|
carsey
|
 |
« Reply #15 on: July 19, 2014, 05:27:43 PM »
|
|
|
Probably whats causing it to tail off then. Left untouched. ill try get a good 4th gear pull tomorrow and see where things are.
|
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #16 on: July 20, 2014, 04:25:14 AM »
|
|
|
Got myself a good 4th gear pull. And the logs backed up my thoughts of dropping boost after 5500rpm. Looking at a 5psi drop from 15-16psi to 10psi. Wastegate Duty including before and after linearisation is showing the drop, so would appear something else further up the PID is dropping off causing the drop in boost - unfortunately bit beyond my basic knowledge at this time.
Ive included the log showing the issue. Hopefully, someone can point me in the right direction so I can try get my head around and understand the maps further/higher up in the LDRPID.
Only other likely cause is the KFLDIMX axis which I have set up as:
0hPa - 0% DC
800hPa - 0% DC
900 - 42.5% DC ....
1000 - 42.5% DC ....
1100 - 42.5% DC ....
1200 - 42.5% DC ....
1400 - 42.5% DC ....
1500 - 42.5% DC ....
|
|
|
|
« Last Edit: July 20, 2014, 04:30:13 AM by carsey »
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #17 on: July 20, 2014, 06:55:32 PM »
|
|
|
Please log these: ldimn_w ;{BoostPID-I-Min} ; {Aktueller Wert zur Minimalwertbegrenzung I-Anteil LDR}
ldimx_w ;{BoostPID-I-Max} ; {Aktueller Wert zur Maximalwertbegrenzung I-Anteil LDR}
ldimxak_w ;{ChargeLimitPID-I} ; {Aktueller korrigierter Begrenzungswert I-Anteil LDR}
lditv_w ;{BoostPID-I-Result} ; {LDR, Tastverhältnis vom I-Regler (word)}
ldptv ;{BoostPID-P-Result} ; {LDR, Tastverhältnis vom P-Regler}
ldrdtv ;{BoostPID-D-Result} ; {LDR Tastverhältnis vom D-Regler}
also just need a snapshot of these (don't need a full log, just one sample) ldimxa_0 ;{BoostCorrectFactorPID-IRange1} ; {Adaptive Korrektur der LDR I-Regler Maximalwertbegrenzung}
ldimxa_1 ;{BoostCorrectFactorPID-IRange2} ; {Adaptive Korrektur der LDR I-Regler Maximalwertbegrenzung}
ldimxa_2 ;{BoostCorrectFactorPID-IRange3} ; {Adaptive Korrektur der LDR I-Regler Maximalwertbegrenzung}
ldimxa_3 ;{BoostCorrectFactorPID-IRange4} ; {Adaptive Korrektur der LDR I-Regler Maximalwertbegrenzung}
ldimxa_4 ;{BoostCorrectFactorPID-IRange5} ; {Adaptive Korrektur der LDR I-Regler Maximalwertbegrenzung}
|
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #18 on: July 21, 2014, 05:04:08 AM »
|
|
|
Made a change to my kfldimx axis to begin at a lower hPA, will go out and give that a log when laptop has charged up a bit. Also added the above variables to my logs, I guess they show the results of the maps higher up in the LDRPID. What do the correction factors ldimx_* allow you to look at?
|
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #19 on: July 21, 2014, 05:39:43 AM »
|
|
|
Changing the axis fixed the boost drop off. Logged those ones you wanted a snapshot of and looking through the log, they change numbers at different RPM points. Ive uploaded the file as you will know what to look for.  Thanks Chris |
|
|
|
|
Logged
|
|
|
|
|
phila_dot
|
|
« Reply #20 on: July 21, 2014, 05:42:01 AM »
|
|
|
Made a change to my kfldimx axis to begin at a lower hPA, will go out and give that a log when laptop has charged up a bit. Also added the above variables to my logs, I guess they show the results of the maps higher up in the LDRPID. What do the correction factors ldimx_* allow you to look at?
I max adaptation |
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #21 on: July 21, 2014, 09:23:30 AM »
|
|
|
You're going to need more DC via KFLDRL
In any case, you are the limit of your turbos and you're requesting too much boost.
|
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #22 on: July 21, 2014, 10:21:32 AM »
|
|
|
Always knew of be at the limits of the turbo top end. K03s isn't going to produce a great deal of boost at too end of rev range.
Strange how I see duty is 95%. IMX is set at 75% top end I believe with a linear kfldrl so unsure how it's going up itself.
Do any of you know the addresses for kftarx? Tried to copy it over by matching things up with a full ols of a 018cb but it never works out right.
Aside from the turbos being on the limit, does everything else look safe? To me fuelling is following specified afr, and timing correction factors are low, which should give me some room for timing advance to add in
|
|
|
|
« Last Edit: July 21, 2014, 10:23:08 AM by carsey »
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #23 on: July 21, 2014, 10:27:09 AM »
|
|
|
Strange how I see duty is 95%. IMX is set at 75% top end I believe with a linear kfldrl so unsure how it's going up itself.
I max adaptation, possibly. There are a few other things that go into the i-limiter as well (including KFTARX as you already mentioned) Aside from the turbos being on the limit, does everything else look safe? To me fuelling is following specified afr, and timing correction factors are low, which should give me some room for timing advance to add in
before going any further please make sure your ldr deviation isn't this terrible... especially once you reset your i-limit adaptations. |
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #24 on: July 21, 2014, 10:36:01 AM »
|
|
|
Thanks, I'll try get the kftarx map correctly addressed up in my map pack.
Must be one or two maps I've missed, is some are also having an affect on engine load corrected; which should be up out the way whilst I tuned boost.
What should I be looking for in regards of the ldr deviation? Guessing a lower number is better. I'll have a look over the FR see if I can see the maps that's goes into to I part of the pid. (Getting complicated now for me haha)
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #25 on: July 21, 2014, 10:37:56 AM »
|
|
|
What should I be looking for in regards of the ldr deviation? zero. I'll have a look over the FR see if I can see the maps that's goes into to I part of the pid. (Getting complicated now for me haha)
You're overthinking things. Stop requesting too much boost, and lde will be zero. |
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #26 on: July 21, 2014, 10:44:56 AM »
|
|
|
Thanks. As I said, I raised the specified boost up out the way so there was nothing stopping me tuning my duty cycle to get the boost plot I wanted to aim for. I should be able to bring my requested engine load back down to a level just above the actual engine load. Would this bring the LDR adaptions back into check?
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #27 on: July 21, 2014, 10:54:34 AM »
|
|
|
Thanks. As I said, I raised the specified boost up out the way so there was nothing stopping me tuning my duty cycle to get the boost plot I wanted to aim for. I should be able to bring my requested engine load back down to a level just above the actual engine load. You're actually not really worrying about spec load. You should be tuning LDRXN to result in sane boost request values. ALWAYS set your req boost to what you want. There is no reason to set them out of the way, the whole point of the PID is to get boost to match req. Would this bring the LDR adaptions back into check? Once lde is zero, yes. |
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #28 on: July 21, 2014, 11:13:45 AM »
|
|
|
Sounds good. Thats roughly where I want the boost to end up (what im actually seeing now). I can see there might be a small window to add a touch more on spool between 2750rpm and 4000rpm, but looking at the actual figures, theres only really 200mbar of boost left....so thought its not worth bothering with as under different loads, it can sometimes change (hence why im now using 4th gear to tune as 3rd was proving a bit troublesome in higher gears if you remember my other threads).
Scaled my ldrxn to a few points above the requested boosts on the plots I took today. Ive included a pic. Hopefully that doesnt get it the way, and is enough the sort the LDR-adaption out.
Even though still probably only a very basic tune, its really enjoying working with the data you log, and working out where you can improve, and what differences it does to the car.
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #29 on: July 21, 2014, 11:21:26 AM »
|
|
|
Scaled my ldrxn to a few points above the requested boosts on the plots I took today. Ive included a pic.
Actually, that is definitely one way to figure out what you want spec load to be... just set it to actual load from your logs, if your boost is right But keep in mind the calc method can be pretty far off, and the relationship between spec load and spec boost isn't the same as the relationship between actual boost and calculated load. Hope that makes sense... |
|
|
|
|
Logged
|
|
|
|
|
