Reverant

iTrader: (10)

The log above is from an Enhanced MS2. It has a different name for the knock parameter, everything else is the same (including the stock NB2 sensor used).

If you look at the value at the moment highlighted, its 236 (out of 1023), which roughly traslates to 23%, but is fairly constant with no big spikes. Not so a bit earlier and later - the huge spikes are actual, audible knock - if you look at the TPS reading, I immediately took my foot off the throttle as soon as I heard it.

Gibbs

In my opinion the spark advance table looks a little extreme for only 91 octane.

Chowcow

So with that being said, would you say that I'd be safe to advance a bit more timing?

The only reason I'm trying to push timing here is because I'm wondering how the other owners of FM2 setups are getting around ~210wtq. I'm assuming its the timing that makes the biggest difference (assuming similar parts).

codrus

I don't have any specific link for the VVT head -- mine is a 99, so I haven't needed it. I don't think the VVT will help with knock, you need more octane or less compression ratio.

--Ian

Reverant

iTrader: (10)

My experience tells me that this is just noise, not knock.

What's your elevation?

Chowcow

1048 meters.

As for the higher octane-- I rather tune at 91 octane so I can safely run the car at that level and maybe use 94 when I decide I'm going for a hard drive (94 is hard to come by). As for reducing compression, I can't unless I spring for new rods/pistons.

Still.. Some people are making ~200wtq with this setup on 91.

18psi

iTrader: (76)

Skimmed thread, so pardon if this has already been established, but:

Have you verified it's actual knock?

Because like Rev is saying, I think the sensor is just too sensitive, you shouldn't be anywhere near det territory with that power curve you got, unless something is majorly wrong.

I've tuned bp05's on crappy cali 91 with similar setups to yours and made 230-240 without knock/det, you should be able to do even more.

*Edit: I just realized a bp05 is a bad example since it's lower comp, but still, 210whp on your setup should be EZ-PZ

Chowcow

Verified with det headphones? No. The best tuner in town, or the one most highly recommended doesn't seem to use det phones.. I tried looking around his place for them but I wasn't able to spot a pair. Are there other ways to verify knock without det cans?

Also, would the proper way to tune be to increase the timing until knock occurs (such as Rev's map) then retard it from there? It seems a bit scary as this is being done on WOT.

My only knowledge of tuning comes from this board, and although I feel like I've read enough to understand what's going on, I let the recommended tuner do his thing. He's never worked on a Miata before, so maybe the knock sensor thing is unique to these engines?

*18PSI* So these 91 cars had way more advanced timing than the data I posted?

18psi

iTrader: (76)

"best tuner in town" means didly squat if he hasn't experience tuning a miata on MS. Some of the newer cars have seriously insane knock-detection systems and OEM ECUs that make even MS3 look archaic.

The best way is to use det cans on a load bearing dyno and try to get as close to MBT as possible then back off a tad. With 10:1 and 91oct you likely won't get to MBT but you certainly should be able to make at least 210-220 on pump on your setup. At least based on the parts and setup you listed.

not a whole lot more, but more boost. The last one I did ran I think 13-14midrange 16topend and about 14psi on a gt2560r

18psi

iTrader: (76)

....oh and basically I take even the MS3+OEM knock sensor data with a grain of salt. it's much MUCH better than what we had before (which is basically nothing), but still a ways away from being trustworthy 100%.

IMHO

Chowcow

Alright. I'll try and get det cans and figure this out.

Thanks!

patsmx5

iTrader: (16)

You ever read up about the J&S Safeguard? They're $$$ but from what I've read, they are fantastic. I found a mustang guy who had his Fuel pressure regulator pop off in boost and the J&S kicked in and pulled like 15 degrees.

I'm tempted to buy one as I can't get the MS3 PRO knock detection to do anything.

aidandj

iTrader: (5)

<p>The tuner probably just watches the exhaust for signs of det.</p>

18psi

iTrader: (76)

A buddy of mine has a TurboXs knock lite of some sort, was basically useless on a BP lol.

I'm sure the fancy ones are more effective, but I think the main problem is our engines are noisy, and rattley, so they throw off your typical sensor that can't differentiate between different frequencies sounds.

patsmx5

iTrader: (16)

From what I've read, the J&S is damn good, better than anything else I can use without putting a Mitsubishi, Subie, etc ecu in the car. Way ahead of MegaSquirt from what I can gather. Apparently the J&S can differentiate between the noise and signal where most others can't.

Besides it's only 600 dollars..........

Chowcow

Another question before I hit the dyno again..

Would it make sense to dyno the car on 94 (WOT dyno), tune it towards that gas but then DD it at 91. The idea here is that when I DD, I'm rarely breaking 5-7PSI. But this gives me the ability to throw 94 into it for the weekends.

turbofan

iTrader: (1)

Tune for the poorest-quality fuel you'll run.

That's all I can contribute to this, ha.

codrus

Alternately, tune for both and install an easy way to flip between them. For example, I used to tune the mechanical wastegate for 91 and the boost-controlled level for 95 (mixing 91 and 100 in a 50/50 ratio). Most of the time I ran the 95, but if I needed to drive it long distances without having 100 available, I could just unplug the boost controller and run it mechanical. The lost timing in the 7 psi row cost me a bit of spool, but it was worth it.

That said, don't rely on your right foot to keep it out of the danger rows. Put in a mechanical switch, otherwise the temptation to roll into the boost when you've got 91 in the tank and teach that nasty Corvette a lesson may become too much...

--Ian

Chowcow

I might just do this..

Last question.. Before I take the car back to the dyno and tell the tuner that the above logs may just be knock caused by a noisy sensor, could anyone else give me an opinion on whether that log shows real knock or a noisy sensor? I'm tempted to tell him to ignore it and bump the timing up until I get a knock spike such as Rev has demonstrated.

--- Does that put my engine at risk of blowing up if I get that large of a spike?

18psi

iTrader: (76)

no one can answer that question but you. you have to listen to knock and deem if the sensor is really picking up knock.

also an engine can knock for a long time and not blow up and another one have just 1 really strong knock event and splooge.

it is not clear cut like you think it is.

based on your timing map and setup AND assuming nothing is actually wrong, then yes I would GUESS that is probably just noise. As would most other people. look at how sharp a real knock event looks on revs log.

again though, this is an estimate at best.