I've got a version 8 EJ207 shortblock + WRX heads and a 16G, as a lot of you know. As part of that swap, I had purchased heat range 7 *and* 8 NGK iridium plugs. The heat range 8 plugs were purchased as a backup based on two things:
1. NASIOC user ebeck had "lots of knock" with heat range 7 and had to use 8 on his EJ207 setup (much larger turbo, however).
2. The compression ratio of the v8 EJ207 block with WRX heads is somewhat unknown. Quirt at Crawford Performance has told ebeck several times that it is roughly 9:1. I have not convinced myself firsthand of this yet.
So here we go:
I'm running a Hydra Nemesis.
Over the last 5 months (it feels like a tedious 2 years), I have been datalogging the hell out of my car. Here's the cliff notes version in chronological order:
- Swap done, using heat range 7 NGK iridiums
- Original Element Tuning map was reporting lots of knock. This was a rough map done for a PE1820 turbo.
- Sunoco GT100, 16psi peak, and conservative timing did not make this reported knock go away.
- Insert skepticism and some fear
- After discussion with Phil @ Element Tuning, I used the 100 octane tankfull to generate a new Knock Threshold map to suit my car.
- A month or so after that, I revisited this. I was uneasy about the knock voltage levels (regardless of octane) that seemed distant from the other main cluster at that RPM and load.
- I contacted Phil again. This time I provided him with a plot of my unloaded free-revving knock voltage from 0-6500RPM. I also provided him with portions of logs containing WOT pulls. I also provided him with 5 knock voltage plots showing 2+3+4 pulls from 2800-6800RPM at WOT. Lots of data, all done with 100 octane, reasonable boost, and reasonable timing.
- Phil commented that, aside from his car, he had never seen knock voltages that high in his Hydra tuning. Perhaps the heads are noisy. Regardless, it was his assessment that it was extremely unlikely that it was knock given the data in the logs sent.
- For reference, my knock threshold map was above 2.05 volts from 2400 to redline. I would consistently see 2.1 and 2.2 voltages in spots above 5600RPM.
- Before doing a track day at Sebring, I decided to do a compression test now that the swap had 3000 miles on it (many many many WOT pulls while tuning, etc). Compression was 150psi in all cylinders. Heat range 7 iridiums looked fine.
- Changed oil, did the Sebring track day, changed oil, yadda yadda.
- Ever-paranoid and cautious, I was still not convinced everything was okay, I decided to try heat range 7 NGK copper plugs. No significant change in noise level.
- Yesterday morning I pulled the copper plugs after 2 weeks of use. They looked normal/boring.
- I installed the heat range 8 NGK iridiums finally. DRASTIC change in knock voltage levels. I have seen a single 1.9 out of 20 pulls, and that was at 6500RPM. From 0-5700RPM the voltages are well below 1.7, mostly below 1.2 even.
- I got another tank of Sunoco GT100 last night and started rebuilding my entire knock threshold curve (I am currently at $280 in GT100 fuel just for the sake of tuning a basic knock threshold curve).
- Keep in mind that I have never heard a single audible detonation event.
a) What the hell is going on here?
b) I don't really care about the answer to this one, but was ebeck's "knock" as reported by his UTEC not knock at all? I clearly ran my car hard for 3000 miles on 93 octane and 18psi with high knock voltages from my heat range 7 plugs. The result? Perfect compression still. Car runs like a champ. Super crisp.