Mad_SB

iTrader: (8)

What up folks,

O.K. have some questions on the 03 rom's, should apply to others as well, but I think the 05's and up have more tables than I do...

I've been logging a fair amount of part throttle knock while cruizing around town and on the interstate. It seems to occure at right around 40% TPS. 40% seems to be the sweet spot for my part throttle acceleration. At this point, the ECU is still closed loop and the short term fuel trim (as measured by o2 feedback in evoscan) shows around 80% so it deff closed loop.

I started poking around the Open Loop Load tables and the Open Loop throttle tables. I'm thinking that the load tables are the calculated load required for open loop transition while the open loop throttle are the TPS output volts required for open loop transition. Can anyone confirm if this is correct?

I'm thinking If I adjust the TPS Voltage (via Open Loop Throttle tables) so that I transition to open loop at between 35% and 40% TPS it will solve my part throttle knock...........

jcsbanks

I would agree with what you are saying. Is it actually knocking or are you simply recording knock sums and the ECU is sorting it out? Additionally, you may need to richen up these fuel zones or retard the ignition. I played around a lot with fuelling to get rid of my lift off knock, but it was part throttle timing that solved it in the end.

Another thing to consider is that some ECUs use certain load/RPM blocks where the ignition is a bit advanced to base a lot of their octane judgement on/switch to low octane etc. If you "fix" these areas the ECU might not be so sensitive to react overall when there is a need for the low octane maps for example. One instance of this I saw was where the knock control only functioned up to a certain RPM and extrapolated what it had learned beyond that. If you tuned out all midrange knock correction it would then detonate up top. May not apply, just things to think about.

jcsbanks

Another point to consider is boost control and dump valve choice. If you have a strongly sprung dump valve and a boost control method that is not compensating well for throttle position (eg ball spring MBC or non-TPS aware boost controller), you can build a lot of part throttle boost. Some people like this because of the crispness, but it is a classic problem with delayed/high throttle open loop thresholds.

Mad_SB

iTrader: (8)

Agreed,

The fueling corrections alone won't work in this case though due to the ECU staying closed loop.. It's already pulling down the fuel correction into the 80% range. I am a little worried about pulling out too much timing in the low load areas where this is occuring, 90-110 load, but I suppose in all reality, the stock maps in this region are geared more for emissions and fuel economy rather than performance so I will revise the timing in the affected map areas.

Now, I thought my utec would show me the tps voltage in one of the loggers but it does not. Has anyone mapped out the voltage vs tps position (It will varry from car to car, but I'm looking for some general values. From the open loop throttle table it looks like the votage increases with throttle position.....

MalibuJack

iTrader: (5)

Yeah the stock map throttleposition below 4000 is about 30% tps, the higher is about 10% tps.. In your case, your TPS may actually need to be adjusted at the throttlebody.. what does it read completely closed? (I'll look up my log and see what mine does, but its just under a volt)

Mad_SB

iTrader: (8)

Well, the simple answer is i'm not sure, I assume it to be real knock based on the increasing counts.. starts at like 3 and moves up to 9 with the ecu pulling timing, this was one specific case and there were other factors, just left homedepot and the car was heat soaked so the intake temp was about 120 degF when I logged the worst case. After a few minutes of driving the temps were back down around 80-85 degf.

So to sum it up, i think it is real knock due to the following factors:
1. IAT high
2. Too much advance
3. closed loop fueling keeping the ratio near stoich.

Mad_SB

iTrader: (8)

About 15 in the evoscan logs, have done a voltage check on it yet. with the throttle closed. Interesting thing is from some of my logs, it looks like the transition is around 65% TPS, that is the point in a slow tip in that the o2 correction goes to 100 and stays there throughout the run. From what i can tell, when the o2 correction (aka short term fuel trim) goes to 100 and stay there it means the ecu is running open loop.

I'll try to dig up some more logs when i get home and show you guys what I'm looking at.

jcsbanks

It is just an 8 bit ADC display in the logs (although most recent ECUs read at least 10 bits on TPS internally for quality of transient enrichment), 5V=255. Mine is 13% closed, 0.65V IIRC.

Have you substantially changed the induction? If so unless you've rescaled the MAF sensor then you'll be reading a different part of the map - often lower = delayed open loop AND advanced AND lean.

Mad_SB

iTrader: (8)

Nope, stock airbox and filter, just cleaned the maf and screen a few weeks back with some crc cleaner.

So is that a straight linear interpolation? Looks like ot from your 13%=.65v (.13*5=.65).

In that case the my 03 rom Open Loop throttle tables would translate to :

RPM TPS
500 60%
1000 60%
1500 60%
2000 60%
2500 60%
3000 60%
3500 60%
4000 50%
4500 36%
5000 36%
5500 36%
6000 35.6%
6500 35%
7000 35%
7500 34%

MalibuJack

iTrader: (5)

I'll try to compare the EvoScan logs with the ECU+ logs since I know the voltage in ECU+ to be accurate..

AutoXer

iTrader: (1)

I believe my OBDII logger has always shown 0% TP when the plate is closed. hmmm... Might the ECU set the zero position when the car is started? Kinda like how a XBOX sets the zero of the analog controllers when you plug them in or turn it on. I'll check my logger later this evening.

I know that my ECU switches to open loop at about 27-28%(OBDII value) in the lower RPM ranges.

Mad_SB

iTrader: (8)

What do your open loop throttle tables look like?

thatsMR2u

iTrader: (19)

yah my voltage on my tps with throttle body closed is .68 volts

AutoXer

iTrader: (1)

I wish I knew. I've been locked out by EcuTek.

Come to think of it, when I did the testing, I think I was on a virgin 03 ROM.

jcsbanks

It is a straight linear interpolation. I've confirmed this with a voltmeter - eg 2.5V is 50% on Evoscan. Whether you choose to work in % of 5V, actual TPS voltage, or 0-255 (where 255 is 5V or near enough depending on the voltage regulator on the TPS - and usually many of the other sensors' supply rail) doesn't really matter. The ECU appears to do no autozeroing using the map or the logs. 100% throttle virtually reaches 5V, 100% TPS on logs.

<geek mode>The other factor to consider of course is that the TPS pot probably isn't linear, but is logarithmic. Also they sometimes sit on a little cam - not looked at my Evo one to see. And then the airflow function for a throttle plate is of course not linear either.</geek mode>