Bosch 1000cc injectors on an Evo
I don't see how you say that at all?
Go log the car, coolant temps drop from 80C down to a lower range very quickly.
Realize, that the true coolant temps don't drop as quickly, but just the temp near the coolant temp sensor. This is all that matters though as that's the temp the ECU sees while cranking.
Regardless, please explain why you want to disregard a potential solution before even trying it?
It's like you have some kind of vested interest in seeing these injectors not work???
Go log the car, coolant temps drop from 80C down to a lower range very quickly.
Realize, that the true coolant temps don't drop as quickly, but just the temp near the coolant temp sensor. This is all that matters though as that's the temp the ECU sees while cranking.
Regardless, please explain why you want to disregard a potential solution before even trying it?
It's like you have some kind of vested interest in seeing these injectors not work???
I think this is a very good point tony. I also think this is a very real possibility. I forgot to mention it, but I'm pretty **** with my wiring. I have removed most if not all the grounds on my evo, buffed the connectors, and applied electrical compound to them. I have also added a couple engine to chassis grounds with some very nice multi strand 10AWG cable I have around.
Ok, did some logs and more to follow.
Drove into work, local temp is -6C
Car sat for 45 minutes
Coolant temp was 50C when cranking
No throttle at all was used
Car took 8 seconds to get an initial fire, sputtered and stalled after the little kick.
Next crank the car fired right up
This is pretty typical of the warm starts I have been seeing.
With regards to what Tony said, it may be an issue.
While cranking, evoscan records a battery voltage of 1.4 V.
Measured voltage at the injector started at 9.8V and dropped down to 9.4V before kicking over. Voltage at injector was 9.2V when cranking the second time when it started right up.
On a side note, I'm using a Westco AGM Miata battery that is about a year old. Voltage before cranking was 11.8 V. Appartently the battery is a little weak? Might be part of the issue of why one car has troubles when another does not. Seeing the cold battery power, I'm very tempted to change over to a lighter battery because I don't think they perform any worse then that.
As I mentioned, the coolant temps dropped to 50C in 45 minutes from roughly 85C (typical temperature I see while driving around). The car took about 2 minutes of idling to get coolant temps back up to 80C. Like I said, I don't think true coolant temps drops that fast. I think that the coolant temp near the sensor just drops lower quicker since there is less metal in the area to hold heat and the heat transfer to atmosphere can happen quicker.
I changed the cranking enrichment table from 32C-80C and will go back out in 30 minutes see how it reacts to the changes.
Drove into work, local temp is -6C
Car sat for 45 minutes
Coolant temp was 50C when cranking
No throttle at all was used
Car took 8 seconds to get an initial fire, sputtered and stalled after the little kick.
Next crank the car fired right up
This is pretty typical of the warm starts I have been seeing.
With regards to what Tony said, it may be an issue.
While cranking, evoscan records a battery voltage of 1.4 V.
Measured voltage at the injector started at 9.8V and dropped down to 9.4V before kicking over. Voltage at injector was 9.2V when cranking the second time when it started right up.
On a side note, I'm using a Westco AGM Miata battery that is about a year old. Voltage before cranking was 11.8 V. Appartently the battery is a little weak? Might be part of the issue of why one car has troubles when another does not. Seeing the cold battery power, I'm very tempted to change over to a lighter battery because I don't think they perform any worse then that.
As I mentioned, the coolant temps dropped to 50C in 45 minutes from roughly 85C (typical temperature I see while driving around). The car took about 2 minutes of idling to get coolant temps back up to 80C. Like I said, I don't think true coolant temps drops that fast. I think that the coolant temp near the sensor just drops lower quicker since there is less metal in the area to hold heat and the heat transfer to atmosphere can happen quicker.
I changed the cranking enrichment table from 32C-80C and will go back out in 30 minutes see how it reacts to the changes.
Last edited by 03whitegsr; Dec 21, 2009 at 09:05 AM.
I'm not too sure I trust the log from evoscan.
I've noticed in the past that when the ECU is cranking, it can mess up evoscan and create errounous data. I'm seeing it in the most recent log with coolant temp and engine speed jumping all over the place on initial cranking then it holds that value until either it is running or I cycle the key from run to start. Voltage consistently shows up at 1.4V while cranking though.
If the IPW data is right though, I'm seeing a noticable trend. The IPW starts out very low, 5-6ms after a few seconds of cranking, it steps up to 10ms. Shortly after that, it kicks over, sputters, and stalls. Now, the next time I start it, IPW is up at 21ms and the car fires immediately.
Anyway, no noticeable change in start raising the raw table values from 2 to 8 in the 32-80C range. I'm starting to think I need to raise the values considerably higher.
When I looked through the sub-routine for starting, I noticed Mrfred noted a timing table where the longer you crank, it changes the IPW. At around the same point, the sub-routine branches into two other branches. Mrfred didn't evaluate these though and I'm wondering if they have something to do with "if the car didn't run for X seconds last time, jump to this other branch." Or something similar. The IPW is definitely different after cycling the starter.
I've noticed in the past that when the ECU is cranking, it can mess up evoscan and create errounous data. I'm seeing it in the most recent log with coolant temp and engine speed jumping all over the place on initial cranking then it holds that value until either it is running or I cycle the key from run to start. Voltage consistently shows up at 1.4V while cranking though.
If the IPW data is right though, I'm seeing a noticable trend. The IPW starts out very low, 5-6ms after a few seconds of cranking, it steps up to 10ms. Shortly after that, it kicks over, sputters, and stalls. Now, the next time I start it, IPW is up at 21ms and the car fires immediately.
Anyway, no noticeable change in start raising the raw table values from 2 to 8 in the 32-80C range. I'm starting to think I need to raise the values considerably higher.
When I looked through the sub-routine for starting, I noticed Mrfred noted a timing table where the longer you crank, it changes the IPW. At around the same point, the sub-routine branches into two other branches. Mrfred didn't evaluate these though and I'm wondering if they have something to do with "if the car didn't run for X seconds last time, jump to this other branch." Or something similar. The IPW is definitely different after cycling the starter.
I noticed something interesting in travis's car yesterday evening during a warm start. He cranked the engine for 5-6 seconds and if he let off on the starter at the right point in the rotation the engine sputtered to life. To me this screams voltage drop issue. To test I'd like to see if someone, with the starting problem can park the car on an incline and catch the car in gear (second works well) to try to start it. Also if someone can put a set of jumper cables on their car and attempt to start it during a "warm start" condition.
The evo electrical system seems pretty weak. A heat soaked starter might be drawing more current than a cold one, reducing the ACTUAL voltage at the injectors. That or the main relay could be completely dropping out during cranking with insufficient voltage. During a hot start (like right after shutting it down) the battery voltage is still quite high from the alternator keeping it at 14.8 volts. Thus the car can still fire fine.
It would make sense to me that a properly designed high impedance injector could draw more current than the stock low impedance with injector resistors in-line. That coupled with the weak evo electrical system (which is a fact on Mitsubishi) could explain the problem.
I also have a fix in mind that does not require modifying the cars harness if we can narrow this down.
The evo electrical system seems pretty weak. A heat soaked starter might be drawing more current than a cold one, reducing the ACTUAL voltage at the injectors. That or the main relay could be completely dropping out during cranking with insufficient voltage. During a hot start (like right after shutting it down) the battery voltage is still quite high from the alternator keeping it at 14.8 volts. Thus the car can still fire fine.
It would make sense to me that a properly designed high impedance injector could draw more current than the stock low impedance with injector resistors in-line. That coupled with the weak evo electrical system (which is a fact on Mitsubishi) could explain the problem.
I also have a fix in mind that does not require modifying the cars harness if we can narrow this down.
Another customer of mine brought up the potential issue of the pump not priming properly when run in the low voltage state. Has anyone with problems tried bypassing the low speed fuel pump control so it's at full voltage during cranking?
" EVO’s have a fuel speed control which switches from high speed to low speed. I use this feature on aftermarket systems if available to help stop the fuel getting hot and for noise, etc. I use the Motec Dual Half Bridge to control on our twin pump system. Now, everyone puts a high volume pump which I assume would draw more current than the factory unit. I just went outside and started a flash tuned EVO 9 with Walbro pump (Standard injectors) and noted that I could hear the pump on prime which suggests it might be on low speed mode on crank.
The ViPEC ECU has Fuel speed control feature which will keep on slow speed until x% injector duty and over xRPM which you can configure. I tend to produce my own GP table to allow the pump to prime at full speed and revert to low speed once it is running and the alternator is charging. I have found low speed mode will stop the pump priming properly. Once the alternator is charging low speed works fine. I run the ID1000’s at over 75PSI base fuel pressure on some EVO’s with E85 so I can flow the HP required. This was leading to starting isues but only because there was not enough power available to the fuel pump in slow speed mode when I cranked the engine."
" EVO’s have a fuel speed control which switches from high speed to low speed. I use this feature on aftermarket systems if available to help stop the fuel getting hot and for noise, etc. I use the Motec Dual Half Bridge to control on our twin pump system. Now, everyone puts a high volume pump which I assume would draw more current than the factory unit. I just went outside and started a flash tuned EVO 9 with Walbro pump (Standard injectors) and noted that I could hear the pump on prime which suggests it might be on low speed mode on crank.
The ViPEC ECU has Fuel speed control feature which will keep on slow speed until x% injector duty and over xRPM which you can configure. I tend to produce my own GP table to allow the pump to prime at full speed and revert to low speed once it is running and the alternator is charging. I have found low speed mode will stop the pump priming properly. Once the alternator is charging low speed works fine. I run the ID1000’s at over 75PSI base fuel pressure on some EVO’s with E85 so I can flow the HP required. This was leading to starting isues but only because there was not enough power available to the fuel pump in slow speed mode when I cranked the engine."
Well everyone can benefit from a rewired fuel pump which will go a long way in helping voltage drop at the pump.
I dunno, it seems to odd to me that I have done 6 of these now with basically no start issues on warm/hot restarts. I am far from a genius when it comes to things electrical, but it seems that there are definitely ways to tune around the issue for the time being. All the cars we do are modded 255s with stock wiring. It does also happen to be that they are all SD, with the exception of Eric's car that is SD with all the MAF stuff in place (I assume it would bypass this like it wasnt there).
I dunno, it seems to odd to me that I have done 6 of these now with basically no start issues on warm/hot restarts. I am far from a genius when it comes to things electrical, but it seems that there are definitely ways to tune around the issue for the time being. All the cars we do are modded 255s with stock wiring. It does also happen to be that they are all SD, with the exception of Eric's car that is SD with all the MAF stuff in place (I assume it would bypass this like it wasnt there).
I don't know if you guys all recall...
I tested putting a battery tender on the car as soon as I turned the car off.
I put a fan on the intake manifold with the hood open upon shutdown.
I primed the fuel pump every 2-3 minutes via Evoscan.
The car started better, but not right. To clarify... I did ALL of these things at the same time.
My car started crappy with the Odyssey battery and with the STOCK battery.
I tested putting a battery tender on the car as soon as I turned the car off.
I put a fan on the intake manifold with the hood open upon shutdown.
I primed the fuel pump every 2-3 minutes via Evoscan.
The car started better, but not right. To clarify... I did ALL of these things at the same time.
My car started crappy with the Odyssey battery and with the STOCK battery.
JB, it could just be an issue with how good of condition the electrical system is.
My car always has the battery and ABS lights on while idling and the battery voltage at idle is in the 12v range. It has done this ever since I replaced my alternator after it went out. I got an OEM alternator and one thing I noticed, my original alternator had a slightly smaller pulley on it then the replacement. The original is long gone as a core exchange so I can't do anything about it now.
Point being, I noticed R/T Ernie's battery light seemed to be on while idling too. And I've seen numerous posts about other people having the same issue and wonder if those experiencing the poor start up problem have weak batteries or alternators.
If these injectors are very sensitive to low voltages (which they seem to be based on the latency table) it very well could just be an issue with voltage and the injectors just aren't opening under some conditions.
All that said, I'm on revision 13 of tuning the cranking and primer tables.
I can get very marked improvements in warm starts by adding fuel. I can't get start up perfect, but I can get it much more acceptable.
The problem though, as I richen up the cranking and primer tables, I start to get flooding when the motor is hot. Table settings that work the best for warm starts make the car almost unstartable when the engine is hot, taking 10 seconds of 100% throttle cranking to clear out the fuel before it will start.
Tony, would it be possible to get some lattency numbers for very low voltages? If we could get 1-2 more latency values in the 4-7V range at 43 psi, it might help sort some of this out.
My car always has the battery and ABS lights on while idling and the battery voltage at idle is in the 12v range. It has done this ever since I replaced my alternator after it went out. I got an OEM alternator and one thing I noticed, my original alternator had a slightly smaller pulley on it then the replacement. The original is long gone as a core exchange so I can't do anything about it now.
Point being, I noticed R/T Ernie's battery light seemed to be on while idling too. And I've seen numerous posts about other people having the same issue and wonder if those experiencing the poor start up problem have weak batteries or alternators.
If these injectors are very sensitive to low voltages (which they seem to be based on the latency table) it very well could just be an issue with voltage and the injectors just aren't opening under some conditions.
All that said, I'm on revision 13 of tuning the cranking and primer tables.
I can get very marked improvements in warm starts by adding fuel. I can't get start up perfect, but I can get it much more acceptable.
The problem though, as I richen up the cranking and primer tables, I start to get flooding when the motor is hot. Table settings that work the best for warm starts make the car almost unstartable when the engine is hot, taking 10 seconds of 100% throttle cranking to clear out the fuel before it will start.
Tony, would it be possible to get some lattency numbers for very low voltages? If we could get 1-2 more latency values in the 4-7V range at 43 psi, it might help sort some of this out.
Last edited by 03whitegsr; Dec 23, 2009 at 08:05 AM.
Dan and I will be testing the "electrical" side over xmas - we both will be back at hour hometown and have extended time off between the holidays.....
I cant even begin to tell you how many hrs + revisions/flashes to try and tune it out - I have been messing with this religiously since April 09. I have replaced part after part, etc etc and have yet to even get a solution to slap me in the face saying hey your onto something! I also had no luck with SD other than the car ran much better, but starting stayed the same.
I just don't think the start-up fueling require precision as some people don't even modify the start-up tables and they have no starting issues hot,warm, or cold. It also comes back to the evo is the only platform reporting these issues - if this was a random "injector" issue then you would have complaints from all makes/models running these but that is not the case. Its isolated to the evo - so now its just what is the problem / problems?
I think dans theory is very valid and i am curious to get some testing done and see what happens......we will report back after xmas with any findings. If there is anyone that can find an electrical problem its dan - he isn't lying when he says he is ****...must be the electrical engineer in him
I cant even begin to tell you how many hrs + revisions/flashes to try and tune it out - I have been messing with this religiously since April 09. I have replaced part after part, etc etc and have yet to even get a solution to slap me in the face saying hey your onto something! I also had no luck with SD other than the car ran much better, but starting stayed the same.
I just don't think the start-up fueling require precision as some people don't even modify the start-up tables and they have no starting issues hot,warm, or cold. It also comes back to the evo is the only platform reporting these issues - if this was a random "injector" issue then you would have complaints from all makes/models running these but that is not the case. Its isolated to the evo - so now its just what is the problem / problems?
I think dans theory is very valid and i am curious to get some testing done and see what happens......we will report back after xmas with any findings. If there is anyone that can find an electrical problem its dan - he isn't lying when he says he is ****...must be the electrical engineer in him
The subarus have similar but to a lesser degree of problems with these injectors too.
It could just be a number of things adding up to cause problems.
From what I have seen though, the car actually is pretty sensitive to the IPW on start up when it's warm. If you convert the stock ECU tables to raw numbers, then you may be able to compare to these numbers.
First off, I multiplied the stock tables by 0.54. This is due to my elevation and the Baro reading no longer being accurate due to SD patch and then the difference in injector size.
Car cold starts VERY well by doing this.
Cranking Enrichment
A value of 4 or less and the car hot starts fine. A value of 5 and the car floods when hot.
Jack it up to 10 or so and it warm starts pretty well, but again, it floods when hot.
You can drop it to 0 and the car will still hot start fine, but barely starts when warm.
Primer table
Stock value in the warm range of the table is 300.
Going under 300 doesn't seem to change anything.
350 and the car warm starts better but is starting to almost flood, depending on the cranking enrichment value.
400 and warm starts improve but the car floods when hot.
I've tried a ton of different combinations in this range. Any setting that dramatically improves warm starts leads to flooding on hot starts though.
It could just be a number of things adding up to cause problems.
From what I have seen though, the car actually is pretty sensitive to the IPW on start up when it's warm. If you convert the stock ECU tables to raw numbers, then you may be able to compare to these numbers.
First off, I multiplied the stock tables by 0.54. This is due to my elevation and the Baro reading no longer being accurate due to SD patch and then the difference in injector size.
Car cold starts VERY well by doing this.
Cranking Enrichment
A value of 4 or less and the car hot starts fine. A value of 5 and the car floods when hot.
Jack it up to 10 or so and it warm starts pretty well, but again, it floods when hot.
You can drop it to 0 and the car will still hot start fine, but barely starts when warm.
Primer table
Stock value in the warm range of the table is 300.
Going under 300 doesn't seem to change anything.
350 and the car warm starts better but is starting to almost flood, depending on the cranking enrichment value.
400 and warm starts improve but the car floods when hot.
I've tried a ton of different combinations in this range. Any setting that dramatically improves warm starts leads to flooding on hot starts though.
I don't know if you guys all recall...
I tested putting a battery tender on the car as soon as I turned the car off.
I put a fan on the intake manifold with the hood open upon shutdown.
I primed the fuel pump every 2-3 minutes via Evoscan.
The car started better, but not right. To clarify... I did ALL of these things at the same time.
My car started crappy with the Odyssey battery and with the STOCK battery.
I tested putting a battery tender on the car as soon as I turned the car off.
I put a fan on the intake manifold with the hood open upon shutdown.
I primed the fuel pump every 2-3 minutes via Evoscan.
The car started better, but not right. To clarify... I did ALL of these things at the same time.
My car started crappy with the Odyssey battery and with the STOCK battery.
Has anyone tested voltage at the injectors yet during cranking?
There's no sense in getting dead times below 8v if the injector isn't seeing it. If the injector IS seeing less than 8v, and the computer is not, then again, no sense in getting that data, the electrical system needs to be fixed if that's the case. Not saying this is what's going on, but it's worth checking out.
I'm not too confident that's it though because the guys with aftermarket ecus on the same car don't seem to be having these same issues.
There's no sense in getting dead times below 8v if the injector isn't seeing it. If the injector IS seeing less than 8v, and the computer is not, then again, no sense in getting that data, the electrical system needs to be fixed if that's the case. Not saying this is what's going on, but it's worth checking out.
I'm not too confident that's it though because the guys with aftermarket ecus on the same car don't seem to be having these same issues.