dc2696

iTrader: (2)

I just finished pulling my engine then removing the manifold and my exh runners look fucked up.

Def signs of det (which I expected with the way it was running last), but there is also a weird flaky ash coating on my runners.

I'll be pulling the head off for further inspection, I'm guessing swiss cheese looking piston heads.

I was running alot of timing and couldn't keep the IAT's down when in boost above 15psi on a 2871 .64 hotside (crower rods, wiseco pistons, 8.5:1 comp, blablabla)

I picked up a hy35 holset for this summer as the 2871's imo don't fair well in the 20+psi range.

rweatherford

White ash is usually oil consumption in my GM owning leak prone intake manifold experience. ;^)

hustler

iTrader: (6)

Show us your spark table, please.

Faeflora

iTrader: (2)

So I googled evo knock detection for a little while and holy shizmolio the stock ecu's knock detection is ridiculously complicated/sophisticated and touted as being extremely effective. There are multiple fuel maps and multiple trim maps and the OEM ecu switches between them. It makes my Hydra look retarded.

These threads are a good intro for mt.netters:

Caution - "Knock Sensor Filters" - evolutionm.net

Knock sensors and knock detection - evolutionm.net

The above thread "knock sensors knock detection" links out to some hard science detonation papers. Search the nasa site like it recommends. Read them and you will understand why I say FML I should have gone to school for physics.

JasonC SBB

This diagram explains why the recorded knock I posted had a descending note.

evolutionm.net - View Single Post - Knock sensors and knock detection

dc2696

iTrader: (2)

Edit: Here is the map I was running before I blew up the last tranny. No knock at 156kpa then @ 183kpa timing has to take a huge hit to suppress the knock. AFR's are 12.5 till 7psi, 11.5 after that, tip in is 11.0-10.7

Not white ash, black ash, like soot but stuck to the exhaust runners, it flakes off when poked. Prolly excess unburnt carbon as I am running the shitty 1.6 coilpacks and @20psi spark is prolly blowing out.

-Dean

Travisivart190

iTrader: (1)

The soot you're mentioning sounds like what I saw in my exhaust runners as well when I swapped out the engine. At the time I did have the 1.6 Coilpacks, but that engine was also horribly fucked. I'd be interested to see what exactly causes that. It almost comes off in little chips.

dc2696

iTrader: (2)

Yup, little flakes or chips, I was thinking oil was/is getting cooked and causing the flakes.

magnamx-5

iTrader: (7)

that would be my gues as well carbon wouldnt scale that thickly and come off so flackiliy did you have any oil consumption? i know i had that **** all over my exhaust ports when i blew my motor and was burning as much oil as gas.

dc2696

iTrader: (2)

Ya she was consuming alittle oil, unhooked the pvc from the intake and oil consumption went way down so I'll be needing a ring job soon I'm thinking.

I'll be pulling the head apart before I put the engine back in, and maybe I'll throw on the 99 head I have laying around.

Travisivart190

iTrader: (1)

Yeah, that sounds about right. That engine at oil for breakfast, lunch and dinner until it finally kicked the bucket.

SKMetalworks

iTrader: (7)

What i think the carbon buildup is a multiple of things. First, the carbon is constantly running through the ports. Second, oil could be pulled into the chamber; mixing with the carbon from the fuel and start a small buildup. Third is the rough ports and the casting left on the walls. Both my Miatas i have pulled off the exhaust manifold and removed/ground/sanded/polished my exhaust ports. I'm not sure if this well help but i think the carbon mixture will have a harder time accumulating on a smooth surface vs a rough one. And It only takes an hour to accomplish.

sn95

iTrader: (2)

[QUOTE=faeflora;509142]So I googled evo knock detection for a little while and holy shizmolio the stock ecu's knock detection is ridiculously complicated/sophisticated and touted as being extremely effective. There are multiple fuel maps and multiple trim maps and the OEM ecu switches between them. It makes my Hydra look retarded.

The EVO knock detection is good, but don't be fooled into thinking that the 2 fuel maps (high octane, low octane) and 6 spark maps (Maps 1-3 for both high & low octane) are what makes the knock sensing work. It seems that no one but Mitsu knows when the car is pulling from timing map 1,2, or 3 or interpolating between them. As a result, all of the Mitsu tuners end up normalizing the three timing maps. A lot of tuners are also running the "Tephra mod" for the Evo VIII and Evo IX ecus which provides an alternate mapset (with a normalized timing map 123) for switching between a pump gas/race gas tune or pump gas/E85 tune with a toggle switch (e.g., don't need your laptop in the car to switch tunes).

Ecuflash is open source software and the most widely used tool for tuning Evos; you can download it from openecu.org and use it free of charge. Evolutionm.net has some extremely good tuning information and tutorials in the ECUflash section (EcuFlash - evolutionm.net) that are applicable to tuning any turbocharged high performance car.

psiturbo

iTrader: (1)

Many say no knock detected or heard, do you mean to the naked ear?

I sometimes think part of the problem could be the use of an incorrect intake manifold or bad intake manifold or cylinder head port job.

Like for example on the Mazda Protege and Escort GT N/A intake manifolds are very different from the turbocharged Mazda GTX and GTR ones. Both use identical engines, the B6 and the BP, but is clear to notice that Mazda did address using a different intake manifold for the turbo engine.


Anyway, I am a little clueless on new models, but... Does the Intake manifold changes between the new OEM turbo Miatas and the N/A counterparts?

dc2696

iTrader: (2)

Little update:

Finally got a 12pt 13mm socket, anyway I was expecting to see some detonation on the pistons but to my surprise they look great.

Slightly black/brown/goldish, which wipes right off and reveals a nice shinny new piston.

Combustion chamber on cyl 2 shows some det but the others look fine, plugs are nice and brown.



-Dean

Erotomania

Just out of curiosity, how many of you who blew built motors had your head port and polished? Primarily the combustion chambers? I've been reading a lot from various sources. The verdict I've seen that may be pretty important is the valve shrouding in the chamber. I've seen some DIY jobs where the shroud was not completely smoothed in the outside cylinder wall edge of the valves, obviously the hardest to reach spot. If you start removing hot spots but leave one, its now one very hot spot. Food for thought.
portview.jpg?t=1273759474
You can see on mine the edge that still has to be worked down on the top outside corner of the valve relief. From what I've been told, leaving it this way is almost just as bad as running a bone stock head.


Also, everyone is listing their piston Comps. but I'm curious to the respective head/block deck hight and gasket thickness in combination with any head work. I know if you really get at the chamber you are reducing compression more than you might thing, and obviously any deck height changes have their own effects. I only mention it based from others I've known grenading motors tuning under the assumption that their low compression pistons are still low compression after machining work on both head and block.


I'm still learning the whole tuning bit, so this is all great discussion.

ZX-Tex

iTrader: (15)

In my case the head chambers are completely stock except for the oversize valves, and there is no significant decking on the head or the block as far as I know.

Lennarth

Hi Guys! I made knock studies on a turbocharged 1.6L engine running RON98. I found out that up to 7-8PSI no need to change ignition point. Above this boost pressure knocking sensitivity was very much engine speed dependent. Worst condition occured at the speeds were inlet runners gave maximum assist and this is pretty common. So I prefer to use very short runners to avoid this problem and compensate the absent gasdynamics with increased boost pressure to maintain volumetric efficiency.

Another more complicated route to follow is ethanol. Here in Sweden it is fairly cheap and easy to get.
I am using a separate set of injectors and a Haltech system to control it. Intent was to do nothing with Mazda ECU and just add ethanol. The engine will see a mix between gasoline and ethanol were the ethanol part will increase with boost level. I have been running stock ignition set up, stock pistons ar CR 9.5:1 and used 18PSI with no knocking. I also took the engine apart to check if there were no indications of knock but it looked very nice. Midlands were also nice. However, stock rods are not recommended at these levels....The stock conrods are very nice in their own way; they just collapse gently adjusting CR to a survivalable level. Ethanol used as described seems to influences ignition delay so peak cylinder pressure is actually lower than expected.

I have similar results with turbocharged motorbike engines from my engine dyno. Ethanol consumption is around 10-15% in average usage on road.


ciao/Lennarth

ZX-Tex

iTrader: (15)

Very cool!

JKav

iTrader: (1)

Which turbo are you using & did you log exhaust manifold pressure as well?

Interesting re: trading VE for boost.