cold start enrichment table
Feb 12, 2009, 10:16 PM
#211
Evolved Member
Join Date: Apr 2008
Location: Sydney, Australia
Posts: 604
Likes: 0
Received 0 Likes
on
0 Posts
There are a few tables that control the acceleration enrichment. Probably the easiest to use is what I labelled Accel Enrich Base vs Engine Temp. I have the table for my rom and I am working on the definition of other tables related to this, but have not searched for it for others. I am sure there is another thread somewhere for this function so it might be better to post there.
Mar 8, 2009, 11:19 AM
#213
Newbie
iTrader: (1)
Join Date: Feb 2007
Location: New York
Posts: 67
Likes: 0
Received 0 Likes
on
0 Posts
96420011:
<table name="IPW Adder for Cold Start #1" category="Fuel" address="38e8" type="2D" level="1" scaling="ColdStartTime">
<table name="Coolant Temperature" address="68f8" type="Y Axis" elements="9" scaling="Temp"/>
</table>
<table name="IPW Adder for Cold Start #2" category="Fuel" address="58c6" type="2D" level="1" scaling="ColdStartTime">
<table name="Coolant Temperature" address="68f8" type="Y Axis" elements="9" scaling="Temp"/>
</table>
<table name="IPW Adder for Cold Start #1" category="Fuel" address="38e8" type="2D" level="1" scaling="ColdStartTime">
<table name="Coolant Temperature" address="68f8" type="Y Axis" elements="9" scaling="Temp"/>
</table>
<table name="IPW Adder for Cold Start #2" category="Fuel" address="58c6" type="2D" level="1" scaling="ColdStartTime">
<table name="Coolant Temperature" address="68f8" type="Y Axis" elements="9" scaling="Temp"/>
</table>
Last edited by eitan; Mar 8, 2009 at 11:30 AM.
Mar 8, 2009, 08:21 PM
#214
That table should only be for the first pulse of fuel. It wont do anything after the motor has actually fired up. The smaller tables (should be 2 of them in that rom), are the ones that should work after the motor is running.
For 96530006 they are at 3732 and 3742. It looks like there are 2 decay maps as well at 3712 and 3722. Not sure what conditions determine which gets used, but try adjusting both.
For 96530006 they are at 3732 and 3742. It looks like there are 2 decay maps as well at 3712 and 3722. Not sure what conditions determine which gets used, but try adjusting both.
Is it this?
Code:
<table name="Cranking Enrichment IPW Adder (Main)" address="3732">
<table name="Coolant Temperature" address="69be"/>
</table>
<table name="Cranking Enrichment IPW Adder (Unused)" address="3742">
<table name="Coolant Temperature" address="69be"/>
</table>
Code:
<table name="Cranking Enrichment IPW Adder (Main)" address="3712">
<table name="Coolant Temperature" address="69be"/>
</table>
<table name="Cranking Enrichment IPW Adder (Unused)" address="3722">
<table name="Coolant Temperature" address="69be"/>
</table>
Mar 8, 2009, 09:03 PM
#215
Evolved Member
iTrader: (30)
Join Date: Dec 2003
Location: USA
Posts: 929
Likes: 0
Received 0 Likes
on
0 Posts
I can't find a final answer on what the Defs are for this rom. It looks like all 4 of those addresses pull up the same table. Has anyone finalized which address is which for 96530006?
Is it this?
or this?
Is it this?
Code:
<table name="Cranking Enrichment IPW Adder (Main)" address="3732">
<table name="Coolant Temperature" address="69be"/>
</table>
<table name="Cranking Enrichment IPW Adder (Unused)" address="3742">
<table name="Coolant Temperature" address="69be"/>
</table>
Code:
<table name="Cranking Enrichment IPW Adder (Main)" address="3712">
<table name="Coolant Temperature" address="69be"/>
</table>
<table name="Cranking Enrichment IPW Adder (Unused)" address="3722">
<table name="Coolant Temperature" address="69be"/>
</table>
Code:
<table name="Cranking Enrichment IPW Adder (Main)" address="58d0" category="Fuel" type="2D" scaling="TimeInjCranking">
<table name="Coolant Temperature" address="69d8" type="Y Axis" elements="9" scaling="Temp"/>
</table>
Last edited by fixem2; Mar 8, 2009 at 09:06 PM.
Mar 16, 2009, 08:32 AM
#216
I messed around with the post start decay tables a bit and things didn't work the way I was hoping. How certain are we that these are timers? Does anybody have a reliable time scaling for these decay tables if that is for sure what they are?
I did two datalogs with the car in open loop. First cold start, I had factory values and it was 9C at start up. Second datalog, I halved the factory values and it was 12 degrees on start up.
Because I wired my wideband into the ACC line instead of the RUN line, the wideband has to cylce the warmup circuit so I miss the first 7 seconds of the datalog. On both datalogs though, the AFR tapers from a rich condition to a lean condition in about 70 seconds. At about 120 seconds, both data logs get a big rich dip and then go lean and start tapering to 14.7:1. Both get to 14.7:1 around the time the car gets to 65C.
Reducing the timer in half doesn't seem to affect the time for the rich condition to taper off. Unless it's happening within the first 7 seconds where the wideband isn't on. The biggest difference between the logs though is that the half valued setting has an AFR of 13.5:1 when the wideband first starts logging. The factory value datalog are running 12.8:1 at the first sample.
I can't post up the datalog right now, but I'll try to get them up tonight and maybe somebody else can make sense of it.
I did two datalogs with the car in open loop. First cold start, I had factory values and it was 9C at start up. Second datalog, I halved the factory values and it was 12 degrees on start up.
Because I wired my wideband into the ACC line instead of the RUN line, the wideband has to cylce the warmup circuit so I miss the first 7 seconds of the datalog. On both datalogs though, the AFR tapers from a rich condition to a lean condition in about 70 seconds. At about 120 seconds, both data logs get a big rich dip and then go lean and start tapering to 14.7:1. Both get to 14.7:1 around the time the car gets to 65C.
Reducing the timer in half doesn't seem to affect the time for the rich condition to taper off. Unless it's happening within the first 7 seconds where the wideband isn't on. The biggest difference between the logs though is that the half valued setting has an AFR of 13.5:1 when the wideband first starts logging. The factory value datalog are running 12.8:1 at the first sample.
I can't post up the datalog right now, but I'll try to get them up tonight and maybe somebody else can make sense of it.
Mar 16, 2009, 10:05 PM
#217
Here are the first two data logs. The only values changed were the values in the post start enrichment decay tables that are said to be timers. The car is in openloop mode. Again, the first 7 seconds of the log of AFR is missing because of the wideband being in warmup mode.
Time based showing AFR and coolant temp. The halved value log has a slightly warmer coolant temp to start, but both of these are first thing in the morning without the car being started prior to the logs The thing really interesting to me is the dip that can been seen in both logs. While it looks like the halved values are leaner, the next chart I think gives better insight into what is happening. Also, towards the end of the halved value log, the AFR varies as if it is in closed loop mode. However, that is not the case, the targeted AFR is changing from 13.83:1 to 17.11:1 where the target AFR will remain constant at 14.7:1 normally while in closed loop, despite the AFR varing. I don't know why this happened on the second one but not on the first one.
This is the datalog showing AFR vs. coolant temp. Here it shows that even though the timers are halved, the AFR is relatively constant for a given engine coolant temp. The both also show a very large dip to the rich side at 40C for no apparent reason.
This last graph will lead into what I'll test out tomorrow morning. This is roughly what the multiplier would need to be to get the measured AFR. As you can see, it is saying I would need a multiplier of 0.9 to get my mid temp enrichment to what it is. This would lead me to believe my MAF Rescale is incorrect, however, the car ran lean on cold start even with the stock values in the MAF table and the car runs right at 14.7:1 at the lower MAF frequencies once the car is fully warmed up. It doesn't show anything more then the last table, but I plan to test the multipliers tomorrow and feel this may be a relevant chart at that time.
For tomorrow morning, I plan to change the enrichment values to 1.00 across the board. Hopefully it is down below 10C so I can get some values that should have heavy enrichment.
Time based showing AFR and coolant temp. The halved value log has a slightly warmer coolant temp to start, but both of these are first thing in the morning without the car being started prior to the logs The thing really interesting to me is the dip that can been seen in both logs. While it looks like the halved values are leaner, the next chart I think gives better insight into what is happening. Also, towards the end of the halved value log, the AFR varies as if it is in closed loop mode. However, that is not the case, the targeted AFR is changing from 13.83:1 to 17.11:1 where the target AFR will remain constant at 14.7:1 normally while in closed loop, despite the AFR varing. I don't know why this happened on the second one but not on the first one.
This is the datalog showing AFR vs. coolant temp. Here it shows that even though the timers are halved, the AFR is relatively constant for a given engine coolant temp. The both also show a very large dip to the rich side at 40C for no apparent reason.
This last graph will lead into what I'll test out tomorrow morning. This is roughly what the multiplier would need to be to get the measured AFR. As you can see, it is saying I would need a multiplier of 0.9 to get my mid temp enrichment to what it is. This would lead me to believe my MAF Rescale is incorrect, however, the car ran lean on cold start even with the stock values in the MAF table and the car runs right at 14.7:1 at the lower MAF frequencies once the car is fully warmed up. It doesn't show anything more then the last table, but I plan to test the multipliers tomorrow and feel this may be a relevant chart at that time.
For tomorrow morning, I plan to change the enrichment values to 1.00 across the board. Hopefully it is down below 10C so I can get some values that should have heavy enrichment.
Last edited by 03whitegsr; Mar 16, 2009 at 10:08 PM.
Mar 23, 2009, 11:53 AM
#218
I did some more testing on this and was able to get some interesting results.
This morning, by zeroing out the post start enrichment (set to 1.0) and decay tables (set to 0) I was able to get the car to not start very consistently. Same settings midday however and the car started up just fine.
This lead me to some more testing by zeroing out each startup/post start table that was listed in this thread. Surprisingly, the car started regardless of what the settings were. Also suprising, the injector pulse width was all over the place, but actual AFR didn't seem to vary much.
Every table did have an effect on the injector pulse width, including the 1D entry of "Cranking Injector Latency". Actually, setting this one value to zero did the exact same thing as setting everything to zero.
I'll post up some graphs tonight. They look like they may help sort out the decay timers time scaling.
This morning, by zeroing out the post start enrichment (set to 1.0) and decay tables (set to 0) I was able to get the car to not start very consistently. Same settings midday however and the car started up just fine.
This lead me to some more testing by zeroing out each startup/post start table that was listed in this thread. Surprisingly, the car started regardless of what the settings were. Also suprising, the injector pulse width was all over the place, but actual AFR didn't seem to vary much.
Every table did have an effect on the injector pulse width, including the 1D entry of "Cranking Injector Latency". Actually, setting this one value to zero did the exact same thing as setting everything to zero.
I'll post up some graphs tonight. They look like they may help sort out the decay timers time scaling.
Mar 23, 2009, 01:32 PM
#219
Evolved Member
Join Date: Apr 2008
Location: Sydney, Australia
Posts: 604
Likes: 0
Received 0 Likes
on
0 Posts
The decays only operate for the first few seconds, so the WideBand most likely wont see it. Try maxing the decays out at 255 and it will give you an indication of the timer.
The engine wont always need large amounts of post start, and most of the time will start fine without it. There is also the fuel pressure post start to consider so the injector post start also can have less effect.
The engine wont always need large amounts of post start, and most of the time will start fine without it. There is also the fuel pressure post start to consider so the injector post start also can have less effect.
Mar 23, 2009, 04:04 PM
#220
The logs I have make it look like the decay timers actually happen at different times. It looks like it uses the first table for the first few seconds, then switches to the second decay timer.
I'm wondering if setting the timers to zero is actually causing some problems though and will try it again with the timers set to one. I'm getting erroneous data on some of the other evoscan channels when I have the timers set to 0. Kind of strange.
I'm wondering if setting the timers to zero is actually causing some problems though and will try it again with the timers set to one. I'm getting erroneous data on some of the other evoscan channels when I have the timers set to 0. Kind of strange.
Mar 27, 2009, 02:40 PM
#221
Having done a ton of map switching and datalogging now, I can say I am 100% sure...
That I'm still not sure exactly how these maps work.
But I have noticed some trends.
1. What I've been referring to as decay timers appear to be count down multipliers. The larger the numbers are, the more enrichment you get and the longer the enrichment lasts for. Maxing it out at 255 gives you a TON of post start enrichment. 255 in the timer tables with the factory enrichment values will get you 30ms of pulsewidth that takes about 17 seconds to bleed off.
2. The enrichment settings have thoroughly confused me, as treating them like multipliers doesn't make sense, because a 10% change in value does not give you a 10% change in AFR. Also, using values under 128 does NOT have an effect of multiplying by a number less then 1 to reduce total enrichment. If they were pure multipliers, it would greatly reduce enrichment, but it doesn't.
3. Trying to figure out scalings for these tables has been a waste of time. I'm going with mattjin on this and just leaving as uint8 values and simply tuning to get the car to run the way you want it.
4. Tuning wise, if you want more fuel for the same amount of time, up the enrichment tables. It doesn't take much to make a pretty big difference. If you want more fuel for a longer period of time, up the timers, again only a small value change makes a big difference here. If you want the same fuel, but for a longer period of time...good luck figuring it out.
5. I can't tell when one map is used over the other. It looks like they are both used, but one of them seems to be dominant. I would just make them both the same though as that's the way the factory maps are.
6. I have a very hard time believe that there isn't a warmup enrichment table that controls AFR based on engine coolant temperature. It's somewhere in there. The car will reliably hit the same AFRs at a given coolant temperature in an obvious enrichment curve as shown in the graph I posted before. This was the entire point of testing theses tables, as I wanted to see if I could affect that curve and absolutely nothing I did changed that curve. There is either a table or an algorithm somewhere that is responsible for that curve and I would badly like to find it so I can richen up my warmup AFRs.
That I'm still not sure exactly how these maps work.
But I have noticed some trends.1. What I've been referring to as decay timers appear to be count down multipliers. The larger the numbers are, the more enrichment you get and the longer the enrichment lasts for. Maxing it out at 255 gives you a TON of post start enrichment. 255 in the timer tables with the factory enrichment values will get you 30ms of pulsewidth that takes about 17 seconds to bleed off.
2. The enrichment settings have thoroughly confused me, as treating them like multipliers doesn't make sense, because a 10% change in value does not give you a 10% change in AFR. Also, using values under 128 does NOT have an effect of multiplying by a number less then 1 to reduce total enrichment. If they were pure multipliers, it would greatly reduce enrichment, but it doesn't.
3. Trying to figure out scalings for these tables has been a waste of time. I'm going with mattjin on this and just leaving as uint8 values and simply tuning to get the car to run the way you want it.
4. Tuning wise, if you want more fuel for the same amount of time, up the enrichment tables. It doesn't take much to make a pretty big difference. If you want more fuel for a longer period of time, up the timers, again only a small value change makes a big difference here. If you want the same fuel, but for a longer period of time...good luck figuring it out.
5. I can't tell when one map is used over the other. It looks like they are both used, but one of them seems to be dominant. I would just make them both the same though as that's the way the factory maps are.
6. I have a very hard time believe that there isn't a warmup enrichment table that controls AFR based on engine coolant temperature. It's somewhere in there. The car will reliably hit the same AFRs at a given coolant temperature in an obvious enrichment curve as shown in the graph I posted before. This was the entire point of testing theses tables, as I wanted to see if I could affect that curve and absolutely nothing I did changed that curve. There is either a table or an algorithm somewhere that is responsible for that curve and I would badly like to find it so I can richen up my warmup AFRs.
Last edited by 03whitegsr; Mar 27, 2009 at 02:42 PM.
The following users liked this post:
400s (Apr 9, 2020)
Apr 13, 2009, 05:10 PM
#222
Evolved Member
iTrader: (26)
Join Date: Oct 2005
Location: Pittsburgh
Posts: 1,712
Likes: 0
Received 0 Likes
on
0 Posts
Mrfred,
On your original post you show the log of how it matches up with the table - did you log IPW in evoscan or is there something different to log?
Reason being - It does not seem my ipw at start up follows the table at all? I am trying to diagnose why my car wont warm start, I just dont get it. It cold starts and hot starts perfect - if the car sits for an extended period of time the car will just not start without a lot of cranks. I thought it was my FIC1050's so I decided to try the new bosch 1000cc saturated injectors and I am having the same exact problem. These are excellent fuel injectors but for some reason my car is being its typical satin self and doing un-explainable things. My friend runs these as well with 0 issues so im sure its something dumb im missing.......
But seeing the ipw start up table not matching with the ipw on startup makes this hard to diagnose.
Thanks
- Trav
On your original post you show the log of how it matches up with the table - did you log IPW in evoscan or is there something different to log?
Reason being - It does not seem my ipw at start up follows the table at all? I am trying to diagnose why my car wont warm start, I just dont get it. It cold starts and hot starts perfect - if the car sits for an extended period of time the car will just not start without a lot of cranks. I thought it was my FIC1050's so I decided to try the new bosch 1000cc saturated injectors and I am having the same exact problem. These are excellent fuel injectors but for some reason my car is being its typical satin self and doing un-explainable things. My friend runs these as well with 0 issues so im sure its something dumb im missing.......
But seeing the ipw start up table not matching with the ipw on startup makes this hard to diagnose.
Thanks
- Trav
Last edited by travman; Apr 14, 2009 at 01:48 PM.
Apr 13, 2009, 07:02 PM
#223
Mrfred,
On your original post you show the log of how it matches up with the table - did you log IPW in evoscan or is there something different to log?
Reason being - It does not seem my ipw at start up follows the table at all? I am trying to diagnose why my car wont warm start, I just dont get it. It cold starts and hot starts perfect - if the car sits for an extended period of time the car will just not start without a lot of cranks. I thought it was my FIC1050's so I decided to try the new bosch 1000cc saturated injectors and I am having the same exact problem.
But seeing the ipw start up table not matching with the ipw on startup makes this hard to diagnose.
Thanks
- Trav
On your original post you show the log of how it matches up with the table - did you log IPW in evoscan or is there something different to log?
Reason being - It does not seem my ipw at start up follows the table at all? I am trying to diagnose why my car wont warm start, I just dont get it. It cold starts and hot starts perfect - if the car sits for an extended period of time the car will just not start without a lot of cranks. I thought it was my FIC1050's so I decided to try the new bosch 1000cc saturated injectors and I am having the same exact problem.
But seeing the ipw start up table not matching with the ipw on startup makes this hard to diagnose.
Thanks
- Trav
When you log, its necessary to turn the key to "on", start EvoScan logging at highest speed (Logger superspeed, minimal number of parameters being logged, logging to screen disabled), then crank over the engine.
Apr 13, 2009, 07:33 PM
#224
Evolved Member
iTrader: (26)
Join Date: Oct 2005
Location: Pittsburgh
Posts: 1,712
Likes: 0
Received 0 Likes
on
0 Posts
If you look in the first post of this thread, there are some EvoScan IPW measurements from testing cold start IPW. I did it as a function of engine temp rather than changing the cranking enrichment value. The table definitely affects cranking enrichment. During the middle of winter on E85, I accidentally reverted to a ROM with low cranking enrichment values, and the car would take several tries to start. Once I put in the corrrect ROM, it fired right up.
When you log, its necessary to turn the key to "on", start EvoScan logging at highest speed (Logger superspeed, minimal number of parameters being logged, logging to screen disabled), then crank over the engine.
When you log, its necessary to turn the key to "on", start EvoScan logging at highest speed (Logger superspeed, minimal number of parameters being logged, logging to screen disabled), then crank over the engine.
I will try logging as you said - I log it to screen and watch while its cranking and then go back after to review the logs closer. No matter the value in the table my ipw is always around 4 on cranking, while my table shows something totally different.
Apr 13, 2009, 11:31 PM
#225
Mrfred,
On your original post you show the log of how it matches up with the table - did you log IPW in evoscan or is there something different to log?
Reason being - It does not seem my ipw at start up follows the table at all? I am trying to diagnose why my car wont warm start, I just dont get it. It cold starts and hot starts perfect - if the car sits for an extended period of time the car will just not start without a lot of cranks. I thought it was my FIC1050's so I decided to try the new bosch 1000cc saturated injectors and I am having the same exact problem.
But seeing the ipw start up table not matching with the ipw on startup makes this hard to diagnose.
Thanks
- Trav
On your original post you show the log of how it matches up with the table - did you log IPW in evoscan or is there something different to log?
Reason being - It does not seem my ipw at start up follows the table at all? I am trying to diagnose why my car wont warm start, I just dont get it. It cold starts and hot starts perfect - if the car sits for an extended period of time the car will just not start without a lot of cranks. I thought it was my FIC1050's so I decided to try the new bosch 1000cc saturated injectors and I am having the same exact problem.
But seeing the ipw start up table not matching with the ipw on startup makes this hard to diagnose.
Thanks
- Trav
https://www.evolutionm.net/forums/al...ml#post6936890