ROFLMAO @ David :D

Boost limit isn't based on _throttle_. Remember the boost values are _absolute_ boost pressure (hint!). What should they be relative to (big hint!).

I'll help more when it's up on a web site to read ! I'll hopefully learn something too ;-)

Cheers

Steve

Hi Pavlo,

Just interested to know your strategy:
Are you watching address access on a ROMulator or disassembling the code ?

Steve

Stephen,

The boost limit string on an ecu code I have with me is this:

96 9C A2 A8 AE B4 BA C3

Now these correspond to the following absolute pressures in mmHg.

1200 1248 1296 1344 1392 1440 1488 1560

Now, 1560 is the limit, but what about the other 7 values? Are they sitting there for fun? I don't think so.

I am not sure about this (the limit), but currently we have had good results by raising all 8 bytes by the same amount. But I have had reports of people using FCD and mechanical means to raise boost, having fuel cuts that not even PE could explain (at the time, maybe they know now). Now I can well believe that a dawes/actuator adjustment could lead to high boost at part throttle, and using a boost limit at all levels of throttle could possibly lead to the syptoms experienced.

So far, I have identified some maps from web pages, some by just looking at the code for a long time as a greyscale image, some by snippets of info from people round here (like you, remember?) and some by dissasembing bits of code.

I also got my PC hooked up to the select monitor port and ripped the standard ecu binary from my car.

Eprom emulator would be nice, but until recently, a little unjustified.

I am not too keen to put all the info on here just yet, and I don't want someone to use it, and kill their engine because I got an address a line out.

Paul

>Now, 1560 is the limit, but what about the other 7 values?
>
They are all boost limits - notice the progression. The boost value is looked up from a range of options.

>Are they sitting there for fun? I don't think so.
>
Could the absolute boost pressure limit be based on atmospheric pressure perhaps ? If you are half way up a mountain, you do not want to have the boost limit at the same absolute pressure value, since if the absolute value were 2.5 bar, and atmospheric was only 0.5 bar, this would be a MRP limit of 2 bar instead of the intended 1.5bar.

Instead, you probably want the limit to be relative to atmospheric pressure.

Cheers

Steve

Wow.....

Thought it was about time I visited.... Didn't know you'd gone public yet Paul ?

For everyone else... I am currently running 2 versions of Code on the said adaptors.

Version 1 'Stage 1' - raises the Boost to 14psi, has different Fueling and timing when compared to the standard ECU. This is a direct copy of an origional 1028 device, split ODD/EVEN for the 27C512's used on the adaptor board.
This was found to work well and happily no difference was found between running the adaptor and the single 1028 device.

Expected Power for a UK Subaru, with 'Stage 1' chip - 258BHP. Chris.Palmer ran with the same 1028 device last week at PE.

Version 2 'Stage 2' - raises the Boost to 15.5psi and the Fuel Cut to 17.5psi. Solenoid Duty cycle has also been adjusted to help spool up. This version is still under development as I want to try adjusting the Duty Cycle a bit more.

Expected Power with 'Stage 2' chip - 265BHP.

Both of the above still provide good fueling as monitored on my AFR.

Version 2 'Stage 2' will be tested at Tuning Japanese on 31st August.

Differences between UK and WRX code are currently being examined, but at present Boost/Fuel/Ignition etc.... all look identical on the WRX95 and UK95 maps we are currently evaluating.

We'll keep you posted......

.

[Edited by Pavlo - 10/28/2002 10:57:16 PM]

Doesn't the brown solenoid next to the Boost Contol Solenoid switch between Atmospheric and Boost Presures.
Therefore the ECU is contantly monitoring the Atmospheric pressure to calculate the 'Actual' Boost produced.

dunno what the ecu chooses to do internally, but it checks atm press on ignition turn on, and I think it does it periodically when it detects 0% throttle.

I have a suspicion it's not as clever as we think in terms of what it decides to do with the atm pressure reading.

Paul

Can I ask a slightly off-topic Q?

Is anyone still developing the in-car ECU display?

Stefan

Ozzy, I am doing something like delta dash for older cars, there are a few doing things with lcd displays too.

Right then,

bonus peanut to anyone that can identify this map I found, i suspect it's something like a temp compensation map, but who knows. 0-72'C top scale perhaps, fuel percentage adjustment map and the other scale I dunno.

http://www.cdtech.co.uk/anonymous_map.gif

http://www.cdtech.co.uk/anonymous_map_plot.gif

Paul

Paul - whats the address of this map - someone may know just from that?

YHM, perhaps someone will help.....

ttt

ttt

hmm, thanks for the bump, should get my finger out really.

Been busy building a Wideband lambda sensor kit, which seems good so far, put some stuff up when finished.

Managed to download my Standard MY94 UK map yesterday, after re-compiling the code into Labview 5.01 and using a more reliable laptop.

UK94 MAP has identicle Boost/Fuel/Ignition/Duty Cyle etc...etc... to the UK95 MAP supplied by Chris.

There are slight variations (the odd byte) on lines 1,2 & 64 from what I can make out. But I think these may just be 'type' numbers i.e specifies Model year etc...etc...

Same applies to the WRX 95 code we aquired too, which I am still unsure whether it is actual WRX code or not. Chris may be able help here...

I should be able to get a 93WRX map in the next week or 2.

But in the meantime, I going to go away and play with my Duty Cycle on my Stage 2 Development Version....

cool indeed

I am glad you got the ecu dump working!

paul

If the 93 WRX you have planned falls through, I'm happy to help with mine.

Hi again Pavlo!!

Can you do the ECU dump with the ECu Disconnected from the car? if so you can have mine for a week or so as i am off on holiday next weekend, so i wont need it :-)

peggy,

I will mail you with more details tomorrow (monday), gettin late.

paul

Scott / Paul

I know you're trying to do the older ecu's, but would a map from an MY97 be of any help?
would at-least be intersting to see if the maps are the same/how different they are!

Steve

it woudld.... watch this space for 97-98.

David

David,

I will ! :) and will try to dump my ECU at next time it stops raining! :(

Would try to find the maps myself - but suspect it's probably easier if Paul / Chris etc take a look / push me in the right address range as they've already found them in the 96 ecus!

Steve

ok guys - so what happens if i try to WRITE to my MY97 ecu ? (using select protocol)
it's not flashable, like the 99-> ones, so will it just change the working value & then revert to normal after ignition on/off or ecu reset ?

Also, has anyone got the ID function of the subaru protocol to work?
It just does nothing on mine ! :(

cheers
Steve

you can write to RAM, which I think is anything in the first 16kb, maybe 32kb. So you can write to areas to put in error codes, remove error codes etc. I don't know how these are stored as yet.

You can write to bits for control switches, but it doesn't do anything as it gets overwritten.

paul

Stick it in a loop, so that is continually writes to the FAN address, then you should be able to see it work.
Once you stop writting tho' it'll probably get over-written.

Rom emulator is next on our list, right after the select monitor tool, which I will have a proper go at, honest!

I have identified a suitable emulator for about £350inc, but I was looking again, and thinking a programmable one (so custom software can be written) might be better, but is about £450 upwards.

However, spending lots on this kit probably isn't so wise, as I will eventually sort out the 99-00 ecus, or just buy the ecutek mapping product if things get serious enough (trade serious I mean).

One step at a time though.

Paul

Scott - ok - might try that one .. if I'm brave enough!

Paul - i've had a scower (??) on the net and come up with quite a few 'home-brew' jobs, that should do nicely .. need to look a bit more into them to see if they will do the job, but the couple I have read most on will do even/odd on most of the 27c family (if that's what's used?)

'Select monitor thingie' - well I've got that almost 100% done. Its not that wonderful to look at, but it shows you

RPM, Speed, Boost, Batt Volts, throttle Pos, water temp

all on one screen. (could add more ...)

and then on another screen you can select (no pun inteded!) to monitor an individual parameter on it's own. and includes all the ones that are avalible! ( inj ms, boost, O2 (might put this on the front page , coz it's one we should keep an eye on !) etc etc)

One problem I can see is that the multi funciton bit wont work on the pre97 ecu, as from what you guys are saying, you cant get multiple addresses back from the ecu, only one at a time.

Will be bringing all the equip with me on the 31st if anyone wants to have a look / play !??

cheers
Steve

scoobyduck, could you email me details for the emulator... as the only one i found was serial port based, and didnt support live updates without a reset, and no odd + even split.

David