Charge temp spark adv subtration.

[vlad01]

So for the longest time I always wanted MAT sensor vs MAP spark retard to prevent knock in very spesific and harsh heat conditions on the 808 and 12P etc...


I only ever saw one called charge temp vs map, advance bias (term subtracted).

I assumed this used the calculated charge temp which uses both MAT and CTS and probably other factors to model and calculate an arbitrary number.

What I found is knock would only occur in air temp manifold heat soaked conditions independent of coolant temp. Examples were low gear 50-60km/h behind slow cars and trucks, then gunning it around them once a lane split off. Same spot each day would yield a sec or two of knock that disappeared once the air temps in the intake dropped, even if the coolant was at max temp or not even at operating temps. Another spot coming out of a tunnel did the same each day. Very repeatable.


So I was bored and looked more at the charge temp stuff and decided to run the trace bubble but didn't have a car I could just real time at midnight lol. So I thought, hey I got logs I can look at. And to my pleasant surprise, the name of the table is actually wrong going by the trace bubble. It is MAT vs MAP advance subtraction bias


Finally might be able to prevent knock happening at all in all conditions without any compromise or band aids.

See the screen shot and note the trace vs the MAT and the calculated charge temp.

Screenshot 2023-12-16 at 14-26-22 Charge temp spark subtract.png (PNG Image 1366 × 1860 pixels).png

[immortality]

Good work but if you're going to gun it past a slow truck wouldn't it help to drop a gear or 2 (you have a manual don't you), generally a sudden increase in air speed also sees a drop in air temp, low rpm and closed throttle is a recipe for heat soak. Then I also wonder if the GM MAT sensor is rather slow reacting to actual conditions in the intake? Does it take longer to read a cooler temp than it does to heat up?

[vlad01]

No actually, the condition is when I am already in 2nd or 1st for a good while waiting to pass sitting at about 3.5k

Doesn't happen when holding a high gear, even if I drop down. The issue is, in this sort of traffic the car is too fast to control the speed in the higher gears without having to constantly swap foot from throttle to brake and back and gear up and down constantly. Holding a lower gear in high rpm makes it a one pedal car at most speeds and moderate to moderate/heavy traffic.

But the higher rpm at light load heats the intake a lot, probably from reversion.


As far as sensor speed, at high rpm the sensor is reasonable quick. At cruise speeds its slower due to far less air density in the runners.

Given that knock in this condition is not relying on the sensor to heat up to read the condition, but a result of slow built up and then only happens when going WOT, the temp should drop down rather quick. So I think there is little risk that this table won't catch the condition and the cool off should be fast enough not to retard the timing too long. But a slight loss in power for only WOT for less than several sec and in these rare conditions is nothing I would even care if it stops the loud eat piercing knock that happens oh and obviously protecting the engine from excessive wear and potential damage is also a priority.

Although, evidence of such violent knock on another engine I had that came out of the red car, (sitting as a spare in the shed now), a bore scope of all bores showed zero evidence or damage from that. These engines are truly bullet proof, at least to a certain caliper of riffle.


Btw, this does not happen on 98 ron, but that's too expensive for daily driving with the Ks I do.

[Charlescrown]

I would think that by logging the IAT would show if the problem was caused by high intake air temp. Looking into the bore on an engine that has suffered heaps of knock may be evident by piston and valve damage not the bore. In lots of cases the ring lands break and are not evident by visually inspecting.

[Charlescrown]

I would think that by logging the IAT would show if the problem was caused by high intake air temp. Looking into the bore on an engine that has suffered heaps of knock may be evident by piston and valve damage not the bore. In lots of cases the ring lands break and are not evident by visually inspecting.

[pman92]

Is it the name of the table that's wrong, or the bubble that's wrong?

I guess the bubble is defined by the XDF somehow.

If the name is right but bubble is wrong it might send you down the wrong path.

[vlad01]

I might, it's something I will need to test, but the left axis suggests it's MAT temp given the steps and range.

[vlad01]

You were right, it is actual charge temp, the ADX must be wrong. I tested it at 2 cells, one that matched the air temp, one that matched the calulated charge temp. I used -20 spark so it was obvious, and it confirmed it's not the air temp but rather the calulated charge.


So I stuffed the front of the car's air inlets with newspaper and blocked off 2/3 of the rad with paper too, did the usual 40-60km/h in 1nd for several min, also 80km/h in 2nd, coolant temps got up there, somewhere around 100-103, normally I never see more than 85 in the real life condition where I get knock. Air temps didn't really exceed 45-50. No knock no matter what I tried.

So this confirms, it's not a heat soak issue per say, but def an air temp problem. I guess the heat from the traffic and the ground air temps with the slow moving air due to the traffic must really soar those air temps. I will need to literally log some 40min to an hour to get the spot where it happens. But I am guessing the charge temps are probably exceeding 100+ given I saw 85 peak with 103 coolant temps.

[krusty]

Not past MBT for that given scenario of load by any chance?

[immortality]

You were right, it is actual charge temp, the ADX must be wrong. I tested it at 2 cells, one that matched the air temp, one that matched the calulated charge temp. I used -20 spark so it was obvious, and it confirmed it's not the air temp but rather the calulated charge.


So I stuffed the front of the car's air inlets with newspaper and blocked off 2/3 of the rad with paper too, did the usual 40-60km/h in 1nd for several min, also 80km/h in 2nd, coolant temps got up there, somewhere around 100-103, normally I never see more than 85 in the real life condition where I get knock. Air temps didn't really exceed 45-50. No knock no matter what I tried.

So this confirms, it's not a heat soak issue per say, but def an air temp problem. I guess the heat from the traffic and the ground air temps with the slow moving air due to the traffic must really soar those air temps. I will need to literally log some 40min to an hour to get the spot where it happens. But I am guessing the charge temps are probably exceeding 100+ given I saw 85 peak with 103 coolant temps.

Increase you're following distance?

I had a similar thing occur coming out of Piha one summer day, one of the hottest days of the year stuck behind a large tip truck climbing all the way up a steep incline, much of it in 1st gear getting heat soaked from the truck, engine and trans temps all shot to the roof and finally overtaking the engine was not happy.

I made a cold air intake that takes the air from under the front bumper instead of through the grill like the factory intake does and this seems to help keep temps down.