Curing Detonation: a theory

[SC SiR]

Just a theory I have regarding detonation. One cause as we all know, is lack of fuel. The other is too much timing.

Tuning my car, I set timing to 8 BTDC and get no detonation. I then set it to 14 BTDC and the MSD 6BTM to retard 1deg/psi. At max boost, it retards to 7-8 BTDC. It knocks. Even at 2deg/psi, it would still knock. And yes the BTM works coz at WOT, and I turn the dial, I could feel it retard/advance.

It was knocking at tip-in and part throttle, as all of us experience. Yet, my A/F meter was well into the rich. I figured the BTM wasn't fast enough to retard the timing. The knock stops at 10 BTDC and BTM set to 0.4deg/psi.

Of course we can dump in more fuel and since it is so rich, the mixture won't fire properly or fast enough. The knock stops, and we think we've solved the problem.

Not unitl now, when I got my PMS w/ Interacq software did I find the cause. Of course it is only a theory on my part. Anyway, I used the data logging feature to record everything. The knock happens right at the point where boost is about to build (from 3"Hg to 0"Hg vac). Then, it hit me.

Our engine's timing requirement depends on two major factors. RPM & Load/Dynamic Compression (DC for short). We all know about the RPM portion. However on the second one. We only know half the story.

Normally, we think, as we feed the engine boost (increase in DC), the mixture is more volatile and we need to retard the timing. What we are forgetting is before boost, DC is already increasing as soon as the bypass valve starts closing. In short, we have to start retarding even before boost. That explains why even with an MSD BTM/6BTM we can't advance the timing to close enough to stock.

Some might argue, but you see 3-0" Hg. vac at WOT with the stock engine and get no detonation. That's because the ECU knows you are seeing that manifold vac. With the SC, since the MAP sensor is before the SC, the ECU sees a much lower vac. It thinks you are at a particular DC and timing is given as programmed. Actually our DC at that point is much higher. The turbo guys don't have that problem because the MAP sensor is after the turbo.

That presents a unique problem over which only SC people tear their hair off.

Our solution: We need more $$$. Don't we all? We need a more sophisticated timing control. Not something that would retard timing in relation to rpm only or boost only. We need something that could adjust timing in relation to rpm and load/DC.

In my case I use the PMS. A MOTEC, DFI, etc. would do the job. Problems are: First, it ain't CARB legal. Second, its harder to tune. Third it costs a lot of cash. That's the price to pay for instant boost.

So, this is what I have on my B16. Timing at 14 BTDC. Retard to 7-8 BTDC by max boost of 7-8 psi. Most importantly, at the point where manifold vac is about to go 0" Hg, retard timing by 2 deg. This load point would be different for each rpm. Slowly, I will try to raise my timing to 16 BTDC.

If you don't have that much $$$ to spend on the timing controls, I have some recommendations:

If you have a BTM, set your timing to a point where you get no knock. BTM at 0. Remember that timing. For this example, lets say it is at 9BTDC. Now, start advancing your timing and adjusting the BTM to go to 9BTDC, more or less, at max boost. Find the max distributor timing you can get and still get no detonation.

If you have an rpm controlled timing (Apex ITC), advance the distributor slowly. Use your control to retard at the rpm you hear knock.

If you have both the above, just try different intelligent combinations, but be systematic.

A lot of this stuff as we know is trial and error.

Keep in mind that different engine have different characteristics. So, mine might start knocking at a different DC from your engines. But keep the theory in mind when tuning.

Any questions, just ask. Any arguments, I'd like to hear them.

Towards a knock detonation free SC community.

[lundgren]

From what I know the BTM won't really deliver more power, since you are right, you still must retard quite a bit to relieve tip in detionation. The reality is, you need lower timing mostly to control lower rpm detionation not really high RPM detonation.

Those with the J&S Safeguard have the advanatage, since we can see when the engine has knock, before it's heard audiably. I've noticed exactly what you are saying, and outside of a reprogrammable ECU, here is what I think the solution is, or what some factors may be.

1- The VTEC crossover WILL set your J&S off. If it doesn't, then your J&S is not sensitive enough. Don't worry about that 1 degree retard at 4500 (B18C).

2- I think it's possible, and I haven't confirmed it yet, there may be a little blow by at closed loop operations on tip in under boost, on some systems, where the ECU takes fuel out, because the o2 sensor thinks it's rich, but it's not. This might occur in cars with cam gear dialed over lap, with open exhaust systems. This might cause a lean period before open loop, when no lean is required.

3- It think the JR rise, doesn't react fast enough. With the cartech, a middle pressure (just before 0 psi), where the ramp pressure can be set, in addition to max boost pressures. I think having a little more fuel at that point, helps.

4- V-AFC. Setting part throttle and full throttle positions at lower RPM, this unit nearly acts like a tip-in helper. So at lower rpm's there is a steep inrichment, but only on the full throttle 'Map'. Part throttle, it's normal enrichment, then as the throttle is depressed, it adds fuel.

5- J&S Safeguard. The Safeguard is really nice for tip-in detonation. I NEVER see knock at higher rpm's. It's all tip-in low rpm type activity. The Safeguard lets me take ignore, for the most part, the small amounts of knock, since it removes it very quickly, and I usually never even hear it audibly.

6- Larger injectors. The stock injectors are taxed. At 6psi you're ok, but at 8 psi you should get larger injectors in conjuction with the cartech or vortech FMU's.

All of these require money. I think the JRSC in it's stock form is really good, but once you start playing... then it gets expensive to make it work right.

------------------
http://www.netbsi-kc.com/integra/
http://www.netbsi-kc.com/aa/
http://www.groupbuycenter.com/

[SC SiR]

I forgot to add. I wrote the post above with the assumption that fuel mixture is correct all through out. But, knock is still present.

To solve my fuel problems, I have RC 370cc/min injectors. No JR FPR and no JR enrichment circuit.

Who said our hobby was a cheap one?

Lundgren:

You're absolutely right. Most of the knock happens at lower rpms. On my car, 3k to 5k rpm.

With regards to #2 of your post, by blow-by do you mean the intake mixture going out of the exhaust coz of overlap? If it is, then I think the O2 sensor would sense it as a rich not lean condition. The sensor senses O2. And since in the blow-by, none of the O2 has combined w/ the gasoline, the sensor would see all of the oxygen.

BTW my JRSC is still in stock form, pulley wise. However, I live in the Philippines, and our cars don't have O2 sensors from the factory. I had to put one in myself. We are alway in open loop operation and the ECU can't adjust for minor rich or lean conditions. So, the kit detonates horribly straight out of the box.