Kevin,
We're not mixing RON, MON and AKI up in this discussion are we? Surely 98 RON isn't thought as "race fuel"?

i agree it is low. I run 18 WOT timing with similar setup as OP and yes that is low too. But I found engine pinging under load so backed off until I could no longer could hear it pinging and it sounded more happy I agree that I am leaving power on the table. It appears both OP and I understand something is not as could be expected. We just don't understand what the reason is why we get detonation with higher timing advance. For me I will delay cam timing slightly and change to a CDI based ignition. As part of troubleshooting. Maybe cams were timed wrong or ignition issue

You don't list your geographical area, but 100LL is available at almost any airport around. It's labeled "Low Lead," but their idea of low isn't really low. And adding even a tiny bit of tetraethyl lead to an unleaded fuel really spikes up the octane. Try adding a few gallons of 100LL to your 98 RON gas and see how the engine likes it. If it was truly an octane problem, the engine should respond positively. If there's no change, I'd look elsewhere, like the location of the knock sensor(s).
This of course, assumes that your car is pre-smog or the catalytic converter is removed, as the lead could damage the substrate.

I am going to agree with Dannobee. The easiest thing to do was your 1st thought.

1.Do I try some avgas or racing fuel as an experiment to see what extra timing I can run and what extra power is available? If the extra power is negligible, I’ll leave it alone.

This will answer a lot. If running more timing with the higher octane doesn't improve power then there is nothing to do.
If power is better then have your tuner listen for knock, If they still hear what they think is knock at the same timing point then to Moregas point it is probably just engine noise. If they don't hear knock at same point and you have more power you have some decisions to make on Cost/ benefit of any changes.
With this information it will be much easier to plan a path forward.

john

I am watching this thread with great interest as I am building a 3.4 with much the same specification (twin plug but ITB, Motec) and am a little concerned.

Now my cam choice is DC35 as recommended by the esteemed Mr. Dougherty, but from everything I have read including Wayne's book on engine rebuilding, a twin plugged 3.4 ought to be able to use advance ~22 to 24 degrees without problems at the 10.3 compression level.

Nothing I have read to date, excepting this thread, has suggested there was a knock limit inherent in a 3.4 build, subject to all the proviso's raised here.

My thought is that the OP has another problem with his engine...perhaps cam choice, timing or some other issue, but I await with bated breath the judgement of those with superior knowledge

Dennis

Sorry- yes let's not get confused on the Octane rating method! I assumed the orig. poster meant 98 unleaded raceful I see at the track all the time, this is not RON but MON. Roughly equivalent of 93 I read MON(we don't get that in So Cal) but back East they have it and elsewhere in the USA.

It would be worth asking Muzz if someone actually CC'd everything out and verified he really has an accurate compression ratio known.

Also a quick chk of The 82-83 Spec book shows ignition advance for SC's at 19-25 for USA Japan, and not more than 25 degrees elsewhere. So removing 5-6 for twin plugs is still 19-20 degree off w twin plugs so very shy set at 15.

Does this ECU have good knock sensing capabilities or are you solely depending on an audible knock external device? The former has incorporated a crank angle variable in the sensing and can therefore be counted on as very accurate as the window in which knock wll happen is pretty small, compared to 720 degrees of crank rotation. The better knock sensing stand alone devices use a crank angle reference. This way you really can discount the noise if happening where it is not logical or likely. Aircooled engines are very noisy.

KR
GAS Motorsport

This is a great tool to understand dynamic compression ratios. Plug in some bigger cams #’s and you’ll be surprised how much compression drops. Could be a win for your goals for more top end power and lowering compression.

Wallace Racing: Dynamic Compression Ratio Calculator

Guys,
Some great discussion for follow up.

Yes, I can verify the chambers and piston dome were cc'ed and the 10.3:1 compression ratio calculated from this info.

The car has a Motec M130 with two head mounted sensors on 964 bridges so identifies knock against crank angle so those individual cylinder(s) are identified and individually trimmed as required.

So I'm off to the airport to buy some 100LL, remap the full load ignition setting to around 24 or so degrees and do a ramp run. After the run I'll download data and see if the knock sensors detected anything untoward, if I don't hear it myself of course! Might do a couple of ramp runs at various advance settings to understand how sensitive it is.

Europe uses RON. US uses (RON+MON)/2. What's the approximate conversion? So is 98RON comparable to 93 US?

Best of luck with the continued troubleshooting. I have to believe the race fuel will solve the knock. It would be great to know what the octane requirements are so you can blend in the future if necessary. If the AV gas has lead I would only consider using that as an experiment. It’s very corrosive on exhaust parts and quite bad for your health.

Using my earlier posted calculator you can try and see if you corrected (Dynamic) compression fall into the 7.5:1 target John looks for on our 91 oct pump gas (R+M/2). If so then you can feel good about the mechanical combination. If not then play with other cams like the DC21 or DC24 to see how much those drop compression. Both of these are mild well mannered cams I would not hesitate to run on the street.


http://forums.pelicanparts.com/uploa...1581608225.png

I can confirm that the dc24 is indeed very streetable. Used this grind in a recent 3L twin plugged build with ITB'S, SSI et.

No issues at all. In fact, I was a little disappointed - should probably have gone with DC40's.

Next time!

Lead will also kill any oxygen sensor really quickly...

Just a few ideas to float here.

Knock is either pre ignition event or post. Pre is typically a temp issue, plugs, hot spots etc. Post is cylinder pressure, slow burn to end gas particles, fuel octane number, plug position etc.

Try the most simple tasks first. Look at your spark plugs. They will tell you if the engine is actually knocking, they will tell you if the ignition is to advanced etc.

Regardless of what way you are determining the engine is knocking, the plugs are your best way of telling. There is no dispute in their appearance. The ground strap will tell you the ignition position. The outer shell of the electrode will show deposits id the engine is knocking. Under normal knocking events, as the pressure spikes, so does the cylinder temperature, so the plug will show signs of over heating.

So start with the easy parts first. Fit the coldest most retracted plug possible and try again. Something that should be a give away was your first post. You state the engine is ignition limited above the mid range. Can I suspect that you are past peak torque? If so, I would suggest you do not have a pressure related knocking situation. It must be either a temperature issue or some other mechanical noise picked up by the sensor and not knock. Again, look at the plugs.

A couple of things to consider here. When finding out the max possible advance to run, go backwards first. Remove some timing and see if the torque drops. If it does, then the engine will take more, as long as it doesn't det. If the torque does not drop, adding more will make no difference but possibly knock.

Remember knock sensors listen for noise. They are a simple way to detect knock in street engines as a safety device. Knock should be monitored by pressure not noise. But this is not always possible due to cost. So accept there is always some error involved with tuning with noise monitoring knock sensors.

Also, when setting up your knock threshold, you have to put the engine into knock. Figure out where peak Torque may be and run the engine up to that RPM without any load. Then measure the engine mechanical noise and set the threshold just above this, to start off. Put the engine into knock to check and then back off the number to be safe. Its always a compromise and one to always error on the safe side. But again, I cannot stress enough, the spark plugs are your best friend here and they tell no lies.

Great post Neil! I consider myself well versed in the topic and yet you just added to that. Thanks!

Yeah those setups are just a tad bit pricey!

At the end hes gives an entry level cost.

https://www.youtube.com/watch?v=PRWx1x8UezM

Thanks
Mike G.

If someone with standalone says that engine knocks and CR is known value (and not too high), also fuel is verified octane and no excessive oil consumption then I would first suggest to check your base timing with strobe light and verify that 18 degrees is actually 18 degrees.
More often than not people set missing tooth point vs TDC wrong (not locking ignition during measurement) or just put some random number there (because engine works) and forget about that.
Since dyno also shows lower numbers than expected (though it should not be used to get absolute numbers anyway), I would just buy few litres of toluene or xylene and mix it with regular gas (3 litres of xylene/toluene with 7 litres of gas) and you should not hear ANY knock on n/a engine.

Good point on verifying timing with a timing light if not done already