PDA

View Full Version : Boost Control via Factory ECU



Hotwire
25-10-2010, 01:50 AM
Has anyone given boost control via the factory solenoid/ecu a go? The addresses as defined by the Russians appear to be correct given the boost curve observed on a number of stock cars at a recent dyno day held here in South Australia.

So i guess I wanted to know has anyone attempted to change the boost via stock solenoid here on CVR4, or am I likely to be a "pioneer" if I give it a try?

Adam.Findlay
25-10-2010, 12:51 PM
i have just removed my boost controller and plumed in the factory solenoid for this exact purpose. i need to borow the open port cable and figure out what tables to alter but i will let you know if i acheive anything and would be great if you could do the same?

Shtiv
26-10-2010, 01:46 PM
Seems like everyone wants to know. all going to plan I am replumbing my factory solenoid in this weekend to confirm it all so I'll post the results on here also as well as ozvr4

Hotwire
11-11-2010, 01:19 PM
Following on from the posts in the xml creation thread:

Thanks to Michael (justcruzn) @ OZVR4 I put the factory restricted hose in the boost control circuit this avo, and Success! Back to factory boost map.
So, in summary here is the stock boost curve for a WOT run (including 2-3 gear change):
http://i162.photobucket.com/albums/t251/Hotwire33/WOTRun.jpg

and here is just manifold presure for just the 2nd gear component:
http://i162.photobucket.com/albums/t251/Hotwire33/WOTManifold.jpg

And here is the WGDC's from my ROM:
http://i162.photobucket.com/albums/t251/Hotwire33/StockWGDC.jpg

as you can see there is very close correlation, but its not clearly C or D.
So I logged all my general driving then did an XY scatter of all points, and you can clearly see the trend of the two WGDC maps:
http://i162.photobucket.com/albums/t251/Hotwire33/LogWGDC.jpg

Conclusion?
Time to make some changes to target boost and WGDC and see the results :D
I'm pretty happy with the results thus far.

Hotwire
11-11-2010, 01:36 PM
Just compiled a few logs together and got the following result:

It seems there is some compensation factor going on here, or interpollation with the A & B duty cycle maps (all 0s) as below 2000rpm, the WGDC is lower than the 100% required by both C & D maps....

http://i162.photobucket.com/albums/t251/Hotwire33/WGDCGeneral.jpg

Nutter_John
11-11-2010, 04:13 PM
Your getting close Lee to finding the method they use for controlling boost via the stock soleniod , table C looks wrong to me though

Hotwire
12-11-2010, 01:06 AM
John,
Do you have any insights into the boost control you can share? I agree that with C being different to D doesnt seem right, as its not quite the curve that i've logged also. Any idea what the correction may be attributed to and the address it may be found at?
Thanks
Lee

Nutter_John
12-11-2010, 01:37 AM
keep going in the direct you are going Lee things will start to slot into place

Subaru ETA
12-11-2010, 06:56 AM
I think Kenneth was playing around with this a while ago....

Nutter_John
12-11-2010, 11:31 PM
Ok after loooooong chats with Afro here comes the info you need !!

WGDC-C is not what you think it is , it is in fact the TARGET Boost for cold engines
WGDC-D is not what you think it is , it is in fact the TARGET Boost for at operating tempreture

Boost Desired engine load C = gain - duty cycle - Hi Octane Auto
Boost Desired engine load D = gain - duty cycle - Hi Octane manual

all values are load8

If you blow up your engines by alter any of these values don't come moaning to me !!!

:D

kinkyafro
12-11-2010, 11:32 PM
Enjoy :)

Hotwire
13-11-2010, 04:23 AM
Thanks guys, you're absolute champions :D

Shtiv
13-11-2010, 07:08 AM
Yeah that makes sense, I thought the D values looked to be the right thing.... and they seemed to work and yeah BDEL being gain - WGDC makes sense, a few preliminary observations here:

http://ozvr4.com/forums/showthread.php?10312-Manifold-Pressure-Logging-(-amp-boost-control)

Hotwire
15-11-2010, 06:05 AM
Ok had a rather random result today.
Auto & manual tables for gain duty cycle set to the same target, target boost for operating temp altered (but not cold spec).
Logged 2 WOT runs to see how boost came on, and these are the results:

as you can see, on one run I only hit 3/4 of the WGDC, and only a boost of 10psi, not 13 as with 100% WGDC graph.
The car was warm for both runs (sitting around the mid-high 80ºc mark) so I'm stumped. what other factors are involved here.....

Nutter_John
15-11-2010, 07:49 AM
Lee can you post up the expanded views of all four tables (target boost hot and cold , gain man , gain auto )

Hotwire
15-11-2010, 09:37 AM
Here ya go John. I also logged coolant temp this avo, and seems that the temp is going down to around 79-80ºc too, which makes me wonder if its interpollating with the cold table? Seems that way from the graph as WGDC definitely drops when temp gets down that low.
Would be interested to hear your thoughts.
Lee

Nutter_John
15-11-2010, 10:29 AM
Yep you got in one the temp needs to be above mid 80's to be fully on the hot map

it took us at least 20+ dyno pulls to work out what was going on with the temp as we would do a run and check then another and it would be different to the first one

Hotwire
15-11-2010, 10:38 AM
Intersting - I always wondered about my car running "too" cool. I'm then guessing there is countless other maps then that would be affected by running in this temperature region? I see for the fuelling for example, at 79-80 I'd be getting 6% enrichment, do you know of any others?

Seems like I either need to adjust this (can with fuel, but not sure how/where to adjust boost compensation) or get a "warmer" thermostat.....

bradc
15-11-2010, 06:52 PM
What is supposed to be the 'ideal' temp range then?

Patryn999
16-11-2010, 12:16 AM
Not sure, but thought the thermostat only opened at 82 degrees? Ideal operating should be just above that. Any idea why your car runs so cold Lee? maybe it does run right and the thermostat is knackered. Otherwise if you're legit running it at the temperature it is currently then it'd just require a series or parallel resistor to shift the curve and lie.

Went to check the validity of this idea (given that thermal sensor is probably non-linear) and appropriate values but couldn't get into the workshop manuals for some reason.


Also the 82 degrees was pulled out of my butt. Its what the thermostat's on the corolla's are, but they're not exactly turbo'd performance vehicles.

Hotwire
16-11-2010, 12:20 AM
Stock thermostat begins to open at 18º0F/82ºc, and is fully open by about 90ºc. My car came from Japan with a 40mm Koyo aluminium radiator, and G-Reddy hard coolant pipes, so I have a feeling they fitted at 160ºF/72ºc thermostat, which would be fully open at around 80c - which is the temps im seeing.

I could probably only check this by removing the thermostat, but given the temps i see and the modifications to my car, im 90% certain that's the case.

So my next steps are to rectify the thermostat issue (new one), get a leak test done (to ensure no leaks at high pressures) and then to get a high flow cat to ensure its not back pressure causing more boost drop at high revs....

kinkyafro
16-11-2010, 12:29 AM
What is supposed to be the 'ideal' temp range then?

we unscientifically pegged it at 85 degrees C. I generally find my car runs at approx 90 degrees once warm and can hit 105 if stationary after a spirited drive.

Nutter_John
16-11-2010, 12:43 AM
Sorry Rich I thought it was 84.9 degrees C :p

twisted32
17-11-2010, 05:34 PM
So you still have not changed your thermostat Lee?
I was considering a 77deg or 72deg thermostat when I do the alloy radiator as my temps stabalise at around 93deg (new, stock 82deg themostat) in cool conditions, but quite easily pushes upto 99 deg (thermo fans switch point) in traffic.
Maybe 77 deg is the way to go (tridon)

Hotwire
17-11-2010, 10:40 PM
Im still undecided as to go 77 or 82, as with the additional cooling capacity of the Al radiator, if the thermostat is fully open at 90ºc, then really the temps should be pretty stable at this temp. Hitting 99c while in traffic would have to be a result of lack of capacity of the stock radiator, not the opening point of the thermostat. It MAY have added a little heat soak initially, but I would point the finger at heat soak occuring due to lack of size.

Shtiv
18-11-2010, 11:17 AM
Hmmm even with my alloy radiator mine sits around 90 degrees in cool weather, wonder when my thermostat opens.... It usually runs at 90-95degrees and i've seen it over 100 heaps.... Next question, how accurate is the stock sensor, has anyone checked?

and I know that maxxing BDEL D (to 159.4 and taking WGDC D to 80% gives almost 16psi boost from 3-4k dropping off to 13 or so.... I haven't had a chance this week to do much at all though and it's been raining most of the week and my logger involves having a window down a bit so. i'll get there early next week hopefullly. Still holds the 16 too, not just a spike and drop off at constant RPM :)

Shtiv
19-11-2010, 04:10 AM
more info on boost control here:

http://ozvr4.com/forums/showthread.php?10312-Manifold-Pressure-Logging-(-amp-boost-control)&p=201237#post201237

merlin
19-11-2010, 05:30 AM
there wont be any fuel enrichment when the engine temp is 70-80*C, it will be running closed loop lambda control.

Hotwire
19-11-2010, 05:52 AM
Removed - answered my on Q I think - PM sent

Shtiv
22-11-2010, 07:24 AM
Sorry Rich I thought it was 84.9 degrees C :p

bahahaha! but I agree, about 85 degrees would be best....

Shtiv
22-11-2010, 07:24 AM
there wont be any fuel enrichment when the engine temp is 70-80*C, it will be running closed loop lambda control.

what happens in the open loop zones though?

Adam.Findlay
22-11-2010, 11:45 AM
hey so have you guys got your cars running higher stable boost yet? i have removed the restrictor pill from my solonoid and had a wee fiddle with the wgdc but the boost was very unstable ranging from 15psi to 8psi. if someone has got it working holding 15ish psi stably could you post up what duty cycles and gains you have used. if not i will hopefully spend some time getting it right this weekend

Shtiv
23-11-2010, 02:07 AM
Yeah, change the BDEL as well (max is 159.4) and look here:

http://ozvr4.com/forums/showthread.php?10312-Manifold-Pressure-Logging-(-amp-boost-control)&p=201237#post201237

Sylwek
19-12-2010, 01:24 AM
Odd, when I see disassembly of auto/manual ECU it does not refer to any wgdc table in any place by the means of coolant temperature. There is only retire/downgrade of wgdc after the engine temperature exceeds predefined level. I could notice only load what is temperature related there.
And yes, you can actually set the boost however you want just the same way any other Mitsubishi car from these years.

Hotwire
05-01-2011, 11:10 PM
After being rather busy leading up to Christmas finally got back to revisiting this on the way to work this morning. Attached are 2 plots, one of Temp Vs Temp Scaled in Evoscan,

and the other is WGDC Vs Temps and as can be seen, 85 on engine temp, 80 on engine temp scaled is definitely the magic mark for achieving full boost control. As such, I will ONE DAY SOON change my thermostat back to the stock 82c. I still believe 77 will result in temps that are too cool for propper boost control (and who knows what else it will affect) in cooler months.

Nutter_John
05-01-2011, 11:39 PM
Yeah roughly 84 was where we first noticed the uplift in boost on the dyno , took quite a lot of runs to fully confirm properly

Shtiv
06-01-2011, 12:16 PM
Wow your car runs cold Lee.... Mine is rarely below 90 and that's with the aluminium radiator

And yes, definitely less duty cycle when cold

Hotwire
06-01-2011, 12:51 PM
Yeah it does run very cool - that was in 24º ambient temp too - in winter was cooler. Only time I get consistently over 80 (scaled) is when its 34+ and i'm in traffic. Intersting that with the stock 82ºc thermostat you get 90s consistently, maybe a 77º is the best compromise?

lateshow
06-01-2011, 12:54 PM
I've tried searching but where can I get definitions that have those boost control parts (i'm currently using a bleed valve type of thing and it works great but just would like to try this out, not sure if there are any benefits in using the original solenoid where as the ECU doesnt actually care about "predefined boost levels". Geekmapped doesn't have new definitions with boost control (only the ones merlin did ages ago) so help me out? :)

twisted32
07-01-2011, 06:05 AM
Yeah it does run very cool - that was in 24º ambient temp too - in winter was cooler. Only time I get consistently over 80 (scaled) is when its 34+ and i'm in traffic. Intersting that with the stock 82ºc thermostat you get 90s consistently, maybe a 77º is the best compromise?

With the alloy radiator, and new, stock 82 deg thermostat, coolant temp scaled seems to like sitting around 85deg, until you are sitting in traffic etc. I was thinking I would go for a 77deg thermostat, but I am sticking with the 82 for the moment. At least until we can adjust thermo's to come on earlier than 90odd degrees (temp scaled).

Hotwire
07-01-2011, 06:26 AM
Rob can you post up the temps at which the fans turn on/off again?

twisted32
09-01-2011, 11:29 AM
On at 91, off 89/88deg using coolant temp scaled. Well that is the stationary temps I was seeing. Steve gave me some thermo fan control tables from the Magna's, and they were speed dependent as well.

lateshow
10-01-2011, 04:54 PM
I got some addresses for BDEL and WDGC but now I'm wondering if they are correct at all. BDEL A B C D are different, all of them however haver 0 for last 2 values 7000rpm-> What should the tables look like. D goes highest and the biggest value is 119 versus evo6 has 159 in every BDEL value. WDGC seems quite good, it has all the values bigger then zero and the first values seem to be around 70 % and then it gets a little smaller. WDGC A and B are the same. Is this correct? I have already seen that different ROMs have different addresses, mine is EM2428/2381003

Shtiv
12-01-2011, 07:41 PM
I'll have a look on an EM2428 ROM for you, they could be different on them, the scales are all moved usually

Hotwire
12-01-2011, 10:13 PM
I have a copy of the same rom - I've PMd you the addresses you are after ;)

lateshow
13-01-2011, 11:35 AM
Thanks. Those RPM addresses seem correct now, BDEL and WGDC addresses itself are now what I'm after. Is there really 4 x WGDC? I was only aware of 4x BDEL and 2x WGDC

lateshow
16-01-2011, 11:36 AM
By the way, does anyone have a PFL automatic. It would be interesting to see what values does PFL automatic have for BDEL and WGDC, if it's supposed to develop only 260 bhp ? The boost levels must generally be same for automatic and manual so C and D BDEL must be temperature related somehow.

Kenneth
17-01-2011, 05:05 AM
The disassembly appears to have laid them out the same as the fuel and ignition tables, 2 low octane and 2 high octane maps.
A and B are full of 0.0, which would be 0 duty cycle on the boost solenoid, the high octane contain the maximum values which makes sense as you would expect that in loss of octane, the ecu would eventually put you down to wastegate pressure only.

If indeed the 2 maps are temperature related, it might just be that way for every dual map in the ECU (fuel, ignition etc). I'll get a final answer at some point when I unravel the assembler.

Hotwire
17-01-2011, 06:15 AM
Kenneth,

I have no doubt there is some temp correction there, but it is not interpolating between C & D depending on Temp. I copied the D map to C to see if there was interpollating, and there is not. I'm still only getting a % of WGDC D at the lower temps my car is running (as posted above). Therefore there is some other scaling going on here. If I can help with any specific logging let me know - would like to unravel this SOB :lol:

Kenneth
17-01-2011, 08:40 AM
Sorry, I meant that I believe it is interpolating between high and low octane rather than temp.
Or to put it another way it only interpolates between A OR B and C OR D.

A and B are low octane maps, C and D are high octane maps. Or perhaps something else other than octane, I haven't worked it out that far to be honest... Assembler was a bit before my time so I am taking a while to get my head around what is actually happening

Hotwire
17-01-2011, 10:31 AM
Makes Sense. What i've seen from logging my car running cool, is that the WGDC is coming in at a % of the D map, depending on coolant temp scaled. That is, at 72ºc, there seems to be about 70% WGDC, 77 about 80%, 80º gets about 90% (roughly) but once you hit 82ºC you get 100% WGDC. The curver of the map follows D constantly (with octane not budging off 100) and no knock, SO if that helps you decipher anything......

I'll post some plots of the data I have logged a little later tonight when I get my laptop fired up.

Hotwire
17-01-2011, 11:14 PM
Here is the plot I was referring too. In particular (as most running was under 2000 rpm, therefore 100% WGDC on both C & D maps) you can see the slow increase in WGDC as temps slowly went up. Looking at the log (data):
Temp WGDC
72ºc = 70%
73ºc = 71%
76ºc = 74%
78ºc = 79.5%
79ºc = 85%
80ºc = 89.5%
81ºc = 95%
82ºc = 100%

40719

lateshow
18-01-2011, 11:14 AM
Has anyone found or is there a boost control OFFSET- kind of function for vr4. Evo's boost is mostly controlled by this (even stock evos have maxxed out WGDC and BDEL and the offset says how much boost is added to the sum)

Hotwire
18-01-2011, 01:26 PM
No but I would appreciate it if anyone else found it and let us know

Hotwire
19-01-2011, 12:01 AM
Copied WGDC D to WGDC B this morning, and still get correction as above. Will try copying to A for the drive home this afternoon, if its the same still, then it's not interpollating but correction via a temp scale that we dont know the address of.

Hotwire
19-01-2011, 10:33 AM
changed all 4 tables to be same (100% to 2.5k then 70%) and exactly same plots as above so temp causes 0 interpolation. Above effect is due to another temp scale not yet discovered

Shtiv
19-01-2011, 11:13 AM
I don't know about there being an offset like the evo's, bear in mind that these don't run boost like they do either but in my testing at maxxed out bdel even at 17psi there was no drop off in boost so I suspect there may not be another offset, and I never needed any more anyway....

I am starting to suspect there is a seperate offset table for dropping boost when cold, just like fuel enrichment.... Would still like to know WTF there are 4 tables though. Anyone got any suggestions how we can test if the A and B addresses are truly lo octane?

MackTheKnife
19-01-2011, 11:00 PM
I'm wondering if the 4 tables are for gears. I'm starting to suspect table c is for gears 1 & 2 and table d is for 3, 4 & 5.

It's just a hunch based on how my car has been behaving.

MackTheKnife
22-01-2011, 08:59 AM
Played with the WGDC & BDEL tables quite a bit in the last few days.

A & B seem to be limp mode settings.

both table D's seem to be normal running and both table C's seem to be where the engine map drops to if the BDEL level on the D table is breached.

Hotwire
22-01-2011, 12:11 PM
Thats very intersting info there Paul - cheers!
That would make sense for the correlation between the two sets of maps

Shtiv
08-02-2011, 10:35 AM
A and B would be Lo Octane as Kenneth says, interesting theory about the C map Paul, I should test that but my 3 port is hooked up again so maybe Lee? then the next question is what about the 1st hi octane fuel and ignition maps, when do they come into play? Surely it's a similar concept that never got used in that stock maps are the same across both hi octane maps

Hotwire
10-02-2011, 11:54 AM
I still have the stock solenoid plumbed in, but still having issues with temps. Eventually I'll change the thermostat (I Have one now) and then i'll check to see if thats true.

MackTheKnife
03-03-2011, 06:52 AM
I still have the stock solenoid plumbed in, but still having issues with temps. Eventually I'll change the thermostat (I Have one now) and then i'll check to see if thats true.

Hi Lee, have you had any chance to test as well? I tested over four days, about 150k's a day, mainly twisty roads and reflashed 2-3 times each 75k leg of the trip to try a few different variations including making C&D the same (boost cut kicks in) and having low C's & high D's to exaggerate things.

I'm 99% sure but would like some confirmation.

Also was thinking A & B might be for low loads, after all if i'm sitting at the lights with the aircon on, I can have up to 30 load, but no WGDC. The ECU surely need to be told somehow to release boot off throttle. My A & B tables were all zeros.

I'll test this theory next week to see if the car will hold boost off throttle at higher revs.

MackTheKnife
09-03-2011, 03:12 AM
Did a brief test (didn't want to bugger things up). A&B don't seem to be low load settings.

Kenneth
09-03-2011, 04:26 AM
I have been doing some code investigation for the boost control tables.
So far I have 13 tables. 2 of which are "switched" and I assume are not used with our configuration, though that is not verified yet.

I'll post the details when I have a better idea of what the tables are and which are used and which are not.

Mitsiman
23-03-2011, 09:34 AM
I have been in discussions with Kenneth and Im sending him my own completed definition file (Minus boost control) which may help to provide a better set of definition files for everyone. Going to work with Kenneth as well with some testing on a car on the dyno to help finalise hte boost mapping. Hopefully to all our benefits.

lateshow
24-03-2011, 10:25 AM
I have been in discussions with Kenneth and Im sending him my own completed definition file (Minus boost control) which may help to provide a better set of definition files for everyone. Going to work with Kenneth as well with some testing on a car on the dyno to help finalise hte boost mapping. Hopefully to all our benefits.

Have you found something new?

Mitsiman
25-03-2011, 12:55 AM
Not yet but the plan is with our ability to put cars on dyno and being experienced OpenEcu tuners and developers, we can aid in fast tracking and reliably sorting out this rom. Its a shame that those who have already done this are not sharing there information but not much we can do about that.

Kenneth
25-03-2011, 05:55 AM
Here are 13 tables which are referenced in the boost control section of the ROM

I have found these by referencing Evo ROMs against the VR-4 ROM. The definition I used for cross reference was given out on geekmapped forums (http://www.geekmapped.com) by Ceddy (http://www.geekmapped.com/forums/member.php?u=347) so thanks to him for getting this underway.

Some of my naming is a bit inconsistent but you will see that the main maps tend to come in groups of 4. I believe these are the same layout (low octane primary, low octane secondary, high octane primary etc) as the fuel and ignition maps.



<table name="Turbo Boost Error Correction" category="Turbo" address="121aa" type="2D" level="1" scaling="DynamicBoost_8bit">
<table name="Boost Error" address="2d25a" type="Y Axis" elements="16" scaling="BoostError_16bit"/>
</table>

<table name="Boost Desired Load Adder" category="Turbo" address="10856" type="1D" level="1" scaling="Load8_16bit"/>

<table name="Max Total Upward WGDC Correction vs TPS" category="Turbo" address="1219c" type="2D" level="1" scaling="DynamicBoost_8bit">
<table name="TPS" address="2d282" type="Y Axis" elements="9" scaling="TPS_Percent_16bit"/>
</table>

<table name="Desired Engine Load A - Primary" category="Turbo" address="1223e" type="2D" level="1" scaling="Load8_8bit">
<table name="Engine Speed" address="2d232" type="Y Axis" elements="17" scaling="RPM4_16bit"/>
</table>

<table name="Desired Engine Load A - Secondary (Stub!)" category="Turbo" address="12296" type="2D" level="1" scaling="Load8_8bit">
<table name="Engine Speed" address="2d232" type="Y Axis" elements="17" scaling="RPM4_16bit"/>
</table>

<table name="Desired Engine Load B - Primary1" category="Turbo" address="12228" type="2D" level="1" scaling="Load8_8bit">
<table name="Engine Speed" address="2d232" type="Y Axis" elements="17" scaling="RPM4_16bit"/>
</table>

<table name="Desired Engine Load B - Primary2" category="Turbo" address="12254" type="2D" level="1" scaling="Load8_8bit">
<table name="Engine Speed" address="2d232" type="Y Axis" elements="17" scaling="RPM4_16bit"/>
</table>

<table name="Desired Engine Load B - Secondary1" category="Turbo" address="1226a" type="2D" level="1" scaling="Load8_8bit">
<table name="Engine Speed" address="2d232" type="Y Axis" elements="17" scaling="RPM4_16bit"/>
</table>

<table name="Desired Engine Load B - Secondary2" category="Turbo" address="12280" type="2D" level="1" scaling="Load8_8bit">
<table name="Engine Speed" address="2d232" type="Y Axis" elements="17" scaling="RPM4_16bit"/>
</table>

<table name="Wastegate Base Duty Cycle 1 - Primary" category="Turbo" address="121c8" type="2D" level="1" scaling="WGDuty_8bit">
<table name="Engine Speed" address="2d232" type="Y Axis" elements="17" scaling="RPM4_16bit"/>
</table>

<table name="Wastegate Base Duty Cycle 1 - Secondary" category="Turbo" address="121de" type="2D" level="1" scaling="WGDuty_8bit">
<table name="Engine Speed" address="2d232" type="Y Axis" elements="17" scaling="RPM4_16bit"/>
</table>

<table name="Wastegate Base Duty Cycle 2 - Primary" category="Turbo" address="121f4" type="2D" level="1" scaling="WGDuty_8bit">
<table name="Engine Speed" address="2d232" type="Y Axis" elements="17" scaling="RPM4_16bit"/>
</table>

<table name="Wastegate Base Duty Cycle 2 - Secondary" category="Turbo" address="1220a" type="2D" level="1" scaling="WGDuty_8bit">
<table name="Engine Speed" address="2d232" type="Y Axis" elements="17" scaling="RPM4_16bit"/>
</table>

lateshow
25-03-2011, 06:06 PM
Great Guys!!!! :)

But you haven't found actual "Offset" value that could say what value is added to original BDEL for desireded load-> desired boost? It must be a bitch to find. Running BDEL maxxed out gives you approximetely 1 bar boost = 200 load which means there's extra 40 laying somewhere? Am I totally wrong?


EDIT: Stupid me you have found it!!!! Absolutely fabulous!!! (Has anyone watched that show :D )

EDIT2: Is that boost adder supposed to be zero as standard????

EDIT3: Got it, it's 40, as I suspected all along (all these months) :D

Hotwire
26-03-2011, 03:35 AM
kenneth, you sir are a champion! Thankyou very much :D

Kenneth
27-03-2011, 02:46 AM
The following is the basic layout of the ECU boost control structure as I can determine. There is some stuff which I haven't quite figured 100% (it figuring out the load error for example) but this is essentially how things are done.

Engine Load Target lookup


Load RPM Scale


Low Octane Value lookup from Desired Engine Load A primary
High Octane Value lookup from Desired Enigne Load A secondary

OR (using a switch in ECU code, possibly permanently off but needs testing)


Low Octane Value lookup from Desired Engine Load B primary1/2
High Octane Value lookup from Desired Enigne Load B secondary1/2

Interpolate (Octane Number, low octane value, high octane value)


Base Duty Cycle lookup


Engine speed scale lookup

Low Octane Value lookup from Wastegate Base Duty Cycle 1 (primary/secondary)
High Octane Value lookup from Wastegate Base Duty Cycle 2 (primary/secondary)

Interpolate (Octane Number, low octane value, high octane value)


Boost Correction

Add load target and load adder

Load Boost error scale
Boost error correction map lookup

TPS scale lookup
Max upward Wastegate Correction vs tps table lookup

Then there is some stuff which haven't quite got to working out... At this stage I just assume it is setting up variables for wastgate duty cycle etc.

Shtiv
28-03-2011, 08:53 AM
Kenneth, you are indeed a champion! (and one that doesn't hide information so doubly so!)

Mitsiman
28-03-2011, 09:02 AM
I will do some testing on these in the next few weeks and hopefully confirm how they are all working together (AS much as one can) and then let everyone know unless someone does it themselves first.

lateshow
28-03-2011, 09:33 AM
Engine Load Target lookup


Load RPM Scale


Low Octane Value lookup from Desired Engine Load A primary
High Octane Value lookup from Desired Enigne Load A secondary

OR (using a switch in ECU code, possibly permanently off but needs testing)


Low Octane Value lookup from Desired Engine Load B primary1/2
High Octane Value lookup from Desired Enigne Load B secondary1/2

Interpolate (Octane Number, low octane value, high octane value)



Those primary load maps are strange, low octane is all 0 and hi octane goes a little all over the place, I think it's mainly those secondary maps it uses?

Kenneth
28-03-2011, 09:40 AM
sorry, was a bit unclear there. I think the top two are un-used and the sets after the OR are. (that is Desired engine load B primary 1/2 and secondary 1/2)

Low octane is 0's. Thats ok because if octane gets to 0, its probably a good idea to drop boost as far as possible.

They should be renamed to


Desired Engine Load Low Octane Primary
Desired Engine Load Low Octane Secondary

Desired Engine Load High Octane Primary
Desired Engine Load High Octane Secondary

phosty
28-03-2011, 06:02 PM
So we are saying that the VR4 actually has real 'Closed Loop Boost Control' and not just simple open loop. Or is it open loop but with a few over-boost protections (depending on temperature, Octane count etc)?

Shtiv
02-04-2011, 11:39 AM
A little bit closed loop it seems.... but closed loop to match load(airflow), not manifold pressure like an EBC would, remains to be seen how well it works though. Like most systems it looks to be open loop initially (initial WGDC setting). Then it reviews where it is, compared to where it should be (desired engine load) and it can only increase the duty cycle above the inital setpoint at 100% throttle (which makes sense). From what I have seen and guessed from what Kenneth has marvellously found, it looks as though the octane number alters the initial WGDC setpoint and the target for engine load (and if I was a Mitsubishi programmer that's probably how I would do it also).

The other point I would make is that there also appears to be a correlation between engine temp and final boost control so i'm not sure where that comes in, I assume that just alters initial setpoints. Using Kenneth's naming, maybe that's the job of the WGDC primary, the WGDC secondary is the one that makes the difference for most of the time (just like the second fuel and ignition maps make the changes for most of the time.

Does that make sense? Kenneth might have a bit more input and hard facts on this.

Oh and Kenneth, not sure if it helps but remember your engine load B secondary 2 is the one that seems to have the effect for normal driving

Kenneth
03-04-2011, 03:05 AM
Ah yes... I had been meaning to finish that one table... I guess now is a good time :P

The ECU checks the current coolant temp against the temperature stored in the 1D table defined below. If the temperature is below, then it goes and uses the 2d Table to work out the correction factor.
Not 100% sure about the scaling of the 2D table, looks about right though.



<table name="WGDC Correction Stop Temp" category="Turbo" address="10858" type="1D" level="1" scaling="Temp"/>

<table name="WGDC Trim v Engine Temp" category="Turbo" address="1229a" type="2D" scaling="Percent128_8bit">
<table name="Engine Temp" address="2d090" type="Y Axis" elements="8" scaling="Temp"/>
</table>


Pretty sure that is it as far as boost control goes though.

Kenneth
03-04-2011, 03:10 AM
I renamed all my tables as follows to make them more correct. The top 2 I still believe are not used.

This way it also makes sense that the base WGDC High Octane 2 is being used during general driving. Not sure where it decides which map to use, but ill no doubt get there eventually.



Desired Engine Load A - Primary
Desired Engine Load A - Secondary (Stub!)

Desired Engine Load Low Octane - 1
Desired Engine Load Low Octane - 2
Desired Engine Load High Octane - 1
Desired Engine Load High Octane - 2

Wastegate Base Duty Cycle Low Octane 1
Wastegate Base Duty Cycle Low Octane 2
Wastegate Base Duty Cycle High Octane 1
Wastegate Base Duty Cycle High Octane 2

Hotwire
04-04-2011, 04:59 AM
Just to help out/clarify. Those of us using a vr4base1.xml, you need to add the following scalings for Kenneths boost control:
I began re-naming to existing scalings already defined, but found it easier just to add the correct scalings from the EVO def file ;)



<scaling name="BoostError_16bit" units="%" toexpr="(x*5/32)-20" frexpr="(x+20)*32/5" format="%.1f" min="-20" max="20" inc="0.5" storagetype="uint16" endian="big" />
<scaling name="DynamicBoost_8bit" units="%" toexpr="(x-128)/2" frexpr="(x*2)+128" format="%.1f" min="-64" max="63.5" inc="0.5" storagetype="uint8" endian="big"/>
<scaling name="Load8_8bit" units="%" toexpr="x*5/8" frexpr="x*8/5" format="%.0f" min="0" max="160" inc="0.625" storagetype="uint8" endian="big"/>
<scaling name="Load8_16bit" units="%" toexpr="x*5/8" frexpr="x*8/5" format="%.0f" min="0" max="500" inc="0.625" storagetype="uint16" endian="big"/>
<scaling name="RPM4_16bit" units="RPM" toexpr="x*1000/256" frexpr="x*256/1000" format="%.0f" min="0" max="11000" inc="3.90625" storagetype="uint16" endian="big"/>
<scaling name="TPS_Percent_16bit" units="%" toexpr="x*100/255" frexpr="x*255/100" format="%.0f" min="0" max="100" inc="0.392157" storagetype="uint16" endian="big"/>
<scaling name="WGDuty_8bit" units="%" toexpr="x/2" frexpr="x*2" format="%.1f" min="0" max="100" inc="0.5" storagetype="uint8" endian="big"/>
<scaling name="Percent128_8bit" units="%" toexpr="x*100/128" frexpr="x*128/100" format="%.0f" min="0" max="200" inc="0.78125" storagetype="uint8" endian="big"/>

Shtiv
06-04-2011, 03:42 PM
Yeah I cleaned up my XML's the other night, took ages.... but I got a bit more consistent and defined everything as 8 or 16 bit in the title, previously if used an 8 bit only, I would just name it and then if i needed a 16 bit later I would then add the same name but with 16 on the end and vice versa, looks better now.

Thanks for the temp stuff too Kenneth. :)

lyndon
07-04-2011, 03:52 AM
is the load adder table the load offset table??? also this maybe off topic but does anybody have a acceleration enrichment table in their rom??

Kenneth
07-04-2011, 06:40 AM
The load adder is not a table, it is a single value which, when added to the lookup done on the Desired Load lookup gives the target load.

Haven't looked for acceleration enrichment yet.

Mitsiman
07-04-2011, 01:39 PM
Kenneth can you email me your modified base VR4 file - I transfered the data into my file but its not looking right, especially if we are all changing the wording or the titles. We really need to all agree to one lot of naming so we can all talk about the right information. If all of us are using different titles it can get confusing.

Shtiv
09-04-2011, 01:14 PM
Yeah I realised how different eveyone is being with their scaling names when i went through mine....

Kenneth
13-04-2011, 07:10 AM
I didn't modify the base file, I just copied and pasted all the scalings from Ceddy's into my file. As such, it is a mess and you don't want it in that state.

I'll get around to cleaning it up at some stage... Unfortunately when I get free time (not often these days) to work on it, I tend to spend it on actual disassembly.

Mitsiman
14-04-2011, 06:55 AM
I have been making some good progress. My programmer, Acamus who I employ is working on the roms for us right now. Its costing me a little money to do this but we have already made some progress on the addresses and within another week we should have for the major VR4 roms, all the main addresses and coding sorted out.

We wil need to do some testing on some vehicles with different roms when I get the coding, and to avoid confusion I will upload my updated Definition Files here so we can all talk using the same referencing of names etc.

Shtiv
15-04-2011, 12:55 PM
I have been trying to get to it but just no time. My car is back on the dyno tomorrow testing a few things (not so much ecu related except for dwell time stuff) and they are taking up more time. But happy to help with testing of items if it helps anyone.

Davezj
26-11-2011, 02:15 PM
Has anyone got a definition file and a Rom file with this boost control in it already, so I can have a Look at it.
It just easier to do then typing it all in myself where I can make mistakes.

foxdie
26-11-2011, 06:32 PM
I think these are already part of most definitions these days? You can set the target load (which almost directly translates to boost) and set the Wastegate Duty Cycle anywhere between 0-100%.

Davezj
26-11-2011, 09:11 PM
I kind of thought that, but there are so many threads on here with which concern the ecu definition files and roms it is difficult to know what is what. Especially with regard to naming conventions, I don't know if this has been settled upon.
I don't want to look at something that has already been done.
So if you can post up the latest definition file that would be great.
I am not after a tuned rom, just one that has all the parameters in it that can be tuned.
As I have not updated mine since may this year, I am way behind.

Davezj
27-01-2012, 12:33 AM
Are there any files out there on this yet, where the naming convention has been settled on.
I am slowly getting my head round this stuff.

Shtiv
27-01-2012, 10:27 AM
Bill and I are working on it.... Very soon I should have my end of the deal done :)

Davezj
27-01-2012, 07:52 PM
Thank for the update, we are all very thankful that you and your mates are putting in this effort.

Shtiv
01-02-2012, 11:30 AM
We're just all slack bastards with too much other stuff on.... Sorry for the delay but we'll get there soon. If you want a current def desperately for boost control email me which ecu you want it for and I'll send one to you. (transformancesteve@gmail.com)

Shtiv
13-02-2012, 02:40 PM
OK we're back on it, I think we're almost there, I'm checking a heap of idle stuff that we had before plus a few new bits Bill has found then we'll post it up.

lathiat
13-02-2012, 03:17 PM
Sounds good. :-)

BCX
16-02-2012, 03:17 PM
As steve said, should be able to post update soon to the VR4 defs soon, once i clarify a few things with steve and we get some stuff tested.

i'll hopefully have some goodies finalised for this coming release or the next one after.

Hotwire
17-02-2012, 12:33 AM
Bill
Do you need another tester to help things along? Happy to assist you and steve

BCX
17-02-2012, 04:38 AM
Lee,
Grab a copy of the latest defs and start testing anything that has (untested) appeneded to the end of the table name.

If you give us some feedback, we'll remove the untested flag for the table in future releases.

Steve was going to volunteer to check out the Idle tables... i cant test anything at the moment as i dont have my openport. i've got a bunch of mods i want to try like increase Baud rate, outputs on conditions, and some more R&D on the live tuning front.

Perhaps one day could meet up and check a bunch of stuff?

Shtiv
18-02-2012, 10:59 PM
A bit closer.... I did a heap of idle testing but a lot was inconclusive, some tables make minor changes but we need more addresses that most likely make the bigger changes, it's hard to see exactly what is changing when you can't see what else is going on in the background....