). You make the most power when the air-fuel mixture ignites when the piston is at top dead center (the higher the piston -> the more force put on the piston when travelling downward -> more torque). I'm wondering if the VVT work is related to this spark advance. How soon after the VVT work did this misfire issue first occur? Next is the manifold absolute pressure. According to the scan tool, the ambient air pressure was 30 inHg (14.73 psi), but the MAP was reporting 31 inHg (15.22 psi). Again, your engine is under load and at a speed of 3664 rpm, so the turbo should be "boosting" the engine. 31 inHg is hardly any turbo boost compared to what the ambient air pressure is. In other words the turbo was boosting the engine by only 1 inHg (0.49 psi). There should be AT LEAST 3-4 psi of boost at those engine speeds. Can you hear your turbo spooling when you rev the engine? See if your wastegate control arm at the back of the turbo moves in and out when someone else revs the engine. The low boost could be a reason why the ECU is subtracting fuel as seen by the STFT. It could be that the turbo is struggling to build boost while you accelerate, but then it fixes itself. As supplementary questions, what octane gas are you running? And what oil are you running in your engine (brand, viscosity, API rating, etc.)? Look into doing a compression test so that we can eliminate the worst case scenario, and look into doing some detective work on whether the turbo is working properly (bypass valve maybe stuck open, wastegate not actuating, etc.). If you remove the hose off the bypass valve, you should be able to see if the valve inside is working when you rev the engine. Good luck and let us know what happens next. I'll give you a link to one of my posts that talks about turbocharger diagnosing. 

@joshhuck
Here is the link to one of my answers that talks about diagnosing a turbocharger:
https://carkiller.com/scottykilmer/qa/p0299-with-p0420-how-to-fix/#post-29243
Good luck.

@Razmig
Thanks a lot for looking into this as much as you are, i really appreciate it. Its been 3 to 4 weeks after the VVT was done by the shop. Also, this car isn't a daily, it probably gets driven 1 or 2 times a week on average.
For the turbo spooling, I was never sure what these little stock turbos sounded like. But I can tell it doesn't feel like its boosting possibly whenever I rev above 3200rpm. Before this misfire issue, you could definitely feel the car pull harder as it climbed through that rev range on the highway. Of course now, giving the car that kind of throttle makes it shake around that rev range and hardly wants to accelerate. One more note, the idle after a cold start was significantly more rough than it has been before the misfire issue. Kind of reminded me of how the car shook at speed when it threw the misfire code. I run 87 octane regular Shell Gas, with a PEA fuel treatment after oil changes, Castrol Edge 5W-30 (didn't want to deviate from what the manual recommended). Its a stock vehicle, so I dont really do anything different from what the manual recommends. I've thought about doing a compression test as well, just from looking around on Youtube. Is it possible to describe where I can find the wastegate control arm and bypass valve are? If not, can you direct me to something that can?

@joshhuck
You're using 87 octane? Turbocharged engines need to run on a minimum of 91/93 octane. Were you using 91 octane when you weren't having issues during those few weeks after the VVT job? Have you ever used 91 octane before and if so, have you still had this issue? The oil you're using is great though. The black bypass valve is bolted on the right side of the intercooler, and it's very hard to miss. The wastegate is all the way in the back where the turbocharger is, and this one is a bit harder to see because there is a heat shield covering the exhaust side of the turbo. Here's a link to a google drive folder with the locations: https://drive.google.com/drive/folders/1lweJnsMihOyqB5dIkripWz9HJ4xJtejE?usp=sharing

@Razmig
I didn't want to deviate from what the manual recommended, and they listed regular unleaded I believe. I haven't use any other gas outside of 87 since I bought it. I have heard of other people with these cars running 87 and they didn't have any issues, they were completely stock of course. Thanks for the pictures, I'll try to do the test to see if its stuck if I have some time today. Ill keep you posted.

Also, how can I tell if the arm is stuck, or opened or closed?

@joshhuck
That's a bit hard to tell since you cannot see the physical "gate" inside the turbo housing. But the wastegate arm should "pull in" when you rev to close the gate, and it should "push out" when you release the gas pedal. Maybe check this part while someone else starts the car up. It's stuck if it doesn't move at all, but as a disclaimer it may not move while revving at idle depending on how the ECU and boost control solenoid are programmed. And while you're at it, check all the vacuum lines connected to the wastegate, bypass valve, and boost solenoid to make sure they're not cracked or disconnected. Follow all the vacuum lines as well to make sure they're connected wherever they're supposed to be. Also, an easier way to test the bypass valve is by just removing the recirculation hose connected to it (the big one). When you rev the car pretty high and let off the gas you should hear a "whoosh" sound kind of like a blow off valve. If you hear it, then it means it's working AND that the turbo is building boost. If you don't hear anything, then put your hand near the hole and see if air is blowing out (that means it's stuck open).

@Razmig
Will do. Also, about the octane ratings again. All I understand about higher octane fuel is that it is less explosive which Im assuming is better for engines with higher compression ratios.. Is this why running a higher octane fuel recommended for a turbo?

@joshhuck
Yes, exactly. Turbocharged engines run a higher compression ratio, so you need to run higher octane gas to prevent pre-ignition and knock. Lower octane fuel may auto-ignite during compression before the spark, and that's not good for the engine.

@Razmig
Makes sense. Is there a specific way to transition to a higher octane fuel or is it safe to mix with a little bit of 87?

@joshhuck
All you do is fill up with 91 octane at the pump. You can mix different octane fuels, but it's best to fill up pure 91 octane. Mixing octane ratings will just average them out...so like mixing half 87 and half 91 will give a full tank of 89 octane.

@Razmig
Got it. I know im asking you a lot here, and I greatly appreciate it. Just really want to take care of this car right. Ill keep you posted.

@joshhuck
No problem!

@Razmig
The weather hasn't been letting me get outside to test the wastegate or BPV, so for the time being I have been doing some research on what else to check for. I'm still suspecting that maybe I have a clogged or leaking injector possibly. Have you witnessed a case like this in a Speed6 before, or another model, where problems with the wastegate or BPV cause a misfire in one specific cylinder?

@joshhuck
I have never experienced an issue like this on a mazdaspeed (at least with the mazdaspeeds my friends have owned). The only things I have experienced with those engines are the fuel pump, running rich, and running lean. I will say that your issue seems to be cylinder-specific, where cylinder 3 is having either an air, fuel, spark, or compression issue. But when I looked at the data the other time, I did find a few concerning things like the spark advance and the low boost under load. These definitely may be causing the hesitation under acceleration because the engine is expecting higher boost (more air) under heavy load and injecting the fuel necessary for it, but it's actually not getting the right amount of air. And we saw this with the negative STFT. Technically it should throw a code for low boost if it is a lack of air, but I don't think it actually has a code for that (never seen an "underboost" code on those first gen mazdaspeeds even when it really was underboosting). But my 2016 BMW would definitely throw an underboost code when it's off by a few psi because the turbo is managed through "load-based" calculations. This basically means the ECU moderates the engine's peak power using a combination of intake air temps and boost pressure. When it's colder outside, my turbo hardly makes any boost (like 10 psi max) because cooler air has more oxygen and therefore don't need as much boost pressure to make peak power. But when it's hot outside, the turbo spools like crazy (up to 19 psi) to compensate for the lack of oxygen in hotter air and make the same peak power. So you can see why a slight psi difference in this sophisticated setup would matter (the ECU controls the boost depending on air temp, so if it sees uncontrolled boost it will freak out lol). Your engine doesn't have "load-based boost" so to speak...it can make however much boost it can until it reaches a set maximum psi by the ECU (controlled by the boost solenoid mainly to protect the turbo and engine) regardless of intake air temperatures. Now I'm not claiming that your ECU can't tell whether the engine is underboosting or overboosting with a fault code, I'm just saying that I have never seen that code before on a mazdaspeed. They just generally throw rich/lean codes when it comes to lack of boost or too much boost. Anyways, I feel like that's your problem in terms of your overall hesitation, but for the misfire, it needs some more diagnosing such as with a compression test. How often does the misfire occur? You say you have to drive it for a while to get the code. Is there any other history you know about the car like previous repairs or whether it was modified? At 162,000 miles there could be a TON of carbon build-up on the intake valves, causing them to stick intermittently, which would cause a lack of air in the cylinder and a potential misfire. My friend's mazdaspeed3 has 110,000 miles and it had a serious amount of carbon build-up when I checked, so that's another "what if" to add to the list. The best thing to do is to continue doing the more simpler things (like checking the coils) and then work your way up the ladder. If you come across anything else, continue to add it here.

@Razmig
I'm also starting to wonder if it may be a cylinder specific issue too, although I hope not. But back when I changed the spark plugs last year, I did shine a light down into the cylinders and they are definitely dirty. Its pretty black in all four of them, so maybe it's safe to assume that the valves look either the same or worse. Thats what I'm hoping for the most out of this. Ive been able to find a few videos of guys cleaning their mazdaspeed intake valves, and funny enough that their cars had less mileage than mine. 162k is pretty up there... I'll check the BPV and wastegate actuator here soon, when the weather lets me, and I'll return with an update.

Also, the car from what I can tell was never modified. When driving it felt like the misfire would start once I reach around 3000rpm while accelerating, but it would take a while to actually trigger the CEL code. There was one time that it felt like it was misfiring (the shaking) while warming up at idle. I never experienced any of this until a couple weeks ago, when I first mentioned the issue in here. Before, the acceleration felt fine, and drove fine as far as I could tell. I could rev up to 4K and could feel the boost kick in just over 3k. I bought the car in October last year, and still haven't really given it the beans yet, which I guess is good that I didn't in this particular case. All that I could tell from the carfax is that maintenance like oil changes were regularly done. It went in for a thermostat issue twice, but that was over a year ago.

@joshhuck
It may very well be the severe carbon build-up on the intake valves IF compression is fine. You'd be surprised with the issues carbon build-up can cause on GDI engines. But then again, I doubt a misfire in one cylinder would cause a whole bunch of shaking, so it could be that other cylinders are misfiring but not enough to trigger a code. Now this issue does also sound like Low Speed Pre-ignition, and I've experienced it before. It feels exactly how you describe it...when I would give gas at lower speeds my engine would start shaking and hesitate like crazy almost like a really bad misfire. I never got any codes with it, but I did use a dealer level scan tool to find out that it was "superknocking" or LSPI. It happens during low speed, high load acceleration and the reason behind it is the oil. That's why I asked you a few days ago about what oil you use. It turns out in my case, it was my oil that was causing it (Liqui moly of all). I'm now using liqui moly's API SN+ rated oil and I don't have the issue anymore. It turns out that the API did research on this phenomenon and found out that LSPI was caused by oils with a high calcium content as their additive package. Calcium lowers the flashpoint of oil, so the oil vapors getting into your cylinder by the PCV system would prematurely ignite the air-fuel mixture WHILE the piston is traveling upwards (not good) so you can imagine that boosted engines experience this a lot more because of higher compression (going back to the idea of octane). So API SN+ rated oils use non-calcium based additives to increase or eliminate the flashpoint of oil to prevent LSPI. Maybe check the rating on the oil you used last and see if it's SN+. If not, then maybe that could be your problem? But I should also say that small pieces of carbon build-up that get inside the cylinder will also cause this as well. It may be worth a shot to just get the valves walnut blasted to see if that solves the problem. But whether it solves your problem or not, you will still feel a huge improvement after a carbon cleaning for sure. Thank you for continuing to keep us updated...it's turning into a fun little detective game and all the pieces will come together soon hopefully.

@Razmig
You know, now that we are talking about oil, it makes me think back to the VVT job the shop did. I mentioned that I was the one changing the oil, but since they had to take the timing chain cover off, they had to replace the oil, which they never specified which oil they used. They did over-fill it, but I know thats because there is a slow oil leak (more than likely an external leak somewhere around the oil pan or oil filter). This issue didn't arise until after I got it back from the shop. Perhaps I can change the oil again to an API SN Plus oil, while also filling the tank up with 91 or 93 octane, and then see how it behaves. This after I confirm that the turbo components are working properly. I know it could use a deep carbon clean, I just don't have all of the tools for that kind of job, like an air compressor. I could take it to the shop and have them do that specific job while putting in a warranty claim. I do miss driving it, but I agree, there something fun about investigating this. Plus, I'm learning a TON about the mazdaspeed platform.

@joshhuck
While you check out the bypass valve and wastegate, I would also suggest checking the wiring harnesses for both the ignition coils and the fuel injectors. I was looking through a mazdaspeed 6 forum, and someone with a similar problem as yours had a problem with their fuel injector wiring. They used a noid light to test the injector harness, and they saw that when the harness was moved, the light would stop flashing, meaning it was a wiring problem. It's kind of hard to see the fuel injector wiring harness since it's under the intake manifold, but it's worth trying to check those for any physical damage/corrosion.

@Razmig
Got it. Beginning to make a list. I believe i saw in the shop manual something about doing a noid test. With the compression testing, if a cylinder turns out to be lacking in compression, is it possible that that could indicate a stuck valve? Or would low compression simply mean that there is a problem specifically with the cylinder/piston itself, such as a worn piston ring? Also, whats your go to brand of oil for API SN Plus?

@joshhuck
So if a cylinder turns out to be lacking in compression, then it may possibly indicate a stuck valve. Remember that your engine is an interference engine, but carbon build-up wouldn't cause the valve to stay completely open (because if it did, then your engine would have been toast by now). It may just mean that the valve is not sealing correctly when it closes because of the carbon build-up around it. Most of the time, carbon build-up causes the valve to "stick" shut during the intake stroke when it should be open, causing a lack of air in the cylinder. But regardless, if you want to go an extra step to figure out what's causing the low compression, then you need to conduct a leak-down test. So rather than the piston producing the compression while turning over the engine, you use an air compressor to add compression and monitor the pressure (to check whether the cylinder maintains that pressure). The whole point of this is to look for a "leak-down" in compression, and if there is a leak-down you can find out what it is. For example, if you pressurize the cylinder and start hearing air hissing out of the intake manifold, then it means the valves are not closing properly, or if you hear air coming out of the oil cap on the valve cover, then it means the piston rings are really worn out, etc. So a compression test just gives you a basic understanding of the engine's health, whereas the leak-down test is more for diagnosing the cause for low compression. I would do a "dry" compression test first and then a "wet" one to see if all the cylinders are within the compression range (130-180 psi). A "wet" compression test will give you an idea on what's causing compression loss. You add like a teaspoon of engine oil in the cylinder to "seal" the piston rings, so if the compression increases on a "wet" test then the piston rings are a bit worn (which is normal as the engine gets older). But if the compression stays the same during a "wet" test, then it may be a sign that the valves are not sealing properly. And remember that you're looking for ALL the cylinders to be within +/-10% of each other (so for example [150, 145, 148, 130] would show that cylinder 4 has some problem). I personally use Liqui Moly's SN+ oil, and now they even make oil formulations for Asian cars instead of just European. But you can use any SN+ oil...

@Razmig
I get my compression test kit in the mail today, however I don't think the rain will let up... I kind of want to do that first, since this is sounding more and more like an engine specific issue.

@joshhuck
Awesome. Let us know what happens.

@Razmig
Also while i'm waiting for a clear day, when I go to do the compression test I know to disconnect the ignition system and the injectors. But I just want to make sure of where exactly those connectors for those two systems are. Can you provide any pictures to show these?

@joshhuck
All you need to do is remove the fuse for the fuel pump and injectors. After you remove the fuses, try cranking the engine to make sure it doesn't start. As for the ignition system, just unplug all the ignition coils. Remember to hold down the accelerator pedal while cranking the engine to allow in as much air as possible for accurate readings. Here is a link to a fuse box diagram: https://fuse-box.info/mazda/mazda-6-gg1-2003-2008-fuses#2006_2007_2008

@Razmig
So I'm fixing to do the compression test, and figured I'd look at the intercooler. Something caught my eye. ill share the images with you. Does this dirt buildup look like a problem? It smells like oil too...
https://drive.google.com/drive/folders/1Rsdnw4MZaHwwx1YaWl8p-V2FCaLF3QJF?usp=sharing

@Razmig
Here are the dry compression test results:
Cyl 1: 180psi
Cyl 2: 155
Cyl 3: 150
Cyl 4: 165

Is doing the wet test necessary at this point?

Well I went ahead and did the wet test. Results were exactly the same.

@joshhuck
Ok, so those results don't look bad considering the mileage of the engine. Interesting how cylinder 1 has the most compression; it probably means cylinder 1 is getting the most air (proper valve operation) since the wet and dry compression tests were the same. The oily residue inside of the intercooler is a bit concerning. It could possibly be the PCV system acting up (staying open and letting in a whole bunch of oil into the intake tract). That would definitely cause LSPI, especially if the oil is not API SN+ rated too. What is your oil level at right now?

@Razmig
I can't but feel a little relief, but now moving on in the diagnosis process. Pretty much all of the intercooler piping is a little bit oily, and you can definitely smell the oil. A couple months ago I was going to install a Damond Motorsports Stage 1 Oil Catch Can, because as we’ve discussed these motors are notorious for their blow-by and gunk. I quit the project for the time being because a clamp on the stock PCV hose is practically impossible to reach. Could it be the PCV valve that is failing? The oil level has been sitting at just above max. I guess the shop over filled it intentionally because we are all aware of a slow oil leak (good thing I never decided to floor it after that VVT job...) This week I do plan on filling up on 91 or 93 octane fuel, and as soon as we get another clear day I will be changing the oil. I found out that the Castrol Edge oil i usually get is an API SN+ oil.

@joshhuck
Yeah, the intake manifold needs to be removed in order to access the PCV valve. It could be that the shop used a different oil, but also realize that overfilling the oil can cause excess oil to blow by the PCV system especially if the valve is stuck open. These engines are notorious for their blow-by junk because of how low the PCV valve and air/oil separator are mounted to the engine (close to the oil level) whereas modern cars have it mounted on or near the valve cover. But at least we eliminated the compression concern out of the way. It could just be the oil or carbon build-up at this point...but my bet is on LSPI.

@Razmig
And just like an auto shop, they'd want to me to bring it back to them... I can't help but feel sketchy after every time I take my vehicles to a mechanic... Luckily PCV valves are only about 20 USD to replace, and I already have the catch can assembly to install. So we will see how she behaves after an oil change and the proper fuel.
Yup, at least we are getting somewhere. Oh, and now that there is about a teaspoon of oil sitting in each cylinder, should I expect a weird start-up?

@joshhuck
Not really. The oil should seep back down from between the piston ring and wall, or it will just get burned off after a minute or so. And after you change the oil and fill up with 91 octane, try doing an "Italian tune-up", which is just driving it really hard at highway speeds...this should help burn off any loose carbon on the intake valves and turbo (because of the heat and fast valve operation). Good luck.

@Razmig
Will do. Thats not too hard to do out here in rural country since the roads are perfect for and Italian tune-up where I live. We will see how this goes, and I'll report back!

@Razmig
Alright so the oil is changed and with the tank about 40% full of 87 octane, I filled up the other 60% with 92 octane. I drove it around my area for about 20 minutes doing about 60mph. It cruises fine, and it felt fine on light acceleration but it still feels like its misfiring whenever I start getting into boost. I didn't go hard enough to throw a code this time, but im sure it would have if I did. I soon as I started feeling the shuttering I let off, and it was around the same RPM range, so. basically at the very beginning of boost I guess. I know we are still talking about the PCV system, and possibly the wastegate control arm/BPV, and maybe even a bad injector. Just wondering what to do next. If it is doing this just due to running 87 octane and non-API SN+ oil, is it possible that it could take a bit more driving before I start seeing results?

@Razmig
Also, I updated the google drive folder with a couple of videos of the engine sound, and the exhaust note after the drive. The exhaust sounded a little odd to me.

@joshhuck
I looked at the videos and the exhaust sounds normal to me (no misfiring). But is that metallic sound in the background coming from the exhaust too? Or is that just rain droplets hitting the car lol? So it seems like the problem is still persisting with new oil and some high octane blended gas. Does it only start hesitating when you accelerate hard? Like if you accelerate lightly even at the 3k rpm range, will it start hesitating? And what does the car sound like when it hesitates? Misfiring/popping exhaust sounds? Knocking sounds? Rpm gauge fluctuates? You say that it's beginning to hesitate around 3k rpm...so I'm thinking it could be the VVT acting up since it kicks in around 3k rpm depending on throttle position and oil pressure. Another cause could be clogged fuel injectors...you have good fuel pressure, but maybe the injectors themselves are not spraying in a nice even pattern. Have you tried a fuel system cleaner in your gas tank yet like Techron? It's really good stuff actually. But then again, it may take a while for that LSPI affect to disappear even with new oil, since some of the old oil may be in the PCV system. If you were to rev your car to 3k rpm at idle, will the problem occur too?

@Razmig
I believe the metal "tinging" sound is the exhaust as well, kind of sounds like it's cooling down, I guess. Thats what a guy in the Mazdaspeed6 subreddit said. I dont hear any popping or other noises, its just the engine shuttering and notably shaking a bit while trying to climb through the rpms. And I can tell it's not gear dependent, since I didn't feel it until 5th gear this time, while last time it could have 3rd or 4th. It's definitely the rpm range/boost. I did run a bottle of Gumout last fall, I get that stuff because of the things I hear about PEA. I haven't done it since though. This issue does seem load dependent, so I think If I rev it high enough at idle it may misfire. I'll have to try that next time I get in it. At least it cruises fine though. What do you mean by the VVT kicking in around 3k rpm?

@joshhuck
The VVT kicks in around 3k rpm according to some mazdaspeed forums I've searched through, but it is dependent on throttle position (or load). So if you lightly hold the pedal until you reach 3k, it won't activate the VVT because it doesn't need to retard or advance the intake valves for light loads. But it's been worked on too so, I'm not so sure if that's the problem or not. Have you looked at the turbo stuff yet? Because it could be that the engine is not getting enough air past 3k. Just continue to drive it around and see if it slowly goes away with the fuel system cleaner, high octane gas, and the new oil. But it hasn't tripped a code yet? If it trips another code, send the freeze frame data again to look at.

@Razmig
I'm thinking that if the new fuel, proper oil and fuel treatment don't do anything to solve the problem, maybe I should give that shop a call and see what they say, if they say anything... With the things regarding the turbo, im not sure how to check for that. Or at least the only thing it sounds like we can do is see if the control arm moves when revving into boost, but as you said that's at idle and it may not activate at all. I didn't throw a new code yet. Ill go ahead and clear this one and try to get it to throw it again.

@joshhuck
Ok, keep us updated. It would give us a better idea if it throws another code other than a misfire. Do you think it would feel safe to continue giving gas when you start feeling the hesitation? I'm thinking maybe you should continue giving some gas during the hesitation so that the ECU can detect the problems and throw the necessary codes.

@Razmig
Makes me a little nervous, but I'll give it a try.

@joshhuck
I know, it would make me nervous too lol. When I was having my LSPI issue, I would get super nervous to even floor the car. And when it would start hesitating, I would immediately let off the gas thinking the engine might blow up. A few times I did continue to hold the gas pedal while it was hesitating, but nothing serious really happened...it would just continue to hesitate for another second and then run fine. It was a weird experience for me, especially when it threw no codes too. Maybe for your issue if you continue giving gas during the hesitation, it would maybe clear out after a second or two? Because you would want to know if this issue persists even after the 3k rpm range, that would give us a better idea on what may be causing the issue. Usually a bad fuel injector would make the car hesitate badly overall, not a certain speeds. If it does clear out when you continue through the hesitation, then it could mean the turbo is having trouble building up boost until it finally catches up with the air demand. When you accelerate in 1st or 2nd gear when you try getting the car moving, does it hesitate at all? Or does this hesitation only happen at the higher gears when you try to accelerate again?

@Razmig
When just starting out, getting the car up to speed limit whether its 25mph, 45, or 60, it feels pretty much normal as far I can tell. Cruising at speed feels normal as well. It seems to be whenever it just starts getting into boost, 3000 to 3200 rpm it seems, thats when the hesitation starts. The more throttle I give it while being in or above that rev range, the worse it feels, however yeah I have always been letting off a little bit after I notice it which is basically before 4000rpm. I don't think I have actually revved the car about 4k now that I think about it... So maybe this could be a turbo specific issue. But if the hesitation persists as the revs build, could it still possibly be an issue with the VVT?

@joshhuck
Oh wow, you haven't revved the engine up past the hesitation point? I mean, an issue with the VVT could be a possibility. But have you driven the car before the VVT repair? If so, did it ever do this? I'm not sure if you mentioned this already in the previous 50 comments lol, but do you know what the VVT repair was for? I personally don't think it's the VVT because it would throw a whole bunch of codes with all that hesitation, not just a single misfire code. The ECU monitors the VVT system very closely, so it would definitely give you codes if something was off. But I just remembered something, my friend with the mazdaspeed3 told me that the engine doesn't boost until in 3rd gear, which makes sense because when I get a ride in his car, you won't hear any boost kick in until 3rd gear. So maybe that's why 1st and 2nd gear are good, but then when the engine needs boost in 3rd gear it just starts hesitating 🤔. You haven't really checked the turbo yet, so maybe thats the next thing to take a look at. So far we eliminated the oil and gas off our list lol.

@Razmig
Yeah, I may be a little bit too gentle with a car like this, lol. So a couple months back I took it into the shop to have the time chain tensioner looked at. I figured it was time since I was getting that chain rattling sound on warm startups, there was also no way I was prepared to work on that part of the engine. Before that trip to the shop, I could rev into boost and seemed to work fine. No codes, no hesitation. As I mentioned, I never really listened for the turbo, but I could definitely feel it as the torque notably changed as it pulled up into that rev range. Ill be taking it out today to get some Gumout for the fuel system. When its warm Ill check the control arm while in idle to see if it activates during a rev. I'll also give revving passed the hesitation point a shot. Plus, letting the new oil and proper fuel cycle through the engine shouldn't hurt it anyway.

@joshhuck
Ok, keep us posted.

@Razmig
Okay, just came back from a drive. I shared a new google drive folder with you of the new freeze frame data with the new code. Still a misfire in cylinder 3, and the CEL flashed on several different occasions. During this whole ordeal, I managed to get up to 5k, over 70 mph. It did shake/hesitate begining in 4th gear and up. I can rev it above 3k while in the first three gears and it doesnt seem to want to hesitate. The hesitation doesn't get worse as it climbs through the revs in higher gears (4-6th), but it doesn't feel very boosted. A new symptom has come up now. The RPMs are very jumpy while in idle or neutral. When I was slowing down from 70mph, thats when I first noticed the jumpy rpms. As I idled while parked, it sounded like it was trying not to die. I included a video in the drive folder.

@joshhuck
Aha, so this actually may be a throttle body issue. Have you cleaned the throttle body yet? It sounds like the throttle body is not opening correctly due to some severe build up on the butterfly valve. Improper throttle positions can cause rough idle, and it may also be the cause for insufficient boost if it's not opening fully when you try accelerating. But when I get back home, I'll take a look at the freeze frame data as well. It could be that when you tried to rev it higher you broke up some of the build up in between the butterfly valve and motor.

@joshhuck
Ok, so looking through the data, I found that your intake manifold pressure was at 4 psi now, meaning your turbo is making some boost. This is most likely because your throttle position now was at 32% compared to 27% last time (your MAP sensor is on the intake manifold, so it doesn't read boost before the throttle). But what's weird is that your load was recorded as 101% when your throttle position was only 32%. Do you think you were giving 32% throttle when the light came on or was it more than that? Other than that, everything looks fine, including the STFT and LTFT. As for the fluctuating idle you recorded, the throttle body may be the problem OR maybe the fuel injectors. But as a next step, I would highly suggest taking a look at the throttle body and see if there's a whole bunch of carbon build-up on the butterfly valve. Please be sure to unplug the connection to the throttle body before messing with it, it can cut your fingers off if the valve decides to actuate lol. And continue to add fuel system cleaner for another few fill-ups, it may be cleaning out the injectors.

@Razmig
As for the the 101% reading, I'm not sure of my throttle position. All I can say is maybe... So it sounds like we can rule out the turbo being the issue. Last time I checked the throttle body, which happened to be before the VVT job, i thought it looked okay. There was a little bit of oil around the bottom edge of the butterfly valve, but I didn't think it was that much. Im not sure if thats the case now, so I'll pull the intercooler off and take a look. Ill get some pictures of it while im at it.

@Razmig
I just updated the google drive folder with a couple images of the throttle body. It doesn't look much different from what it looked like a couple months ago. Is that pretty dirty you think? Also, once I disconnect the connector form the throttle body, touching the butterfly valve won't harm the system at all?

@joshhuck
Yeah, it needs a good cleaning. The front side doesn't look that bad, but usually the back side will have a ton of build-up. And from the picture, it looks like there is a good amount of build-up around the valve's pivot points. It wouldn't hurt trying to clean it thoroughly. I would give it a try...it's a pretty simple job too. It may not fix the cylinder 3 misfire, but it will help with the overall hesitation and allow more air to get into the intake manifold. 

@Razmig
I'll take the unit out today and take a look at it. I already have the spray cleaner for it, and a new gasket on standby. Is there anything I should be cautious of when clean the throttle body? Like is it okay to okay the butterfly valve with my fingers, mess with it, etc.?

@joshhuck
Yes, you can push it open with you fingers ONLY when it is disconnected from the wiring harness. Of course, don't be too forceful with it if you notice that it doesn't go past a certain angle, but you do want to make sure that it can open all the way as well. Hopefully cleaning it will improve the overall performance, and it will also give you an idea of how much carbon build-up is inside the intake manifold too.

@Razmig
So im getting confused now. I noticed the coolant lines connected to the unit, and before I disconnect them I wanted to do some looking into what I should expect once I do that. Now I am starting to see people saying that I shouldn't touch the butterfly valve at all, period. While also seeing people say that I need to bleed the cooling system any time a hose is removed, which I don't know how to do. Sorry, I'm still novice, so I'm not sure whats safe to do. Can I take the unit off and remove the coolant hoses and put them back on after cleaning without bleeding? What would happen if I do that?

@joshhuck
I understand that you may feel uncomfortable messing with the throttle body. I've done numerous throttle body cleanings before, and I've personally never had problems. Now in the case of your Mazda, it has coolant lines to disconnect. I've done a throttle body cleaning on my friend's Mazdaspeed, and I'd say a quart of coolant came out of the throttle body when we disconnected it (do this when the engine is cold...you can catch the coolant and reuse it). As for the bleeding process for your car, you need to remove the cap off the coolant reservoir and let the engine warm up to operating temperatures. Have your heater on the max setting. Once the thermostat opens up, the coolant will flow throughout the whole system and will release any trapped air out from the coolant reservoir. It's totally up to you if you feel comfortable enough to completely disconnect the throttle body. Worst case scenario, you could try cleaning it without disconnecting the coolant lines (but still disconnect the electrical connection), but you wouldn't be able to clean it as good. I don't want to force you to try something you don't feel comfortable doing.

@Razmig
Oh no it's fine. I needed a little more information on what procedures to follow when I do this. Plus it seems like everyone has their own way of explaining it, but at the same time a lot a people leave information out (probably because they have an entirely separate tutorial on that bit of information...) I'm gonna take it off, because I'm really curious to see how bad the inside looks. Thanks for walking me through it. I'll report back after a drive.

@joshhuck
Alright! Yeah, it would give you a really good idea what the intake manifold looks like inside...I'm sure it's gonna be caked with carbon build-up. That's the downside of GDI though lol.

@Razmig
Yeah. I can't wait to get that OCC on it.
So, it's getting pretty frustrating. I currently trying to get the hose clamp off of the throttle body end of the hose that the intercooler piping attaches too. Somehow, whether a mechanic or previous owner/s, they tightened this clamp from underneath the throttle body. I even tried reaching it from underneath the car and still cant reach it. Im pretty sure the tightening screw of this clamp is right in the way of the bottom back bolt holding the throttle body to the intake manifold, so unless there's an at least 12" extension for my socket wrench out there, I may have to take this into the shop...

@joshhuck
You can take off the throttle body with the hose coupling attached, can't you? So when you take off the throttle body, you should be able to get the clamp loose.

@Razmig
Well it feels like the tightening screw of the hose clamp is right on top of the bottom right bolt of the unit (I can't see it,I can only feel for it). Tomorrow I'll be making a trip to Harbor Freight to get a 12" socket extension, I need to update my socket set anyway. I have to unscrew the hose clamp from underneath, otherwise I don't think im gonna be able to get to it. Or I'll just keep trying to find a creative way of getting to this one clamp or bolt, lol.

@joshhuck
Haha, I'm just wondering how someone managed to get a clamp on like that anyways lol. Why didn't they just make it easy on themselves and put the clamp screw on top?! lol

@Razmig
My thoughts exactly, lol. It's like the same issue I had with trying to get a clamp off the PCV hose underneath the intake manifold when I was trying to do the OCC install.

@Razmig
So I completed the throttle body cleaning, and like you said, the back side of the butterfly valve was pretty filthy. I shared another folder of pictures showing the extent of the filth and some other interesting things that I wanted to take note of. I managed to get that clamp that was giving me trouble off by means of using a 10" and a 14" socket extension combined, lol. But the clamp its self on the tightening screw was super dirty with the same gunk that you find in everything else (what I'm assuming to all be from the PCV system...) It almost seemed like maybe there was a leak between the coupler and the throttle body. I made sure to tighten things securely when reassembling, and made it all to where I can actually reach what I need to get to. I have yet to start it up, it took a lot of cleaning and it was a really cold day. Hopefully this cleaning does something, although after seeing how much crap was around the back edge of the valve I'm sure the car will appreciate it, probably get some better fuel economy too. How gloopy it looked makes me think that when I road on the car hard that one day to throw the code again, maybe it loosened up some gunk from the PCV system and ended up sticking to the butterfly valve once I let off. This thing probably needs carbon cleaning in general. Anyways I just wanted to let you know how the cleaning went and to see your thoughts on the pictures.

@joshhuck
Oh yeah, that was a whole lot of gunk huh? From the pictures, you can definitely tell that there is a lot of oil vapors being released from the PCV system. There's just a whole bunch of dried up carbon all over the pipes and even on the throttle body motor actuator. I think you will definitely benefit from an oil catch-can, and I would highly recommend it. Whether the PCV system has an issue or not, I would have the PCV valve replaced whenever you can. If you want to be super complete, you should have an intake valve cleaning done, replace the PCV, and install an oil catch-can all at once to prevent carbon build-up from occurring ever again. I really hope cleaning the throttle body does make a difference...it was pretty gunked up (and it looks like a "sticky" sort of gunk too). Let us know how it drives!

@Razmig
Yeah it was a pretty dirty job. And it was definitely sticky, almost tacky kind of. I even wiped some out of the entrance of the intake manifold, it was caked on the bottom pretty good. So I guess this is what a GDI engine at 162,000 miles with no OCC looks like, lol. Not a ton of coolant came out, probably a pint roughly. But I managed to catch most of it. So when I go to start it up, as far as bleeding the cooling system, I need to have the coolant reservoir open and what else?

@joshhuck
Oh, so not that much coolant came out. So for bleeding the cooling system, you start your car up with the heat on the max setting WITH the reservoir cap off. This will have the car warm up quickly to open up the thermostat. When the thermostat opens up at operating temperature (when gauge is about halfway), the coolant will start to flow throughout the whole system, and this will help carry out any air bubbles out to the coolant reservoir. You then want to rev the engine up to like 2k rpm a few times so the water pump stimulates more coolant flow. After about 5 minutes, check the coolant level. If the level went down, then it means you got a good amount of air out of the system. If the level didn't change, then there wasn't too much air to begin with. You can now close the coolant reservoir. You want to make sure you feel a good amount of heat coming out of the vents, but if there isn't then it means there is a whole bunch of air stuck in the heater core that still needs to be released. I wouldn't worry too much about any trapped air since hardly any coolant came out. This process is generally for a full coolant flush.

@Razmig
Ah, gotcha. If the windy weather calms down at all today, I'll go ahead and do that then. Thanks, and I'll let you know how it all goes.

@Razmig
Hey man. So I updated the "Misfire, plus jumping/dying idle" folder in google drive with a couple of videos of how the car was running when I started it up to bleed the air out of the cooling system. So, it sounds worse, and I have no idea what this could mean. The first time I tried starting it up it died a few seconds after turning over. The second time it managed to keep running, however it sounded terrible. Im sure you'll to hear what I am talking about in the videos, its not smooth at all. When I would try to rev it, it would really want to die after revving. No exhaust smoke, but my buddy had to use his scan tool for his own vehicle, so I wasn't able to pull any data. No codes were thrown either. Im going to check the oil level, but I doubt it moved at all.

@joshhuck
Yeah, it's doing that because the ECU needs to recalibrate the throttle position after disconnecting it and reconnecting it. Let the engine run for like 10 minutes and see if it gets smooth. If not, then try disconnecting your battery, wait 10 minutes, and then reconnect it. Don't worry, it's not a more serious problem, it's just re-adaptation. But in the video, the engine sounds fine. I didn't hear it idle bad.

@Razmig
It just seems a lot like the idling it was doing before I cleaned the TB. Im about to disconnect the battery.

@Razmig
Its not getting any better. I included a couple more things in the drive folder. The picture I outlined shows a splatter that looks like it came from the exhaust while it was trying to start up. The start up was rough, it really didnt want to. Hopefully the video of the dash youll be able to hear the idle, because it doesn't sound good at all to me.

@Razmig
At this point, I'm worried that I have a leaking fuel injector. Which means driving it could lead to catastrophic failure from what I've read...

@joshhuck
I'm not sure what happened to cause this idle issue. This was somewhat occurring before the TB cleaning, and cleaning it didn't fix it. The pipes and TB are tightened down correctly to prevent vacuum leaks I'm assuming. Can you maybe send live data while the car is idling? Take a look at the LTFT and STFT. If it's negative, there's too much fuel or not enough air. If it's positive, there's too much air or not enough fuel.

@razmig
I'll try to borrow my buddy's scan tool again (he needed it for his neighbors car). Although I am beginning to wonder if the issue has anything to do with the VVT job it had done... Could this mean that the timing in the engine is off? I feel like the shop screwed something up, because this car had no issues outside of a minor leak and timing chain rattle before I brought it to them. But at the same time, the only code its throw is the cyl 3 misfire... I should have got a video of the start up, because it was bad also. It would turn over slow, but it didnt even rev up after the engine started up. Its like while the starter was going the engine would just stumble up to 650 or 700rpms and stay there, then going back to its surging idle. I think I am going to call the shop and tell them whats going on, because I have feeling it starts with them. They will need to come pick it up too, i really don't think it's safe to drive...

@joshhuck
You know what, I think it's best to take it back to the shop who dealt with the VVT to take a look at the car. You already tried a few "at-home" fixes, but it seems like this is an issue that requires in-depth diagnosing and perhaps more involved work (carbon cleaning, timing re-adjustments, wiring, etc.). Have them check the injectors because it may just be super dirty injectors (if it's not a VVT issue).

@razmig
I have the freeze-frame data to show them too. Luckily I bought this car with a warranty, which helped with the VVT job, so if it turns out to be something other than VVT at least the warranty should cover some if not all the cost. It's just so bizarre that the idle is so erratic and inconsistent as it is now. I really made sure I put everything back the way it was too, while tighten clamps and such.
Well thanks again, a thousand times, for helping me out and sticking with me through this process. Even though we couldn't pinpoint the problem, I've learned a lot at the very least.

@joshhuck
No problem! It was really fun looking into it. It wasn't a complete waste of time...you learned a lot along the way and I hope you continue to learn more with your car. I don't think it's a really big issue. It's probably something to do with the fuel injectors, carbon build-up, or just an issue with the wiring harness (short when it vibrates or something), but definitely not a mechanical problem with the engine itself (the compression test was good). I wish you the very best of luck, and let us know what happens and what the issue was! I'd definitely want to find out what the pesky cause was lol.

@razmig
Hopefully. Well, I should have some news probably by next week. The snow is keeping the tow service from coming out here to get it. I'll let you know what the findings are!

@Razmig
So I got the diagnostic report today. Get this, supposedly it was the EGR valve stuck open. They told me that it started right up when they first started to look at it. But then after a drive it began to idle poorly just like how it was before, which then led to the prolonged start-ups just like before. They couldn't get it to throw a misfire code, which is interesting, but they did get to throw a P0401, which deals with the EGR failing to function properly. They took it off, cleaned it out, and now it supposedly runs fine. They recommended doing a full carbon cleaning obviously, but I'm going to hold off on that for now and see how it does.

@joshhuck
Wow, the EGR of all things lol. Interesting how they were able to trip the code for it too, but I guess that was the problem. If it's running better now then that's great.

@Razmig
Just got the car back from the shop. Started right up, revved fine, and drove smooth. I do notice that the warm-up on a cold start sounds better, I believe thats due to the throttle body being clean. Now if the same symptoms return, I guess I can just go straight for the EGR and see if the valve stuck again then give it a clean. But I learned from this particular shop that de-carbon cleaning isn't as pricey as I initially thought, so I scheduled that for a later date. After that, I'll be looking for a time to install the OCC, and carbon build-up shouldn't be a major issue for the rest of it's days. At least If we hear of someone experiencing the same symptoms in an engine similar if not the same as the L3-VDT engine, we could recommend looking into the EGR valve. Thanks for baring with me and helping me out!

@joshhuck
No problem! I'm glad your car is running fine now!

Ill have to try that later this week, when the rain clears out.