Ralliart TD04 spool-up - the load calculation problem & the solution
#16
BINGO.
Today at the track, I did half a dozen test runs. Not a single instance of spool-up knock. Not one. Unbelievable. She ran smooth as silk every time.
After four clean runs, I started some gentle prodding at the performance side of the tune:
I increased boost targets by 1psi across the board... No knock.
I then advanced timing 1 degree from 5000rpm in WOT areas.... and saw a single instance of knocksum=4 (7000 rpm / 180 load, AFR 10.9:1, TimingAdv 20*).
That only occurred once in the run. That was the only time I saw any knock, all afternoon - the run after that ran clean as a whistle.
The logs show that "ChosenCalc" is far better behaved now. Still, when I go WOT at 4000rpm, the load does seem to be a TINY bit below where it should be... so I might try a very slightly higher "Load Ramp Rate", just to see if I can get it 100% on-target.
I'm being seriously nitpicky now, I know. Here's a graph showing how damn close to perfect it is... (warning: You really have to squint to see the imperfection, lol)
I should say that the car was noticeably more zippy with the mild boost/timing changes.
I feel I finally have room to start tuning for torque/power. I can take my eye off the KnockCEL flashy light thing while I'm lapping. I can try out leaner WOT mixtures without serving up leaner spool-up mixtures. Awesome! About time!
Rich
Today at the track, I did half a dozen test runs. Not a single instance of spool-up knock. Not one. Unbelievable. She ran smooth as silk every time.
After four clean runs, I started some gentle prodding at the performance side of the tune:
I increased boost targets by 1psi across the board... No knock.
I then advanced timing 1 degree from 5000rpm in WOT areas.... and saw a single instance of knocksum=4 (7000 rpm / 180 load, AFR 10.9:1, TimingAdv 20*).
That only occurred once in the run. That was the only time I saw any knock, all afternoon - the run after that ran clean as a whistle.
The logs show that "ChosenCalc" is far better behaved now. Still, when I go WOT at 4000rpm, the load does seem to be a TINY bit below where it should be... so I might try a very slightly higher "Load Ramp Rate", just to see if I can get it 100% on-target.
I'm being seriously nitpicky now, I know. Here's a graph showing how damn close to perfect it is... (warning: You really have to squint to see the imperfection, lol)
I should say that the car was noticeably more zippy with the mild boost/timing changes.
I feel I finally have room to start tuning for torque/power. I can take my eye off the KnockCEL flashy light thing while I'm lapping. I can try out leaner WOT mixtures without serving up leaner spool-up mixtures. Awesome! About time!
Rich
#18
I think the best approach is...
1. Wait for the published XML definition for "Load RampRate #1" to appear on Golden's website.
2. Book the car in for a touch-up tune.
3. Tweak "Load Ramp Rate #1" to something higher (still testing the ideal value).
4. Get your tune touched up.
Your tuner will probably find the fuel and timing maps can be altered in spool-up areas to work more cleanly.
Rich
PS. There's another load-related tidy-up coming too... to get rid of those weird "steps" that sometimes appear in the load graph, for no good reason. Wait for that stuff as well. Won't be long.
1. Wait for the published XML definition for "Load RampRate #1" to appear on Golden's website.
2. Book the car in for a touch-up tune.
3. Tweak "Load Ramp Rate #1" to something higher (still testing the ideal value).
4. Get your tune touched up.
Your tuner will probably find the fuel and timing maps can be altered in spool-up areas to work more cleanly.
Rich
PS. There's another load-related tidy-up coming too... to get rid of those weird "steps" that sometimes appear in the load graph, for no good reason. Wait for that stuff as well. Won't be long.
#20
Brilliance Rich. I see what you've done der! Ha!
Since you doubled your load there and got nearly right on the money, why not go up to like a 10% scaling? Wouldn't that fix that little burp you had?
Since you doubled your load there and got nearly right on the money, why not go up to like a 10% scaling? Wouldn't that fix that little burp you had?
#26
Hi Rich,
Just want to super confirm here ... would this recent discovery benefit BBX Lite turbos in your opinion? My last track day, only 2 days ago by chance, was knocking like a girl scout with a fresh box of cookies.
Just want to super confirm here ... would this recent discovery benefit BBX Lite turbos in your opinion? My last track day, only 2 days ago by chance, was knocking like a girl scout with a fresh box of cookies.
#27
If it was knocking during spool, once boost started getting up there, I'd start with this thing, yes.
Especially if your AFR shows that weird thing where it seems to stay lean until boost has peaked, then very sharply drops a full point to where it should be.
Rich
Especially if your AFR shows that weird thing where it seems to stay lean until boost has peaked, then very sharply drops a full point to where it should be.
Rich
#30
Hi folks.
Golden's website now has definitions for "Load Ramp Rate #1 (Load>70)". This is defined for the fave Ralliart ROM IDs.
First post updated with a gratuitous EcuFlash image.
The factory value for this is: 3.53%
I've tested out 7.06% and it helped ChosenLoad keep up with fast-rising load. I'm not sure it was quite enough to handle a 4000rpm throttle-stomp, but very close.
Next time I'm out on track, I'll be trying 8.63%.
Rich
Golden's website now has definitions for "Load Ramp Rate #1 (Load>70)". This is defined for the fave Ralliart ROM IDs.
First post updated with a gratuitous EcuFlash image.
The factory value for this is: 3.53%
I've tested out 7.06% and it helped ChosenLoad keep up with fast-rising load. I'm not sure it was quite enough to handle a 4000rpm throttle-stomp, but very close.
Next time I'm out on track, I'll be trying 8.63%.
Rich