EDC16 REMOVING MAF DIAG LIMIT AT 1275

[vwstuff]

I found :

UACC_RSigBlkGrp_MAP 1F28A8
UACC_ESBFormular_C 1F1CA8
UACC_ESBNormVal_C 1F1ED9
UACC_ESBSigNr_C 1F210A

But I wonder where are located :

sig_qDIAScale_C
sig_qDIAOffset_C

And how to process
I'm stuck !!!!:bawling:

[DiegoStud]

1C12AE- Sig_qDIAScale_C
1C12AC- Sig_qDIAOffset_C

[vwstuff]

1C12AE- Sig_qDIAScale_C
1C12AC- Sig_qDIAOffset_C

Thanks a lot Diego:ciao:

[vwstuff]

So : b/4*a*0.1
This is 255/4 *200*0.1 = max at 1275 Mg/Str

1C12AE- Sig_qDIAScale_C value on my car is 0.037
but : (255/4)/200 *0.1 = 0.032 what is wrong ?!

[DiegoStud]

Your welcome, although they are not what you require.
Signal 258 is AFSCD_mAirPerCyl__mAirCyl_DLen1 therefore you need:
1C11C6-Sig_mDIAScale_C
1C11C4- Sig_mDIAOffset_C.

Including modification of NormVal you should be able to read 1625mg/stroke. Beyond that you need to change formula.

Edit: My bad-
1C118A- Sig_mAirCylDIAScale_C- 0.02
1C118A- Sig_mAirCylDIAOffset_C- 0

255/(1275*10)= 0.02

[vwstuff]

Is 1625mg/stroke ( a=255) enough for a 2 liters with a gtb1756 ?


1C1188 for Sig_mAirCylDIAOffset_C

[DiegoStud]

Assuming you want a bit of extra scope to check for boost leaks, 1625mg would be fine for about 3200mbar boost. Plenty.

Remember to modify Sig_mAirCylDIAScale_C to 0.0156.

[DiegoStud]

The following signals are dependant on Sig_mAirCylDIAScale_C, therefore the complete job would also require looking up these signals in UACC_ESBSigNr_C and modifying their respective NormVals also.
300 "AirCtl_mDesVal__mAirCyl_DLen1"
428 "AFSCD_mAirPerCyl1__mAirCyl_DLen1"
429 "AFSCD_mAirPerCyl2__mAirCyl_DLen1"

[vwstuff]

My goal is afr calculation ;
smoke limitation via lambda seen to be not efficient ?!

[Meat-Head]

Wowsere deigostud, your amazing, part of the Spanish Inquisition and you can speak
Numbers and smoke, well done

[vwstuff]

Another fan :courage::courage:
Finaly i'm not able to do it ... some think is wrong ?

[DiegoStud]

Thanks Meat-Head!

Smoke limitation by lambda is not what it sounds like on these engines. Lambda is a calculated mAirPerCyl value based on airflow rather than using the lambda sensor directly, which is an incredible disappointment I know.
So in other words it's as efficient as your MAF.

The lambda sensor is used for long-term adaptation in EGR and FMA correction, regulating O2 concentration in the cat/DPF and also to regulating post injection.
In a way, view smoke limitation as controlling the fuel burned in the cylinder and the lambda controlling from the exhaust valves onwards.

With the the diagnostic limits raised though you might be able to log and make you own calculations. From my own calcs if you are still around 2850mbar you shouldn't really be that much further beyond 1275mg.

[vwstuff]

Yes really disappointment regarding the lambda sensor working condition on edc16 ....
My diagnostic limits raised are NOT working properly as mentionned where I'm wrong ? !

[DiegoStud]

194409-Another block of NormVal that needs changing- I'm assuming it has been forgotten as NormVals for 24,25 and 26 are changed and the Formular nearby has been changed to account for boost limits.
Give that a try and report please.

Beyond that it's starting to get difficult and I'll have to ponder I little further- I can see you have made changes to MAF SRC which might help beyond 4000rpm.

Edit: P.S- having RPM and MAF together in one log would be great to make some calculations

[vwstuff]

Sorry i'm confuse : what do you mean by 194409 ?
According to my vcds version and UACC_RSigBlkGrp_MAP I need to change on UACC_ESBNormVal_C : three times : blocks 10, 44 ,133 value 7 and block 45 value 327 ! ?

[DiegoStud]

Away so can't see your file but 194409 onwards is also UACC_ESBNormVal_C like 1F1ED9. So make the same change there from 200 to 255 as you have done at 1F1ED9.

[vwstuff]

Got It Thanks :welcoming::welcoming:

[vwstuff]

I fact I need another formule what is the best one ?
Log done at 2750 Mbars request

[DiegoStud]

You must have an intercooler the size of a patio door sitting in a bath of LN2.
Before you go down the formula changing route, have you considered the possibility you may have a boost leak? Or that your MAF is lying?

I mean let's do a bit of quick maths- MAF/PR/CC= rough VE calculation.
@1500rpm VE is around 100%. On the optimistic side but there's a large volume of air being filled in the intake system.
@2500rpm VE is about 130%. Not happening in this universe at least.
@4850rpm VE is about 95%. From an 8v TDI? Whoever ported out your head is a living legend. At this point also there is no delay in boost pipe expansion etc to account for excess airflow, in fact boost is dropping.

According to MAF your peak flow is approx 700kg/hour and remains thereabouts from 3600rpm until 5061rpm.
To get that sort of flow plausible at 2750mbar your post intercooler air temperature would need to be about -5 degC which is.... implausible.

[Meat-Head]

You must have an intercooler the size of a patio door sitting in a bath of LN2

As your off on one, for use leechers, can you tell us if the bath has to be cast iron or will
A regular plastic one do please? Can you tell us if the intercooler can be the
same thickness as a patio door, or does it need to fit the bath?

AWESOME post, although some DK members would not fully understand it.

thanks

[DiegoStud]

What the bath is made of is unimportant, but the patio doors must be 1970s metal framed sliding ones not these small new cheap things with more plastic than glass.

[vwstuff]

Will test a new MAF asap

[vwstuff]

I'm still trying to find sig_qDIAScale_C for max iQ any idea ?!!
Through pressure test I found a leak when pipe mouving ...
I'm waiting for some parts .

I attached a pic of my new intercooler I guess, some body work is mandatory.
I just purchase a Metalwork for Craftsmen book for this purpose.

[DiegoStud]

Nice pic, it would fit in a bath rather well :encouragement:
Well done finding the leak, your turbo will thank you for it.

If sig_qDIAScale_C is not 1C12AE then it's 1C12B6.

[vwstuff]

I'm not sure for the two address stock formule ( 39 ) is b/256*a so factor is 255/70 =3.64
and both values for 1C12AE and 1C12B6 are 0.037 :eek-new:

[DiegoStud]

Yes but IQ values have a factor of 0.01 therefore 255/7000= 0.0364.
Why the original figure is 0.037 rather than 0.0364 is not a question to which I have found a satisfactory answer.

[Meat-Head]

Yes but IQ ..........I have found a satisfactory answer.

Think you meant to say you have an iq higher than einstine, do you have the lottery numbers to hand

[vwstuff]

So With no boost leak and new MAF : logs attached.
I tried to change MAF Formular from 19 to 8 with a = 80 to archive 2040 ( 0.1*a*b) with no succes .
I wonder if there is two UACC_ESBFormular_C maps ?! Formular 8 is NOT take into account by the ECU.
And also my iq is over 85 ....well some think is wrong :sorrow:
Will try to set both 1C12AE and 1C12B6 with same factor values ?
File attached


Albert IQ is 160 ...I'm stuck at 85 ?

[DiegoStud]

Yes in your file you have two UACC_ESBFormular_C- 1941D8, 1F1CA8.

IQ stuck at 85- if you want to log more than 85 why have you only scaled until 85? Would not 100 with 0.0255 make more sense?
I wonder how wise it is to be trying to edit MAF and IQ diagnostic limits at the same time, sounds like you are losing track of what's going on. Modify one at a time, log and confirm.

With regards to the log itself- a little smokey maybe?
What would you like to know? Comparing Lambda to MAF well provides an empirical flow model for the engine. Do I think MAF is still high? Possibly so, and with a larger intercooler on the way the disparity will increase. However as I've already commented on, it's very difficult to try to analyse MAF data during heavy transient response, no single figure of expected MAF can be given to correlate with MAP unless you are willing to run simluations or plug the figures into some horrible equations (Heywood C14).
Considering the above and the fact the majority of the log covers a turbo spooling on an accelerating engine and the diagnostic air flow limitation occurs for just 2 seconds, the reason is clear why I'm not keen on MAF for smoke limitation purposes.

Energy conservation is more Newtonian than Einsteinian :biggrin:

[vwstuff]

Thanks for UACC_ESBFormular_C
My car smoke visibly at 0.9
I don't plan to Inject more than 75-80 IQ
MAF is for testing airflow of my UGLY inlet manifold Prototype (build in process pict attached )

In fact I'm try to understand what is the best way to calculate smoke map and how to process ? I know i'm slow ....:confused:

And also I wish to archive a constant boost from 2000 to 5000 rpm at 2750 Mbars to lower final ergt.My goal is below 800 °C
I notice 2 maps 1EB4CE - 1EB716 and wonder if they act like a limiter or this is simply engine efficiency fail down ?

Speed of light being a constant, with time and length different… don’t ask Newtown about that :itsme:

[DiegoStud]

vwstuff I think I owe you an apology: If I read correctly it is possible to have full load lambda regulation enabled by FlMng_swtFullLdRglVal_C (1D7012) allowing feedback from LSU_rLmbdRec aka Lambda sensor:


As shown in Figure 354, a PI controller with the speed-dependent parameters FlMng_FullLdRglPKp_CUR and FlMng_FullLdRglIKi_CUR is used for control, since the transmission behavior of the line is also strongly speed-dependent.

The control deviation is calculated from the reciprocal of the λ setpoint FlMng_rLmbdLim_mp and the measured 1 / λ actual value LSU_rLmbdRec.
In order to avoid overshoot the controller e.g. when switching on, the control deviation FlMng_rLmbdRecDvt_mp is limited in both directions by the values ​​FlMng_rLmbdRecDvtMin_C and FlMng_rLmbdRecDvtMax_C. A correction quantity FlMng_qFullLdRgl_mp results, which is both positive and negative, as the controller manipulated variable Can assume values. The manipulated variable is based on the values ​​FlMng_qFullLdRglMin_C and FlMng_qFullLdRglMax_C limited and added to the smoke limitation amount FlMng_qLimSmkBas_mp.

When the controller is switched off, e.g. in the transition to partial load, no volume jumps may occur. Therefore, in this case the manipulated variable of the controller is slowly reduced using a ramp function with the gradients FlMng_dqSlpPos_C and FlMng_dqSlpNeg_C. The correction amount FlMng_qFullLdRgl_mp is now determined by the output of the ramp. When switched on, the integrator is initialized with the current ramp value.

FlMng_FullLdRglPKp_CUR (1D603A 7x1 Factor 0.012207)
FlMng_FullLdRglIKi_CUR (1D6018 7x1 0.023283)
​​FlMng_rLmbdRecDvtMin_C (1D6796 1x1 0.0001)
FlMng_rLmbdRecDvtMax_C (1D6794 1x1 0.0001)
FlMng_qFullLdRglMin_C (1D6038 1x1 0.0001)
FlMng_qFullLdRglMax_C (1D6036 1x1 0.0001)
FlMng_dqSlpPos_C (1D7000 1x1 32bit 0.000153)
FlMng_dqSlpNeg_C (1D7004 1x1 32bit 0.000153)

Give it a go if you like, would be interested in how you get on.

With regards to your other points:
Personally MAP based smoke limitation is easy and reliable. Make a map with a calculated AFR and then fine tune after logging lambda if you want. There should be a margin of space between torque limiter and smoke limitation to ensure consistent torque but it depends on what you prefer and how close to the limit you wish to run.

2750mbar at 2000rpm is not an acheivable or adviseable goal; you will hit compressor surge before that.

If you are only going to run up to 80IQ then there's no need to increase diagnostic limits and tune for more.

Those two maps are boost setpoints during DPF regeneration.

[vwstuff]

Hi Diego no need to apology your advices are very wise and useful .
I really appreciat to be in touch with you :beer:

!!!

Otherwise not time yet to work on my car ; I will try the LSU route asap !!!!!!

And to complet my previous question regarding boost I mean in the range of 4000 to 5000 rpm,
I notice pressure curve fail donw in this rpm area and I need 100 mbars + to lower egt by 30-40 °C ?!
I guess this is engine efficiency or I can increase N75 pass 50 % of duty ?

[DiegoStud]

Increase boost until it blows then dial one step back...;)
You need to approach this with a set strategy in mind and work methodically. This includes working through basics first.

For example, I will raise a question I've given you before: what is the limit of your duration maps? Therefore, are you are not injecting the fuel you think you are, and it isn't really possible to make lambda calculations accurately.
Another issue is actual boost not matching specified at the RPMs you speak of- at what RPM does your N75 map end? If specified N75 and actual N75 are not somewhere close to each other (they will always have some variance) then, has the boost control been tuned enough? It ends at 37% DC

If these issues are not corrected, adding a feedback loop from a Lambda sensor is only going to make it harder and harder to tune.
For example, lambda reads richer than expected -> reduces IQ -> makes no difference to duration-> lambda still reads rich -> reduces IQ -> N75 miscal drops boost-> lambda reads rich until limit of lambda regulation is acheived= useless.

With this fine tuning you may find you don't need the extra boost you are asking about, and should you choose to try lambda regulation, it might work correctly. At least in all this your EGT limiter is saving you.
It's up to you, but I'm not keen on full torque running until hitting the limit of some variable like lambda, EGT, air temps, air pressure, what day of the week it is.....

[vwstuff]

I understand well the need of correct all of my mistake before fine tuning https://www.digital-kaos.co.uk/forums/images/smilies/culpability.png

what is the limit of your duration maps? : do you mean limit max of duration 0 ? This 34.6 °! In fact according to VCDS I inject 83.1 Mg/Str .. With axis set at 75...
Do you use extrapoled duration maps from PD 170 ? All my durations are wrong ?

Another issue is actual boost not matching specified : I spend a lot of time on boost actuator to set it properly ;right now, duty cycle on map is very very close to VCDS logs
Will have a much closer look at my boost logs .

[vwstuff]

I extrapoled ( one axis basic ) durations 00 to 03 up to 80 Mg/Str and wonder if this make sense ?
Durations 04 to 05 are stock PD105 for SOI below 9°
Please let me know if this more realistic / adequate durations for the rest of the process :fisheye:

I think monday ?!

[DiegoStud]

Simple answer is yes- your durations maps limit you. If they are only calibrated to 75mg then that is all you will get regardless of what VCDS says- you can look up the durations in your logs and compare.
My point about boost what you were asking about boost between 4000-5000rpm but N75 map ends at 4250rpm- change the axis if you need to.
Extrapolation is fine- I'm assuming the PD150 you are refering to would be using the same injectors otherwise you will have troubles. You can use the same process with the exisiting maps also.

[Meat-Head]

SPAMMITUDE:-

once you got to the bottom of this, got that cast iron bath of liquid nitrogen and intercooler
and another two intercooler baths on the roof, find another ecu for a spare
incase you get issues, twiddle then swap

[vwstuff]

I fitted PD 130 Injectors but unfortunately the durations I gathered on web are from PD150...
Therefore I'm looking for PD 130 durations .
Any ideas where I can source those maps are welcome !

Will add WI on the roof too ....well no need I live in Brittany :biggrin:

[DiegoStud]

The duration maps I've already given you are from a PD130- why not extrapolate from them?

With respect however, I really don't see how this will be of benefit to you. As I understand from your logs, you are running into EGT limits and complaining of AFR being too rich, neither of which will be solved by allowing more fuel to be injected. Would it not make more sense to get the fuel you are injecting now under control, even being willing to reduce IQ if necessary?

[Meat-Head]

SPAMMAGE & ADVERTISING

make sure your sat down not eating or drinking when you see the price

https://www.demon-tweeks.com/uk/spa-design-turbo-boost-exhaust-gas-temperature-dual-gauge-246433/

incase you melt sonething or blow a turbo

#300 beer tokens

[vwstuff]

Ah eh melted pistons and turbo are scheduled asap ; price is a killer !!!

At this level of $$ , I'm more fan of : Polar FIS / https://hajes-racing.com/en/skoda-octavia-ii-1z3-onboard-diagnostic-system-polar-fis
(https://hajes-racing.com/en/skoda-octavia-ii-1z3-onboard-diagnostic-system-polar-fis/)

[Meat-Head]

Well deffo worth some kind of external gauges, otherwise you report back in 5 days with melted piston and turbo
people will accuse you of #post clocking

then should your passenger say “whats the boost pressure running out”
you can say “xyz psi” instead of “no idea, boost gague ran out of numbers at 50 psi”

[vwstuff]

For pistons, I plead not guilty, your honor:devilish:

Otherwise in lambda smoke limitation, it's really hard to archive a stray afr ( 1.3 ) trought the rpm range by durations corrections.
Will try 3 more hiterations but I'm close to give up ....
Also thanks for the PD 130 durations....I'm slow :slug:

[vwstuff]

I stuck regarding smoke limitation MAP extrapolosation I need to reach 2800 Mbars.
If I remplace the 2600 mbars for the 2800 values... there is a big gap from 2000 to 2800 mbar !
I wonder if this is acceptable ?
Also i'm surprise stock PD 170 is very very right , i'm waiting more : 1.2-1.3 ??

[vwstuff]

I still working on LSU smoke limitation .....work in process.My EGT max is now at 788 °C
Otherwise i'm not able to increase N75 duty in the range of 4500-5000 rpm .
Both two 1936EO and 1EC564 limiter are set at 51 % and on log my max duty cycle and up at 44.6 %
I quess there is another limiter somewhere ?!

[vwstuff]

No help diego ???

[DiegoStud]

Too busy to give as complete an answer as I would wish, but here are the salient points:
Smoke/Lambda issues- Smoke map was a CR170 not PD- common rail can run richer than PD for a given smoke limit and 16V better breathing. You can have as big a gap as you wish, values in between are interpolated.
If you want to run leaner, then reduce the figures.
At the same time, 1.3 lambda makes no sense when to spool a 1756- this is the whole reason why smoke limits are a map and not a single figure, or a flat lined map such as you have initially made. Richer lambda will increase pre-turbine exhaust enthalpy, thereby work that can be performed on the turbine, meaning target boost setpoint can be acheived quicker and then once spooled lambda can be leaned out again. Think more 1.1-1.15 lambda at lower IQ and RPM to initially spool. Worth noting on your log boost only meets specified ~3100rpm. Boost following the green line on the graph as shown would be less aggressive and yet still meet specified by ~2600rpm.
689415

N75 issues- are you changing the right values? Look at IQ in your log and compare to your N75 map. ~60mg? 33-21%? PID is doing a good job keeping you up above 40%. Why does your N75 map reduce as IQ reduces?

Nice to hear you've got Lambda limitation functioning- well done.

[vwstuff]

Thank for your feedback, help and time !!!
I will try to correct my boost curve as much as possible according to you green line sugestion , for this purpose I focus only on the 2 main maps : 19118A and 1EC564 maybe i'm wrong ?
The only raison I whish to increase boost at top end is for lower egt.
In the range of 1.1-1.15 I'm still over 800 °C... my bad
750 °C will be fine i guess
Unfortunatly, 02 sensor full load don't see to working properly .
I think another trick is necessary to put it on.
Maybe smoke map MAP need to be blanked just an hypothesis...who know ?!

[Meat-Head]

Just to add to this, can you add somekind of feeder tube into that exhaust so when your over 666 deg c
you can drop into another bath mounted on rear bumper some freshly smelted aluminium ingots.

then every 20 miles sell the ingots at every fuel station for diesel

[vwstuff]

I read the text below on vw tdi forum what do you think about it Diego ?
Maybe full load LSU is in close loop with MAF ?
In other world LSU full load mode will work only in edc 16 MAF mod ?


---------------------

One more function. Full-load Smoke Limitation. This is the oh-so familiar smoke map which is implemented with the MAF on the older engines. (And actually, I assume that the MAF is still used in EDC16 in a inner-outer loop setup similar to the EGR control). Basically, the A/F ratio is directly related to smoke number (how much smoke the engine puts out). The lower the A/F ratio, the higher the smoke number. There's an internal limit in the ECU as to how much smoke the engine can put out... and if the measured A/F ratio drops below the stored A/F ratio related to that smoke limit, the ECU will reduce fueling.

----------------------------

[DiegoStud]

I read the text below on vw tdi forum what do you think about it Diego ?
Maybe full load LSU is in close loop with MAF ?
In other world LSU full load mode will work only in edc 16 MAF mod ?


Yes it is in closed loop with MAF- MAF determines what the target lambda should be (outer loop as he calls it) as the lambda maps use an airflow axis based on mAirPerCyl from MAF, and LSU provides feedback to account for any deviation therefore closing the loop. There is no need to blank any maps, switches will correctly select what it will use.

Regardless, your MAF readings bear no resemblance to what your engine is breathing thus rendering all the above defunct.
As I see it the reason you are trying control lambda is really to control EGT- instead of trying to set up an extremely diffcult lambda control, why not deal with EGT directly and use the EGT limiters?

[vwstuff]

Do you mean to move the max stock egt limiter over 800 °C
This is not really safe ???

[DiegoStud]

Raising limiters wouldn't keep it under control now would it?
What I'm saying is if you are hitting the EGT limiter so quickly you have a choice to make: increase boost, reduce fuel in torque limiter or let the EGT limiter reduce it for you.
Smoke limitation in whatever form, be it MAF or MAP, is for controlling minimum acceptable lambda under transient conditions- ie short bursts.
If you want to keep aluminium ingot consumption under control you can't afford to be driving on your smoke limiter- it isn't designed to do so.

[Meat-Head]

If you want to keep aluminium ingot consumption under control


Excuse my ignorance of being just a leecher, do you not mean “aluminium ingot PRODUCTION” rather than
consumstion

great post there

[vwstuff]

Nothing is under control ...
Now I better understand the way edc16 work thanks to Diego !!
A this time, I stop the remap process to work on hardware ( intercooling , intake manifold, turbo efficiency ) .
My goal is to lower egt !!!
So I stop momently aluminium ingot production