Kenneth's ECU Mods V2.0

[SAF68]

Which value do you want?

Saf

[foxdie]

Speaking of values, can someone outline the functions behind Coil Settings 1, 2 & 3? I need to increase dwell to hopefully resolve an intermittent spark blowout issue, a couple of the above tables seem to be time vs voltage with values in the regions of what I am looking for, but as there is no definition behind each table I'm not willing to risk it without a second opinion

[SAF68]

Hello Foxdie , Go with HKS dli , if you've missfire !

The Twin Power Ignition Amplifier is an easy-to-install ignition upgrade that continuously monitors the charge condition of the coil to allow the system to recharge to its maximum capacity without overcharging. The Twin Power provides optimum spark duration and voltage output resulting in crisper throttle response & more horsepower. it is a combination of CDI (Capacitive Discharge Ignition) and Transistor ignition in one package. The CDI segment is effective a producing high voltage, but only for a short time period, like most performance ignitions on the market. Therefore, the Transistor segment controls current to the coil (impedance matching circuit) & promotes a longer spark duration. The combination of the CDI & Transistor system creates higher RPM horsepower & lower RPM throttle response. This system amplifies the stock ignition, but does not replace your existing system.


Saf

[Kenneth]

All of the settings in various tables for running E85 would be nice, if you don't mind sharing them.

Which value do you want?

Saf

[BCX]

Speaking of values, can someone outline the functions behind Coil Settings 1, 2 & 3? I need to increase dwell to hopefully resolve an intermittent spark blowout issue, a couple of the above tables seem to be time vs voltage with values in the regions of what I am looking for, but as there is no definition behind each table I'm not willing to risk it without a second opinion

I can look at the disassmbly and let you know exactly what each table is doing... if you want to increase dwell, i'd play with coil table 2 from memory.

Also note - Shtiv's played with these tables, there is a hidden adder of about 1ms or something like that from what the table show you. It's on my to-do list to workout whats actually going on here (and how the three tables are related), but he scoped it and the timings are 1ms out. he might be able to elaborate on this more.

[Kenneth]

Again, thanks to Merlin for the Evo definitions which gave the names and scalings.

I still can't find any adder, though I haven't spent a LOT of time on it.


<scaling name="BatteryVoltage" units="Volts" toexpr="x*75/1024" frexpr="x*1024/75" format="%.2f" min="0" max="18.75" inc="0.5" storagetype="uint16" endian="big"/>
<scaling name="ChargeCurrent" units="Alternator Max Charge Current - Ampere" toexpr="x*80/256" frexpr="x*256/80" format="%.1f" min="0" max="255" inc="1" storagetype="uint8" endian="big"/>
<scaling name="ChargeTime" units="mS" toexpr="x*0.064" frexpr="x/0.064" format="%.2f" min="0" max="255" inc="1" storagetype="uint8" endian="big"/>
<scaling name="ChargeTimeOffset" units="mS" toexpr="x/256" frexpr="x*256" format="%.3f" min="0" max="255" inc="1" storagetype="uint8" endian="big"/>



<table name="Ignition Coil - Dwell Time" category="Ignition Parameters" address="11fba" type="2D" scaling="ChargeTime">
<table name="Battery" address="2cec4" type="Y Axis" elements="9" scaling="BatteryVoltage"/>
</table>

<table name="Ignition Coil - Dwell Correction RPM" category="Ignition Parameters" address="10524" type="1D" scaling="RPMLimit_16bit"/>

<table name="Ignition Coil - Dwell Time Correction 2" category="Ignition Parameters" address="11fd6" type="2D" scaling="Percent128_8bit">
<table name="Battery" address="2cec4" type="Y Axis" elements="9" scaling="BatteryVoltage"/>
</table>

<table name="Ignition Coil - Dwell Time Correction 1" category="Ignition Parameters" address="11fc8" type="2D" scaling="ChargeTimeOffset">
<table name="Battery" address="2cec4" type="Y Axis" elements="9" scaling="BatteryVoltage"/>
</table>

[SAF68]

All of the settings in various tables for running E85 would be nice, if you don't mind sharing them.

For sure , but now , i'm try to obtain a solid LTFT & STFT ... NEED times

Saf

[foxdie]

I actually started looking into this last night, doing a lot of reading and finally settling on the following as being the best possible definitions for the tables (these were published by Merlin literally 3 months ago to the day) and is the result of a lot of work by some smart guys on evolutionm.net.

I can't guarantee these are instantly compatible with our cars, but should serve as a great reference point to update the defs, usage info is after the following XML;

<!-- SCALINGS -->

<scaling name="ChargeTime" units="mS" toexpr="x*0.064" frexpr="x/0.064" format="%.2f" min="0" max="16.32" inc="0.064" storagetype="uint8" endian="big"/>

<scaling name="BatteryVoltage" units="Volts" toexpr="x*75/1024" frexpr="x*1024/75" format="%.2f" min="0" max="18.75" inc="0.5" storagetype="uint16" endian="big"/>

<scaling name="CF256" units="Correction Factor" toexpr="x/256" frexpr="x*256" format="%.3f" min="0" max="2" inc="0.0039" storagetype="uint8" endian="big"/>

<scaling name="Percent128" units="%" toexpr="x/1.28" frexpr="x*1.28" format="%.0f" min="0" max="100" inc="1" storagetype="uint8" endian="big"/>

<scaling name="BattVoltDwell%" units="%" toexpr="x*75/1024" frexpr="x*1024/75" format="%.2f" min="0" max="18.75" inc="0.5" storagetype="uint16" endian="big"/>

<scaling name="RPMLimit" units="RPM" toexpr="7500000/x" frexpr="7500000/x" format="%.0f" min="0" max="9000" inc="60" storagetype="uint16" endian="big"/>


<!-- Coil settings for Evo6, ROM ID 25570000 -->

<table name="Ignition Coil, Dwell-Time" address="1229e" category="Timing" type="2D" scaling="ChargeTime">
<table name="Battery" address="2c7c8" type="Y Axis" elements="9" scaling="BatteryVoltage"/>
</table>

<table name="Ignition Coil, Dwell-Time Reduction RPM" category="Timing" address="10A46" type="1D" scaling="RPMLimit"/>

<table name="Ignition Coil, Dwell-Time Reduction Factor" address="122ac" category="Timing" type="2D" scaling="CF256">
<table name="Battery" address="2c7c8" type="Y Axis" elements="9" scaling="BatteryVoltage"/>
</table>

<table name="Ignition Coil, Dwell-Time Correction 2" address="122ba" category="Timing" type="2D" scaling="Percent128">
<table name="Battery" address="2c7c8" type="Y Axis" elements="9" scaling="BattVoltDwell%"/>
</table>

And to summarise use as I now understand it;

Ignition Coil, Dwell-Time is a baseline for what coil dwell (charge duration) should be, this is used up until 2800 RPM without any adjustments. Above 2800 RPM the ECU starts to apply a reduction factor (because the theory being the coils will now have less time to charge), this is controlled via Ignition Coil, Dwell-Time Correction 2 and Ignition Coil, Dwell-Time Reduction Factor. The faster the engine spins, the more the reduction factor is applied.

So to summarise, Ignition Coil, Dwell-Time will be the maximum dwell time across the board, with the other two tables controlling how quickly dwell drops off. It's a little more involving than that, but those are the basics. How this translates to our current definitions is speculative based on what I've seen on the other forums..

coilsettings.png

Coil Settings 1 = Ignition Coil, Dwell-Time Reduction Factor
Coil Settings 2 = Ignition Coil, Dwell-Time
Coil Settings 3 = Ignition Coil, Dwell-Time Correction 2

As I say, that is speculative and there's a good chance it could possibly be wrong. I'm gonna plot the actual formulas in a graph and post the results to see if it all tallies up.

[foxdie]

So after further investigation, I'm now utterly confused. I could do with some disassembly advice from @BCX or @Kenneth to complete the following;

coildwell.png

As you can see, it tails off way too quickly. The formula I used is the one Merlin posted on EvolutionM.net, Correction 2 (ms) doesn't get used for example. It's all a bit hairy-fairy and I'd like to nail this down

[BCX]

From an disassembly perspective, the values in these tables are loaded directly into the timer used to drive the coils.

When i get a chance, i'll have a look at how the timer is configured and work backwards from that.

[Kenneth]

Jason, the definition I posted a few posts back is exactly those 3 tables, you can cross reference the table addresses in the original VR-4 definition to find out which were which.

I did this by disassembling the Evo5/6 ROM, finding the code which deals with the coil timings (fairly close to the VR-4 code actually) and renaming/scaling the VR-4 ones as per the Evo ones.

If the evo ones are correct, so are the ones I posted. (Which were from Merlin)

[BCX]

Had a quick look at the disassembly, prescaler for timer is set to divide by 64 (CPU clock is the reference)

H8 is clocked at 16mhz, means each time the counter decrements, time is 4uS (microseconds). Values in table(s) are loaded into timer, then timer counts back to 0.

For example, if value is 5 then 5 x 4 = 20uS coil is set to dwell.

while the timer is counting, relevant pin in held high. timer has 4 i/o, one is used as interupt from a sensor (cant remember what sensor, have to check my notes, it's not CAS), other three pins drive each coil.

I've run out of time atm, should be doing work but i'll post up how the coil timings are calculated and what the three tables do from a disassembly perspective.

[Kenneth]

Have updated the ROM again and should have fixed the WGDC map 5.

[Kenneth]

may have been a bit ahead of myself... will do some checking first

[Kenneth]

False alarm, the scaling has been updated previously to show the - symbol when timing is after TDC

[Davezj]

this is for you guys that want to use kenneth's map switching for running ethanol based fuel and petrol based fuel without having to empty the tank before you fill up with the other fuel.
appartently it is done with an external sensor that monitors the percentage of ethanol to pertol and this is feed into the ecu which interpolates between the petrol and ethanol fuel and ignition maps
to give you the correct configuration depending on what mix of fuel you have in the tank.
well that is how i read it.

this is from the evolutionM.net web site and it seems to be quite a good write up, i thought the bones of it could be used and ported accross and used rather than building from scratch.
http://forums.evolutionm.net/ecuflas...n-evo-ecu.html

hope this helps someone.

[Kenneth]

If you want to go with flexi-fuel, you should also look at this thread: http://www.clubvr4.com/forum/showthr...ight=flexifuel

[Davezj]

sorry post in wrong thread.

[rotorB]

Hi everyone, what address for Minimum IPW is in that ROM?

[rotorB]

116af address for "Injector Latency v Battery Voltage" in def file is wrong, the correct address is 115a2.