PDA

View Full Version : ECUflash / injector deadtime?



Matz
05-21-2006, 10:19 AM
I'm a noob at tuning, and have been wondering how I can modify my ROM to support my Helix 720cc injectors. I have been told that I should change the injector scaling entry to roughly 90% of the volumetric flow rate of the injectors (so around 648). However, I've also heard that you need to play with the injector battery voltage latency compensation.

First of all, does anyone already have the correct data for the Helix 720s? If not, can you post the numbers you're using, along with tech. specs on your injectors?

Secondly, if you're in the same boat and have no idea what the hell the battery compensation is for, here's a link to a site I found when trying to learn more about it. It's pretty good all-around info on injector timing calculations. Check it out! (http://www.stealth316.com/2-fuelinjection.htm)

EDIT -- here's another good page from DSMlink (http://dsmlink.com/details-fuel.html)

MR JayR
05-21-2006, 08:07 PM
Give Andy at SpeedElement a holler. He's an ECU Flash mastah.



I'm a noob at tuning, and have been wondering how I can modify my ROM to support my Helix 720cc injectors. I have been told that I should change the injector scaling entry to roughly 90% of the volumetric flow rate of the injectors (so around 648). However, I've also heard that you need to play with the injector battery voltage latency compensation.

First of all, does anyone already have the correct data for the Helix 720s? If not, can you post the numbers you're using, along with tech. specs on your injectors?

Secondly, if you're in the same boat and have no idea what the hell the battery compensation is for, here's a link to a site I found when trying to learn more about it. It's pretty good all-around info on injector timing calculations. Check it out! (http://www.stealth316.com/2-fuelinjection.htm)

EDIT -- here's another good page from DSMlink (http://dsmlink.com/details-fuel.html)

Matz
05-21-2006, 09:22 PM
Give Andy at SpeedElement a holler. He's an ECU Flash mastah.



I'm a noob at tuning, and have been wondering how I can modify my ROM to support my Helix 720cc injectors. I have been told that I should change the injector scaling entry to roughly 90% of the volumetric flow rate of the injectors (so around 648). However, I've also heard that you need to play with the injector battery voltage latency compensation.

First of all, does anyone already have the correct data for the Helix 720s? If not, can you post the numbers you're using, along with tech. specs on your injectors?

Secondly, if you're in the same boat and have no idea what the hell the battery compensation is for, here's a link to a site I found when trying to learn more about it. It's pretty good all-around info on injector timing calculations. Check it out! (http://www.stealth316.com/2-fuelinjection.htm)

EDIT -- here's another good page from DSMlink (http://dsmlink.com/details-fuel.html)



Thanks, yeah, he seems to be the only one on NCE actually tuning with the ECUflash right now. Actually, I think KevoVIII is, too.

earlyapex
05-21-2006, 09:49 PM
First of all, does anyone already have the correct data for the Helix 720s? If not, can you post the numbers you're using, along with tech. specs on your injectors?


The Helix 720's are made by Denso. I would like to know these tech specs as well since I use denso 720's. I have been lazy lately and haven't looked for it since I'm not using ecuflash yet, and well, because my coolant looks like chocolate milk.

vtluu
05-21-2006, 09:52 PM
Haven't the DSM guys already figured out the dead time for pretty much any injector you would conceivably want to run in the Evo?

Matz
05-21-2006, 10:16 PM
Haven't the DSM guys already figured out the dead time for pretty much any injector you would conceivably want to run in the Evo?


That's a good question. I will start searching through the DSM resources for this info. Bryan, thanks for reminding me that Denso makes the Helix injectors. That should make finding the data quite a bit easier.

EFIxMR
05-22-2006, 01:54 AM
I rescaled some 720 denso injectors for a customer today, and it went easy putting in the multiplier to 650. In terms of the battery compensation as soon as I get some more denso injectors to put in my own car to test, I'll get something figured out. to do it you need some specialized equipement, but nothing too fancy.

Beware try this at your own risk... This is standard practice in the world of tuning stand alone efi to determine the characteristics of an injector.

What you need to do is disconnect your alternator and battery and connect your car's battery wires to a variable voltage power supply. Basically, you turn the variable voltage supply to 14 volts... set the compensation so your afr is what you want. Then change the output of the supply to 13 volts and set the appropriate compensation so the afr is the same as before. Repeat this process until 11 volts. When you plug your alternator back in and your battery back in the injectors will have the correct compensation for any voltage fluctuation that may occur thru normal operation.

Matz
05-22-2006, 05:55 AM
I rescaled some 720 denso injectors for a customer today, and it went easy putting in the multiplier to 650. In terms of the battery compensation as soon as I get some more denso injectors to put in my own car to test, I'll get something figured out. to do it you need some specialized equipement, but nothing too fancy.

Beware try this at your own risk... This is standard practice in the world of tuning stand alone efi to determine the characteristics of an injector.

What you need to do is disconnect your alternator and battery and connect your car's battery wires to a variable voltage power supply. Basically, you turn the variable voltage supply to 14 volts... set the compensation so your afr is what you want. Then change the output of the supply to 13 volts and set the appropriate compensation so the afr is the same as before. Repeat this process until 11 volts. When you plug your alternator back in and your battery back in the injectors will have the correct compensation for any voltage fluctuation that may occur thru normal operation.


Hmm.. I was thinking about doing a benchtop test. However, I don't feel comfortable using actual fuel at work to do the tests... I was thinking about water, but I don't know if that'll give me poor results or not. I could easily set up a signal generator, SSR, and power supply to do cycle tests on injectors with a pressurized vessel. If you think water will work OK for testing, please let me know, and what the ideal pressure should be set at.

Otherwise, I'll just hang out with these injectors until you have been able to figure it out. :) Thank you very much for posting the how-to on doing the process in the car, though!!! :thumbsup:

EFIxMR
05-22-2006, 11:09 AM
from my understanding you can't use water you need to run a liquid that doesn't conduct electricity, and if you are doing a bench test the liquid can't be flammable.

Doing it on the car has the advantage of running it in its actual environment, even though you are running actual fuel it is safe (fire wise from fuel) because it is self contained inside the engine. The only danger (slight possibility) is shorting out your electrical system.

If you don't want to hook it up to an auxiliary power supply what you can do is basically just disconnect the alternator, and have it compensated for battery voltage, and interolate the points between that worse case scenario and regular operation with the alternator and battery connected 14+ volts.

Honestly I don't think it will really matter much, as I put in the denso's into yesterdays car and it worked fine with the base battery comps in place. Much like how a car can run decent/ok when putting in 720's and running an AFC.

edited for my bad morning english

KevOVIII
05-22-2006, 11:34 AM
Info for 12.5V lol

On Time: 0.65ms / 12.5v /2.5kg/mm
http://www.helix-power.com/Fuel/564-EN004.htm

I didn't change the battery compensation but short term and long term fuel trims are fairly close to one another. The 16.8 AFR that was flashed into the ECU has been changed to 14.7 again. I can't believe how the fuel table looked like it was a hacked job (before I reflashed the car with my own settings).

earlyapex
05-22-2006, 11:57 AM
I can't believe how the fuel table looked like it was a hacked job (before I reflashed the car with my own settings).


I can believe it. ;)

KevOVIII
05-22-2006, 11:04 PM
I can believe it. ;)


Hehe :D

"I see," said the blind man.

Matz
05-23-2006, 05:41 AM
Info for 12.5V lol

On Time: 0.65ms / 12.5v /2.5kg/mm
http://www.helix-power.com/Fuel/564-EN004.htm

I didn't change the battery compensation but short term and long term fuel trims are fairly close to one another. The 16.8 AFR that was flashed into the ECU has been changed to 14.7 again. I can't believe how the fuel table looked like it was a hacked job (before I reflashed the car with my own settings).


Great, thanks for that info, Kevin! Perhaps I will install the FP and injectors this weekend, and then just change the injector scaling to 650. I should probably better get my wideband hooked up to the ECU+ again and cross check the values against those on the remote display as well.

Matz
05-23-2006, 05:44 AM
from my understanding you can't use water you need to run a liquid that doesn't conduct electricity, and if you are doing a bench test the liquid can't be flammable.

Doing it on the car has the advantage of running it in its actual environment, even though you are running actual fuel it is safe (fire wise from fuel) because it is self contained inside the engine. The only danger (slight possibility) is shorting out your electrical system.

If you don't want to hook it up to an auxiliary power supply what you can do is basically just disconnect the alternator, and have it compensated for battery voltage, and interolate the points between that worse case scenario and regular operation with the alternator and battery connected 14+ volts.

Honestly I don't think it will really matter much, as I put in the denso's into yesterdays car and it worked fine with the base battery comps in place. Much like how a car can run decent/ok when putting in 720's and running an AFC.

edited for my bad morning english


So then what do the injector companies use for their benchtop testing? I don't understand why water isn't okay. Injectors are just solenoid actuated valves, so why would conduction matter? I could always use DI water. :wink:

I can definitely hook up an external adjustable power supply, as I have a lab supply at home that I use for stuff now and then. I'm just not comfortable with the parts after the connection. :lol:

EDIT -- I take back what I said before about using my adj. power supply at home... there's no way it supplies enough juice to start a car! 30A @ 5V so less at 12V, and I'm sure that starting the car requires a bunch of power. However, after re-reading your original post about the process, it doesn't seem that bad. Now I just have to look around to see if I can find a variable PS that'll provide enough output current to do the job!

Matz
05-24-2006, 12:10 PM
Andy, have you actually used an adj. power supply before to do this profiling? They are insanely expensive instruments.

In addition, why would they have battery voltage compensation below 11VDC? Much below 10V even, you'd have so many issues that I doubt the car would be running. :?:

EFIxMR
05-24-2006, 01:50 PM
quick update, on the denso 720's i rescaled the other day by changing the ECUflash scaling parameter to 650, the car kept on throwing P00171 (system lean) codes at idle even with the afr adjusted to 14.7

I tried changing the battery comp and even just using the "+" function once made the system to rich barely idling the car.

The remedy was leaving the injector scaling to 513, and manually changing the fuel tables to give the desired afr. This kept the CEL away, but had to take out a bunch of fuel on the top end.

We had that adj power supply at my old work, but it seems messing with that stuff is only necessary when stepping up to a huge injector. IMHO, it doesn't look like the denso 720 needs that stuff, and only causes more problems.

Maybe there is a more elegant solution but so far I havent found it. maybe you guys will have better luck?

Matz
05-24-2006, 01:56 PM
Thanks for the update, Andy. It's kind of funny because Bryan had to leave the injector scaling at the default value in the ECU+ as well. I know it's not necessarily related, but still funny. By any chance do you have a base map that just accounts for injectors that I can take a look at?

EFIxMR
05-24-2006, 02:03 PM
the car i did was actually a strange case, he had bought an afc neo from us before i knew that ecuflash would work, so the top end trimming was done on the afc so simplicity and speed sake. This was more of an extra service at no charge to make sure the car didnt run too rich partial throttle and cruise and not what I consider a "tune".

earlyapex
05-24-2006, 02:10 PM
Ugh, I was hoping this wasn't the case. I really want the scaling to work correctly. Maybe it is something you are overlooking?

I will try and look into it more when I get some time.

EFIxMR
05-24-2006, 02:16 PM
i hope so too, but the p0171 code comes up without pause just sitting there at idle, coming up right after you clear the code with the scanner.

EFIxMR
05-24-2006, 02:17 PM
i got some precision 850's that i can try too on my own car, ill see if the result is any better.

KevOVIII
05-24-2006, 02:18 PM
I just tested "+" for 11.72V and it jumped from 0.672 to 1.680. *That might be an issue. *:D

earlyapex
05-24-2006, 02:24 PM
Well, there is way to do this correctly. Mike and Shiv both scale injectors via reflashes. I'm sure others do as well.

EFIxMR
05-24-2006, 02:26 PM
whoops, yea... was in a time constriant as our customer was on a lunch break. but yea just scaling them alone didn't work well. maybe the key is in the battery comp.

earlyapex
05-24-2006, 02:33 PM
Did you mess around with different numbers then just 90% of 720?

Matz
05-24-2006, 02:46 PM
I just tested "+" for 11.72V and it jumped from 0.672 to 1.680. That might be an issue. :D


Maybe you can try using "=" instead, which lets you input an exact number (supposedly)

EFIxMR
05-24-2006, 02:50 PM
i tried various numbers between 650 and 513.

GST
05-24-2006, 03:28 PM
maybe the key is in the battery comp.


For the Denso's that is 100% correct

Matz
05-24-2006, 03:30 PM
i tried various numbers between 650 and 513.


But we're talking about the battery compensation, and how the "+" key doesn't give you a fine enough control over the time delays...

earlyapex
05-24-2006, 03:50 PM
Here is some more info:

Q:how do you determine injector dead time for injectors not listed on the supplied table in the m800 e.g. japanese factory style injectors low or high impedence do you just try different options available until the mixtures stay stable with voltage change or is there a method that can be explained of testing injectors to get the correct deadtime.

A"There are a couple of different opinions on how to do this currently.

In the field, one method involves running the engine at steady state with light load so as to have a small pulsewidth.

Add overall trim such as 5 % and see if you get a corresponding lambda change of 5%. If not, adjust the battery comp in the following way.

If you got more than 5% change, your battery comp is too small. If you got less than 5% your battery comp is too big.

When you get it right, you can change to a different voltage and test again building yourself a battery comp table based on your engine at your fuel pressure for your injectors.

http://www.efi101.com/forum/viewtopic.php?t=754&highlight=injector+compensation

KevOVIII
05-24-2006, 04:05 PM
I just tested "+" for 11.72V and it jumped from 0.672 to 1.680. That might be an issue. :D


Maybe you can try using "=" instead, which lets you input an exact number (supposedly)


Yup, the user can set values with "=" I just wanted to point out that "+" might throw things off... in this case, it did.

earlyapex
05-24-2006, 04:16 PM
A little birdy told me the Denos 720cc Deadtime is 450us.

From what I have read, this map in the ECU is battery voltage based. Lower battery voltage also slows the injectors down, larger injectors usually take longer to open, and the lower rpm's/pulsewidths will be too lean if this is not corrected for. Sounds like what Andy is running into.

KevOVIII
05-24-2006, 04:24 PM
At ~12V I assume. Pretty close to Helix's 650ms at 12.5V.

For stock, 14.06V = 432ms.

EFIxMR
05-24-2006, 05:23 PM
i tried various numbers between 650 and 513.


But we're talking about the battery compensation, and how the "+" key doesn't give you a fine enough control over the time delays...


This was in reply to Bryan if i had tried any other injector scaling parameters.

the "+" was definately an oversite on my part as it obviously isn't the fine tuning needed for these injectors.

EVO GRIM
05-24-2006, 06:26 PM
Might need to keep an eye on voltage for long term. In theory if you had a alternator go out it could yield a lean condition?

trinydex
05-24-2006, 11:08 PM
can someone like give me a 101 on this injector and battery talk?

vtluu
05-25-2006, 12:16 AM
Well, here's the deal: an injector injects fuel into the intake manifold when voltage is applied to it. You set the injector duty cycle depending on how much fuel you need. For example a 50% duty cycle applies 12V (more or less) to the injector for some given amount of time (the injector pulse width I think it's called), then no voltage for the same amount of time. Still with me?

That's all fine and well, but it doesn't account for the fact that when you apply power (voltage) to the injector, there's a delay between the time the power comes on, and the time the injector actually sprays fuel out. That's the injector dead time. If the voltage is lower, then the dead time will probably longer because it takes a bit more time for the injector to get going. That's why the injector dead time is a function of voltage.

The ECU has to know about the injector dead time so it can adjust the injector pulse width. So if you needed the injector putting in fuel 50% of the time, the time that voltage needs to be applied to the injector is actually a bit more than 50%--specifically, it's the injector pulse width plus the injector dead time.

If you have your A/C and fan and 1000W amp and subwoofer all turned on, the voltage from the battery/alternator circuit is going to be a bit less than usual, so the ECU needs to know that the injector dead time is a bit longer than usual, so it can continue supplying a consistent amount of fuel to the engine.

trinydex
05-25-2006, 12:19 AM
coo

vtluu
05-25-2006, 12:21 AM
That's my understanding anyway but I only learned about this the other day. Someone correct me if I'm wrong. 8)

Matz
05-25-2006, 04:53 AM
That's my understanding anyway but I only learned about this the other day. Someone correct me if I'm wrong. 8)


You are correct, as the behavior of solenoids will vary with voltage. IMHO, Mitsu is lame for not having a voltage regulator in the circuit to guarantee 12V (or the minimum allowable voltage to run an injector) going to the fuel injection system. This whole battery compensation thing seems like a hack to me (or they have to be real penny pinchers on the ECU PCB to save lots of money).

Matz
05-25-2006, 04:57 AM
Might need to keep an eye on voltage for long term. In theory if you had a alternator go out it could yield a lean condition?


Yup, the battery voltage will drop and, if not compensated for, you would end up injecting less fuel than expected because the extra time required to energize the solenoid wasn't added to the total injector time. Was that a run-on sentence? :oops:

earlyapex
05-25-2006, 10:53 AM
IMHO, Mitsu is lame for not having a voltage regulator in the circuit to guarantee 12V (or the minimum allowable voltage to run an injector) going to the fuel injection system.* This whole battery compensation thing seems like a hack to me (or they have to be real penny pinchers on the ECU PCB to save lots of money).


I believe every EFI car has battery comp. I could be wrong however.

Matz
05-25-2006, 10:58 AM
I believe every EFI car has battery comp. I could be wrong however.


Ok, but it doesn't mean that it's a good design.

earlyapex
05-25-2006, 11:06 AM
I believe every EFI car has battery comp. I could be wrong however.


Ok, but it doesn't mean that it's a good design.


I didn't say it was or wasn't. Has to be a reason why it's there though. Seems like an easy fix like you said, and you would think with millions of dollars going to ECU research, it could be fixed.. so there must be a reason for it to still be there.

I think it might be more for idle, etc where the RPM can dip below a certain number and the Alternator can't spin fast enough to keep the current voltage up so the battery takes over?

EFIxMR
05-25-2006, 12:05 PM
It might just be for simplicity's sake. with everything set correctly, these sorts of things are invisible to the driver.

When's the last time you guys thought about that before ECUflash and going with plus sized injectors.

earlyapex
05-25-2006, 12:06 PM
When's the last time you guys thought about that before ECUflash and going with plus sized injectors.


Trying to get injector comp to work with ECU+

Matz
05-25-2006, 03:58 PM
When's the last time you guys thought about that before ECUflash and going with plus sized injectors.


Trying to get injector comp to work with ECU+


Yeah, even without ECUflash, it's probably the first thing someone would wonder about. You have larger flow rate injectors, so you would probably add more fuel. Or would you? What if the injectors took longer to actuate because they flow more? At that point, you'd then likely realize that something must happen, but not know what exactly "that" is. :)

Matz
06-04-2006, 06:31 AM
Andy, do you have any updates on the 720cc injectors? :)

EFIxMR
06-04-2006, 03:37 PM
We recently jsut got an order of 720 densos back in. Don't have them in my own personal car, so thats why the development has been slow. I personally own a set of 850's and 1000's CC injectors, so not sure if i want to buy another set of 720's. Gotta talk with the rest of the guys at the shop, and see if we can open a set for R&D.

EVO GRIM
06-04-2006, 04:41 PM
Andy I need a set pretty soon. If you give me a smokin deal I'll let you test them out. For the good of the community of course.

Matz
06-04-2006, 05:22 PM
Andy I need a set pretty soon. If you give me a smokin deal I'll let you test them out. For the good of the community of course.


+1 :thumbsup:

EFIxMR
06-04-2006, 09:58 PM
I think bryan said he was messing around with the multiplier and injector battery comps maybe he can shed some light on the subject? I'm willing to help you out on the injectors if you can deal with the possibility of the check engine p0171 code popping up until we get it squared 100%. I could take a matter of mins to a matter of days.

earlyapex
06-04-2006, 10:06 PM
No p0171 here yet. We'll see how it is this week.

Best/easiest way to get the deadtime settings is to send the injectors and get them flow tested requesting deadtimes for all the voltages. ;)

It would be the fastest and easiest way to do it espically if you haven't put the injectors in yet.

Matz
06-05-2006, 07:10 AM
No p0171 here yet. We'll see how it is this week.

Best/easiest way to get the deadtime settings is to send the injectors and get them flow tested requesting deadtimes for all the voltages. ;)

It would be the fastest and easiest way to do it espically if you haven't put the injectors in yet.


I can do this if it's not too expensive. It would be good since a lot of Evo guys use the Helix / Denso 720s. I think all injectors will be a little different, but I would hope they are all still close. Who do you recommend? RC Engineering? if you know of a local place that's good, please let me know. Thanks!