Lean running Legnum on the dyno

[eddyvr4]

Thanks steve, Lee and i were scratching our heads with my logs and thought you might see something that stood out as an indicator of my knock and if that o2 reading on open loop did mean lean.. but only if you can find the time..
ill check the engine mounts.....

 

[Kaldek]

wtf, zero knock on the dyno. Maybe some 1s and 2s as it came on boost and at high rpm and now it looks like 120 counts yet it only goes to 36 counts. Are you logging the right thing there? Or are the scales on that graph off.

Sorry yeah the graph is off. Max knock was 13, so the scales are all off.

Anyway, I would think it was the o ring that was the issue for the lean stuff. As for what it was doing on the day, it seemed to be worse with more manifold pressure, like it was not increasing fuel pressure with boost. rpm made no difference, so not flow related, purely pressure dependant. Given we tried bypassing the cold start solenoid, changed the fuel pressure reg, swapped mafs all with zero effect and then dodgy fixed the oring and it made it not right but way better I'm sticking with the o ring.

Are you saying that when we changed the fuel pump on the day that it got measurably better? At the time I don't recall you saying anything with that level of confidence, but if that's how you feel then I feel MUCH better.

As for the knock can you email me a csv log of
rpm
2byte ecu load
tps
coolant temp
air temp
knocksum
wgdc
timing

Sure. Is the "2byte ECU load" what EvoScan calls "Load MUT 2Byte Mod"? I'm capturing Load1Byte already but it's permanently zero.

only other thing, I can't tell from that graph on a phone screen but is the knock occurring as it comes on boost or straight after a big throttle increase? If so, what are your engine mounts like, especially the roll stops.

The knock counts are slightly delayed after throttle. Could be the roll stops although I don't recall seeing any damage on them when I changed the clutch.

 

[Kaldek]

pump pic looks ok to me, with the bit sticking out the top.. make sure the rubber o-ring you are using has a inner and outer diameter that seals on BOTH surfaces

Yep, used the O-ring provided with the Walbro kit. Was nice and tight.

 

[TME_Steve]

Yes that's the correct load. If you can send me a log that'd be great. And yes it was markedly better after the o ring. Before the change targets of 8 were resulting in high 10s, after the change targets of low 10s were resulting in low 11s.

 

[Kaldek]

And yes it was markedly better after the o ring. Before the change targets of 8 were resulting in high 10s, after the change targets of low 10s were resulting in low 11s.

Well now that is quite a difference! But technically we're saying it was a marked difference after the replacement o-ring and fuel pump.

 

[Kaldek]

Yes that's the correct load. If you can send me a log that'd be great.

Will do. I'll add a few more parameters like MAF voltage too if it lets me.

 

[TME_Steve]

Just the ones I said, then it goes faster, the more you log the slower you log.

 

[Kaldek]

OK then. Can't do it until tomorrow, the wife is out tonight and I'll be kid wrangling.

 

[jungle]

Is the "2byte ECU load" what EvoScan calls "Load MUT 2Byte Mod"? I'm capturing Load1Byte already but it's permanently zero.

Sure. Is the "2byte ECU load" what EvoScan calls "Load MUT 2Byte Mod"? I'm capturing Load1Byte already but it's permanently zero.

yes Load MUT 2byte is the one you want. Steve has an older version of Evoscan.
Unless you have the Tephra mod you won't need Load1byte.

 

[whats_doing]

Tephra Mod?

 

[Kaldek]

Tephra Mod?

Basically a custom operating system (ROM) for the ECU. It adds a bunch of features to the ECU. Check Engine Light when knock detected, No Lift Shift, Valet mode, Launch control, etc.

 

[whats_doing]

Do Want!

 

[jungle]

Basically a custom operating system (ROM) for the ECU. It adds a bunch of features to the ECU. Check Engine Light when knock detected, No Lift Shift, Valet mode, Launch control, etc.

Correct plus more resolution in timing and fuel maps

 

[Kaldek]

Do Want!

Pretty sure Steve offers a tune that includes Launch control and CEL on Knock for an extra $50 or so. Not sure if it's based off Tephramod.

 

[TME_Steve]

Tephra is evo , he's done the Evo 7-9 stuff, some evo X as well, he uses Rajab's dyno by the way, Kenneth is the man that did the first lot of mods here, based on the evo 6 code written by Ceddy and now BCX is going a different path to all of the above but will include all that stuff.

 

[Kaldek]

OK Steve log sent via PM. Not public for obviuos PoPo reasons.

 

[Dice]

Oh I think there's a bug in evoscan where it shows the speed is higher than actual

 

[Kaldek]

This data is very interesting; I also recorded the knock sensor voltage. It's hard to say with certainty of course (due to the sampling rate being about 210hz which is very low), but the knock sensor voltage never exceeded 2.5 volts. What's even more interesting is that when knock was detected (max of 6 in this run), the knock sensor voltage wasn't any different to other times.

With such a low sampling rate though I could be missing the knock voltage peaks. I'd need to backprobe it with my oscilloscope for certainty, but you'd think if it was *really* knocking it would peak enough that EvoScan would catch it.

 

[Kaldek]

Vey interesting. Seeing some knock come on here when the throttle is planted even at low rpm. The knock sensor is also constantly producing a signal when rolling.

Both knock counts seen here (a very small one at the end) are as the throttle is planted. I don't know if the knock voltages are normal as I have nothing to compare them to, and each car is very different. Some don't show any signal unless there's a well defined metallic sound, whereas others are more sensitive and have code in the ECU that adjusts the internal knock sensitivity based on engine rpm, road speed, etc.

 

[naughtika]

how old are your engine mounts?

Im thinking that the ecu's sensing false knock?