Meth Failsafe and Patches
Oops, I read your post wrong Barry. I thought you were confirming that pin 23 was an alternate 5V source for sensors. My Bad. Yeah, I wouldn't want to use pin 42 for fear of it altering the voltage level to the sensors.
Originally Posted by mrfred;
Richard, Pin 81 is 5 V power for sensors on USDM Evo 8s. Is it something different on the EDM Evo8s?
Since I'm pretty excited about the future of map interpolation I started messing with things we can log if we hook up a flow sensor to the ADC input now. You can log the flow level (in cc's or m/l's) and you can log the percentage of your fuel mixture which is from the injection system.
I used Richards (i.e; Aquamists) DDS-3's scaling of 1200cc's@5v in my example. To log this, all you need to do is change your MUT 83 code to FFFF0x2929. This way it logs the ADC0F input MrFred was oh so nice to find for us,lol.
This is gonna be so cool when we can interpolate mapping with this 
One thing you'll notice with some of the crappy injection systems is they either flow their maximum, or they are off. No expodential increase as the IDC's go up. When you log these systems you'll notice the percentage of alcohol is greatest when the system activates and leans out as the rpm's go up. There is no consistant mixture what-so-ever. This is why the 'good' systems with IDC inputs kick their butts from dyno pull to dyno pull.
I used Richards (i.e; Aquamists) DDS-3's scaling of 1200cc's@5v in my example. To log this, all you need to do is change your MUT 83 code to FFFF0x2929. This way it logs the ADC0F input MrFred was oh so nice to find for us,lol.
Code:
<DataListItem DataLog="N" Color="" Display="Alcohol Injection Flow" LogReference="InjectionFlow" RequestID="83" Eval="(X*0.0196078)*(1200/5)" Unit="cc" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="1200" ChartMin="0" ChartMax="1200" ScalingFactor=".1" Notes="Replace maximum flow rating in the formula to the highest reading your flow sensor can monitor at 5v. Example: (X*0.0196078)*(ENTER MAX CC FLOW RATING HERE/5). Change gauge and chart scaling to accurately display your flow range." Priority="1" Visible="False" /> <DataListItem DataLog="N" Color="" Display="Alcohol Injection Mixture %" LogReference="InjectionMixture" RequestID="00" Eval="([InjectionFlow]/([injectionFlow]+(513*4*([InjDutyCycle]*.10))))*100" Unit="%" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="200" ChartMin="0" ChartMax="200" ScalingFactor="1" Notes="If you have changed your Injector Size Scaling to anything other than the stock injector scaling you must edit your formula with your new injector scaling size. Example of where to alter the formula: (ENTER INJECTOR SCALING SIZE HERE*4*([InjDutyCycle]*.10)" Priority="1" Visible="False" />
One thing you'll notice with some of the crappy injection systems is they either flow their maximum, or they are off. No expodential increase as the IDC's go up. When you log these systems you'll notice the percentage of alcohol is greatest when the system activates and leans out as the rpm's go up. There is no consistant mixture what-so-ever. This is why the 'good' systems with IDC inputs kick their butts from dyno pull to dyno pull.
Last edited by Jack_of_Trades; Jan 4, 2008 at 01:50 PM.
Also you can get a pin splicer terminal where a very thin pin is pushed through the insulation. When you finish, the wire insulation heals itself. You need to google it.
He has a setting for the option of a time delay (in milliseconds) and also an option for a load limit in which the maps won't switch if you exceed the desired load level.
I personally would like to see a setup where in the instance that the alternate map is switched during high load, the ecu won't allow the aggressive map to switch back on until it sees the car go below a set load value. I wouldn't want to be at a load of 240 and have my car switching back and forth between an aggressive and a conservative map while under such harsh conditions.
I personally would like to see a setup where in the instance that the alternate map is switched during high load, the ecu won't allow the aggressive map to switch back on until it sees the car go below a set load value. I wouldn't want to be at a load of 240 and have my car switching back and forth between an aggressive and a conservative map while under such harsh conditions.
I can't seem to get evoscan to let me use the logreference tags in the 'injection mixture' eval formula. Can anyone see anything wrong? The 'alcohol injection flow' formula works just fine. Its totally stumping me as to what is wrong here.
this works fine
this one isn't registering the logreferences
This is gonna be so cool when we can interpolate mapping with this One thing you'll notice with some of the crappy injection systems is they either flow their maximum, or they are off. No expodential increase as the IDC's go up. When you log these systems you'll notice the percentage of alcohol is greatest when the system activates and leans out as the rpm's go up. There is no consistant mixture what-so-ever. This is why the 'good' systems with IDC inputs kick their butts from dyno pull to dyno pull.[/QUOTE]
this works fine
Code:
<DataListItem DataLog="N" Color="" Display="Alcohol Injection Flow" LogReference="InjectionFlow" RequestID="83" Eval="(X*0.0196078)*(1200/5)" Unit="cc" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="1200" ChartMin="0" ChartMax="1200" ScalingFactor=".1" Notes="Replace maximum flow rating in the formula to the highest reading your flow sensor can monitor at 5v. Example: (X*0.0196078)*(ENTER MAX CC FLOW RATING HERE/5). Change gauge and chart scaling to accurately display your flow range." Priority="1" Visible="False" />
this one isn't registering the logreferences
Code:
<DataListItem DataLog="N" Color="" Display="Alcohol Injection Mixture %" LogReference="InjectionMixture" RequestID="00" Eval="([InjectionFlow]/([InjectionFlow]+(513*4*([InjDutyCycle]*0.10))))*100" Unit="%" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="200" ChartMin="0" ChartMax="200" ScalingFactor="1" Notes="If you have changed your Injector Size Scaling to anything other than the stock injector scaling you must edit your formula with your new injector scaling size. Example of where to alter the formula: (ENTER INJECTOR SCALING SIZE HERE*4*([InjDutyCycle]*0.10)" Priority="1" Visible="False" />
One thing you'll notice with some of the crappy injection systems is they either flow their maximum, or they are off. No expodential increase as the IDC's go up. When you log these systems you'll notice the percentage of alcohol is greatest when the system activates and leans out as the rpm's go up. There is no consistant mixture what-so-ever. This is why the 'good' systems with IDC inputs kick their butts from dyno pull to dyno pull.[/QUOTE]
Thread Starter
Joined: Feb 2007
Posts: 9,486
Likes: 67
From: Melbourne, Australia
Just a quick update - I have found the problem with Jamies ROM, basically the ADC0F registers a tapering of the voltage, the code doesn't like this.
Anywho I have recoded it and Jamie is testing now, MrFred I will update your ROM once Jamie says its working! You can ignore the ROM I sent you before!
Just to confirm the failsafe/ADC method of switching maps is INSTANT and currently has no load thresholds - ie it will swap at any load. This may be changed in the future.
Anywho I have recoded it and Jamie is testing now, MrFred I will update your ROM once Jamie says its working! You can ignore the ROM I sent you before!
Just to confirm the failsafe/ADC method of switching maps is INSTANT and currently has no load thresholds - ie it will swap at any load. This may be changed in the future.
EvoM Guru
iTrader: (50)
Joined: Mar 2006
Posts: 9,675
Likes: 132
From: Tri-Cities, WA // Portland, OR
Just a quick update - I have found the problem with Jamies ROM, basically the ADC0F registers a tapering of the voltage, the code doesn't like this.
Anywho I have recoded it and Jamie is testing now, MrFred I will update your ROM once Jamie says its working! You can ignore the ROM I sent you before!
Just to confirm the failsafe/ADC method of switching maps is INSTANT and currently has no load thresholds - ie it will swap at any load. This may be changed in the future.
Anywho I have recoded it and Jamie is testing now, MrFred I will update your ROM once Jamie says its working! You can ignore the ROM I sent you before!
Just to confirm the failsafe/ADC method of switching maps is INSTANT and currently has no load thresholds - ie it will swap at any load. This may be changed in the future.
Thread Starter
Joined: Feb 2007
Posts: 9,486
Likes: 67
From: Melbourne, Australia
could be anything - maybe there is a cap or something in the ECU... either way I have changed the logic:
ADC0F > 4.7v = altmaps
anything else = standard maps
also it will only swap to standard maps if it was on altmaps by way of the ADC > 4.7v...
ADC0F > 4.7v = altmaps
anything else = standard maps
also it will only swap to standard maps if it was on altmaps by way of the ADC > 4.7v...
THe ADC input holds the voltage charge for an instant before it fully discharges. Adding a 10k ohm resistor between ground and the ADC input pin would help speed up the discharge but I don't know how much it'd help though.
[QUOTE=Jack_of_Trades;5119205]Since I'm pretty excited about the future of map interpolation I started messing with things we can log if we hook up a flow sensor to the ADC input now. You can log the flow level (in cc's or m/l's) and you can log the percentage of your fuel mixture which is from the injection system.
I used Richards (i.e; Aquamists) DDS-3's scaling of 1200cc's@5v in my example. To log this, all you need to do is change your MUT 83 code to FFFF0x2929. This way it logs the ADC0F input MrFred was oh so nice to find for us,lol.
[QUOTE]
The flow sensor has a active voltage window between ~0.53V to 4.6V. The voltage outside the area is reserved for sensor check. The internal micro generates 0.53V on stand-by and if the sensor fails, the output will be >4.6V.
You need some interfacing circuitry to null out the offsets if you want the output span to swing between 0-5V.
I used Richards (i.e; Aquamists) DDS-3's scaling of 1200cc's@5v in my example. To log this, all you need to do is change your MUT 83 code to FFFF0x2929. This way it logs the ADC0F input MrFred was oh so nice to find for us,lol.
[QUOTE]
The flow sensor has a active voltage window between ~0.53V to 4.6V. The voltage outside the area is reserved for sensor check. The internal micro generates 0.53V on stand-by and if the sensor fails, the output will be >4.6V.
You need some interfacing circuitry to null out the offsets if you want the output span to swing between 0-5V.
How easy is it to alter the threshold? 4.7V is quite close to a commercial voltage regulator output tolerance of 5%. 4.75V to 5.25V. If the threshold is set to <4.7V=normal map, it doesn't take much "line noise" or a "loaded 5V rail" to cancel the alt map.
If the threshold be can be set at >4V = altmap, it would be more reliable. A hyteresis of ~1V will also help to null out some of the voltage decay after 5V is disconnected. Jamie's suggestion of loading the ADC input of 10K will speed up the decay.
With hysteresis of 1V:
>4V = "altmap" on ramp up (0 ---> 5V)
<3V = "normmap" on ramp dowm (5V ---> 0V)
This will ensure the altmap stays firmly in place between 3-5V after triggering instead of 4-5V.
Yeah, I had asked if it was possible to have these trigger points be adjustable in ECUFlash but I wasn't sure if that would be a coding nightmare or not. I'm still not sure if the ICS switch is supposed to be hooked up to do manual map switching still with the ADC coding he added.
Best thing to do would be to use a stripped down .bin and just add the ADC map switching until we get that to work I think. Then again, I'm not the coding genius here (butt kissing ) We're SO close Dave! YOU CAN DO IT!!!! 
Best thing to do would be to use a stripped down .bin and just add the ADC map switching until we get that to work I think. Then again, I'm not the coding genius here (butt kissing
) We're SO close Dave! YOU CAN DO IT!!!!