| View previous topic :: View next topic |
| Author |
Message |
hotrod
Joined: 19 Apr 2005 Posts: 872 Location: Colorado
|
Posted: Sun Feb 11, 2007 1:29 pm Post subject: How to make your car (Subaru) a FFV at home. |
|
|
How to make your car (Subaru) a FFV at home with very little money or effort.
Well I have been sitting on my hands, and biting my lip now for a couple months waiting until I had the data to backup what I am about to suggest.
I believe I have figured out how (at least on the Subaru WRX) to quickly and easily achieve a home made FFV. The solution is so simple it makes my head hurt, but I wanted to check it out to see if the ECU would do what I expected it to do. (sometimes those electrons do odd things you don't anticipate)
=====================
First a review of the basics.
Pure E85 requires about 30% more fuel flow that pure gasoline to run normally.
The Stock ECU handles up to a 30% blend of E85 with absolutely no problems, not even a CEL.
The Stock ECU will run just fine on a 50% blend of E85 but will throw a CEL because it has used up its fuel trim authority of +/- 25% to get closed loop fueling back to Stoich (Lambda 1.00).
E85 is a cool burning high octane fuel that will tolerate leaner mixtures than gasoline under high load.
Gasoline likes rich mixtures under high load to prevent detonation if its octane is a bit low.
=====================
On my first attempt to convert the car, I switched to the JDM STi 550 cc/min injectors and made no other tuning changes. This increased the fuel flow by 25% (stock injectors are 440 cc/min).
The car ran beautifully on 100% E85, ran good on 50/50 E85 and gasoline but developed a small hitch on transition to open loop fueling from closed loop fueling near 30% -40% E85..
The light dawns !!
At this point I realized that if you could setup a circumstance where the ECU was sitting on zero fuel trims at a 50/50 mix of gasoline and E85 it had enough fuel trim authority to cover both extremes of 100% gasoline or 100% E85.
At this point I planned out my next experiment ---- unfortunately it took me nearly a year to assemble the parts I needed (not because they were necessary for the tests, but because I was scratching for the funds, and had other plans down the road and only wanted to do the upgrade once.)
For the last few weeks I have been driving in all sorts of weather (including -19 temps) with 100% E85 and now recently a 50/50 mix of E85 and gasoline.
Here's the magic combination!
All you need to do on a Subaru WRX (2002 -- I assume others will behave the same), is increase your fuel low by 15% over stock. This will make the ECU have zero LTFT's and STFT's on a 50/50 mix of gasoline and E85.
If you plan to run mostly on E85 you can set the neutral point a bit higher, if you expect to need to spend a good fraction of your time on pure gasoline set the neutral point a bit below the 50/50 fuel blend.
The ECU will use 15% of its fuel trim authority, going positive as you move toward 100% E85 and going negative as you move toward 100% gasoline.
I have spent the last couple weeks driving around on -15% fuel trims and the car runs just fine.
The easiest way to add 15% to the fuel flow, is to up the fuel pressure from 43 psi to 57 psi.
This has two minor down sides. If you have an upgraded turbo (and fuel pump) you will run out of injector at high load (not good --- don't go there). On the stock turbo you will also run out of fuel pressure and flow from the stock fuel pump at high boost (need to add a high flow fuel pump if you run over stock boost levels).
The slightly more expensive way to do it, if your interested in performance, is to install larger fuel injectors, an upgraded fuel pump and either adjust fuel pressure or re-scale the injectors with one of the aftermarket tuning solutions.
This is what I did, I have an Aeromotive adjustable fuel pressure regulator, 800 cc/min decapitated injectors and an ECUtek reflash. Harvey at Super Rupair in Boulder has been worth his weight in gold helping me get the reflash tweeked properly to prove to my self I am on the right track. He's made nearly daily adjustments to the map as I tried out different things.
Thanks Harvey !!! (and all the guys at Super Rupair who have worked with me on this experiment).
More tuning will follow as I work out the final setup I am looking for, but I wanted to get the word out so some of the other E85 experimenters could do some peer review of my concept and prove if it is a safe and reasonable solution for the average user.
Larry |
|
| Back to top |
|
 |
stylin99

Joined: 08 Sep 2006 Posts: 85 Location: Charlotte, NC
|
Posted: Fri Feb 16, 2007 1:45 pm Post subject: |
|
|
Very interesting stuff! It's a great concept.
That theory sounds great for closed loop, part throttle operation, but I'm not seeing the other end of the picture.
What are you doing for open loop fuel requirements? Do you have a standard fuel trim built into the ECU for open loop full throttle runs? _________________ '99 Mustang GT, Vortech Supercharged, now running E85 full time.
www.e85mustangs.com
Best Quarter Mile on E85:
10.52 at 135 mph |
|
| Back to top |
|
 |
hotrod
Joined: 19 Apr 2005 Posts: 872 Location: Colorado
|
Posted: Fri Feb 16, 2007 7:12 pm Post subject: |
|
|
FFV test#2 completed
sudden transition from 50/50 mix of gasoline and E85 to a tank of 100% gasoline.
I ran the car on a 50/50 mix of E85 and gasoline for a bit more than 2 full tanks (600 miles). I tweaked the injector scaling so that on that 50/50 mix the LTFT's and the STFT's approached zero ( By that I mean the sum of them was zero +/- 5%).
I then ran the tank down to about 1/2 -1 gallon into the reserve (fuel fill light on for 10-15 miles) and then filled the tank with 91 octane Premium gasoline.
Upon starting STFT's immediately (within 10-15 seconds) had scrolled up to -25% then the LTFT started to decrease.
LTFT's and STFT's settled in around the near zero area but were slightly higher than I expected 0 ===> +8% on occasion.
The good news is that the switch to gasoline is a graceful failure mode, in that if the ECU does not immediately apply full correction, the mixture goes rich.
Watching the Wide Band, cruise mixtures still settled in at Lambda 0.99 - 1.00 but you could see it was bouncing around more before it settled.
About 10 miles after I put the new fuel in the tank I did a WOT blast and saw mixtures on the Wide Band of 0.69 ( 10.143 AFR ) --- to ---- 0.72 (10.584 AFR) These are essentially the stock AFR's !!!!!
Now I will watch things and see how the fuel trims settle in. There were no drivability issues with the sudden switch of fuels other than an occasional pop from the tail pipe.
I'm going to let the car cool a bit and run it through a couple more driving cycles here in a bit.
Stay tuned for more of " How the ECU Turns "
( I know more controlled tests would be useful but I am trying to get a first pass "coarse evaluation" before I start confirming the fine details. Next I will run this tank down to the fill light and throw in some straight E85 and see what happens.)
Larry
==============
On my second driving cycle (15 - 32 miles) the LTFT's hung around -11.8% ===> -14.1%.
I just finished my 3rd driving cycle since the change in fuels. Total of 55 miles.
In light throttle cruise it is showing a LTFT that varies from -14.1% ==> -15.7%, so it is closing in on the predicted -15% number.
Does anyone know what the ECU defines as a "driving cycle" ? Does the engine simply need to be shut off, or does it need to cool enough to go through warm up enrichment steps.
I have heard that the ECU will be fully adapted to most changes in 4 driving cycles.
===============
In the for what its worth department ------ the way the LTFT's are acting makes me suspect that the ECU may keep a moving average of the LTFT somewhere for certain conditions and use that default value the next time the engine is started.
The reason I speculate this, is on the first driving cycle most of the adjustment was made in the STFT and the LTFT tried to stay near its historical average prior to the fuel change. After shutting the car off, when it was restarted for the second driving cycle the LTFT immediately jumped to a value that was near the average adjustment it needed on the first driving cycle until it warmed up, then the LTFT started moving around again, and slowly got more negative as I drove.
After shutting the engine off and starting the 3rd driving cycle the LTFT did the same thing, it immedialtely went to a -15.7 value and stayed there until the engine warmed up, then both the STFT and LTFT moved around a bit as the LTFT gradually settled near -14.1% most of the time.
We will see what it does on the 4th driving cycle, after the engine has some time to cool off.
Larry |
|
| Back to top |
|
 |
hotrod
Joined: 19 Apr 2005 Posts: 872 Location: Colorado
|
Posted: Thu Feb 22, 2007 4:33 pm Post subject: recipe for the setup --- FFV is working |
|
|
Okay I finished enough testing to post a recipe for a home built WRX FFV.
I have satisfied myself that the following concept is sound as the car is behaving as I expect, and so are the fuel trims.
You can run this in two different modes. You can let the fuel trim authority of the ECU do all the work (we will call this "autotuning mode"), or you can make life easier on the ECU by a minor tweak of the fuel pressure to help it get in the ball park quicker -- (we will call this "user assisted tuning mode").
Autotuning works just fine, but takes about 4 driving cycles for the ECU to settle into final LTFT's. It is a bit less than optimal tuning at the ragged edge so if your tune is a max power tune I would not recommend this. For a happy daily driver at moderate boost levels (<17 psi) I think it will be safe unless there is something amiss with the base tune on the ECU.
Step 1 -- Upgrade fuel pump to a walbro GSS 342 255 l/h pump or equivalent.
Upgrade fuel injectors to have at least 30% more max flow than stock.
Install an adjustable fuel pressure regulator and an under hood fuel gauge.
For autotune mode. Set the fuel pressure at 45 psi and get the ECU reflashed with an Ecutec or equivalent reflash with a 50/50 mix of the gasoline in the tank. Use the fuel you normally have available and summer blend 85 if available (winter blend E85 is about 70% ethanol so if tuned with the winter blend, richen the mixture a tad in the wide open throttle areas.)
On the 50/50 mix of gasoline and E85 at 45psi fuel pressure you should have the flash setup so you get LTFT's that settle down near zero or slighly negative ( 0 to -5% ).
If you simply add 100% gasoline to the tank, the ECU will auto tune to -10% to -15% fuel trims. You will get normal stoich AFR's while in closed loop but your WOT AFR's will be a bit rich (around 10:1).
If you simply add 100% E85 to the car and let it auto tune, you will get normal stoich AFR's in closed loop fueling and your open loop fueling will be a bit leaner than the tune at 50/50 fuel mix by about 7%.
==========
For user assisted auto tune mode, you simply adjust the fuel pressure slightly when you make a major change in the fuel in the tank.
35 psi for 100% gasoline
45 psi for 50/50 gasoline and E85
55 psi for 100% E85.
If you adjust the fuel pressure as above, you will keep the ECU near 0 LTFT's and your WOT AFR's will be only slightly effected by the fuel switch.
The adjustment of fuel pressure does not need to be precise. Simply make a ball park guess what the blend in the tank is, and you will be close enough unless your running a high boost system that needs the full octane of the E85. In that case, I assume you have access to some sort of logging ability and or a wideband so you can tune accordingly or have a dedicated 100% E85 max performance map.
=============
My setup (essential mods for the FFV conversion)
Aeromotive adjustable fuel pressure regulator.
800 cc / min decapitated OEM injectors.
Walbro GSS 342 255 l/h fuel pump.
Ecutec reflash -- tuned at 50/50 mix of gasoline and E85 with -5% LTFT's after is settles in (more than 4 driving cycles). WOT wide band is showing 0.82 - 0.84 lambda on the 50/50 gasoline, E85 fuel mix.
(non-essentail mods)
upgraded fuel system under the hood (fuel rails and new braided fuel lines)
More details to follow as I get more miles on the setup and get better info.
50/50 gasoline - E85 mix giving me 24 mpg, same milage on 100% gasoline. I will update the fuel milage on 100% E85 when I finish this fuel tank, but am guessing it will turn out to be about 22 mpg.
Larry |
|
| Back to top |
|
 |
matt621
Joined: 01 Apr 2007 Posts: 4
|
Posted: Sun Apr 01, 2007 4:51 am Post subject: |
|
|
What do you think of using the Kenne Bell Boost a pump?
http://www.kennebell.net/accessories/boostapump/boostapump.htm
One reason I like it is that walbro say their pump is not rated for alcohol. Yes, I know it's just a cya, but the KB solution seems pretty simple. |
|
| Back to top |
|
 |
hotrod
Joined: 19 Apr 2005 Posts: 872 Location: Colorado
|
Posted: Sun Apr 01, 2007 5:53 pm Post subject: |
|
|
It might help some folks, it is just a voltage stabilizer, that ensures the pump gets its rated supply voltage. On some cars it may be problimatic to set it up.
On the Subaru WRX for example it uses a fuel pump controller that varies the voltage supply to the pump as standard equipment. It uses pulse width modulated DC to vary the supply voltage to the pump at 33% 66% and 100% of th system voltage. Some users of some of the early aftermarket tuning solutions had problems because the "chips" used too high of a duty cycle on the fuel pump power supply and it would over heat the fuel pump and make it cut out due to a thermal protection switch in the pump. The result was that under certain cruise conditions where the pump was run continuously at the high voltage duty cycle they would stall for no reason and not be able to move the car until the pump cooled. On a couple of occassions that resulted in the cars stalling at bad times like during left turns in heavy traffic.
For something like drag racing where you want maximum pump out put it would work but there may be unintended side effects for some users.
Larry |
|
| Back to top |
|
 |
matt621
Joined: 01 Apr 2007 Posts: 4
|
Posted: Sun Apr 01, 2007 7:22 pm Post subject: |
|
|
| gotcha. thanks for the info. so the walbro pump and new injectors is the only way to go, right? Do you have to modify the car in any way to put the walbro pump in? Or do they have a kit that adapts it to a subaru? |
|
| Back to top |
|
 |
hotrod
Joined: 19 Apr 2005 Posts: 872 Location: Colorado
|
Posted: Sun Apr 01, 2007 11:05 pm Post subject: |
|
|
On the WRX the walbro 255 l/hr pump is a drop in replacement for the OEM pump.
Larry |
|
| Back to top |
|
 |
matt621
Joined: 01 Apr 2007 Posts: 4
|
Posted: Sun Apr 01, 2007 11:41 pm Post subject: |
|
|
| what about the sti? same? |
|
| Back to top |
|
 |
mtbottle
Joined: 14 Jan 2006 Posts: 46 Location: West Virgnia
|
Posted: Fri Apr 06, 2007 11:15 pm Post subject: good stuff |
|
|
Larry, You are absolutely brilliant! Maybe you should go to Detroit and show those folks how to build an FFV the right way. _________________ Duane Combs |
|
| Back to top |
|
 |
Revision
Joined: 14 Sep 2005 Posts: 194 Location: Carol Stream, IL
|
Posted: Sat Apr 07, 2007 11:43 pm Post subject: |
|
|
Our friends at AMS are testing various fuel pumps and have come up with some interesting observations.
This directly relates to anyone who wants to use a fuel pressure regulator to adapt for E85.
See Link for Original Thread: http://forums.nasioc.com/forums/showthread.php?t=1237517
AMS's results.. in a nutshell
| Quote: | | The in-tank 255lph HP Walbro pumps have a pressure relief valve that starts to dump flow at roughly 85psi and the output flow drops. |
So if you are approaching 85psi on a Walbro 255lph pump then you will start losing fuel pressure.
For Reference, also from AMS's results:
| Quote: | | The EVO VIII pump has a pressure relief valve that opens up at around 65-68psi and the flow output drops dramatically after that, eventually going to zero flow around 75-80ps |
So keep an eye on that fuel pressure!  |
|
| Back to top |
|
 |
hotrod
Joined: 19 Apr 2005 Posts: 872 Location: Colorado
|
Posted: Sun Apr 08, 2007 12:01 am Post subject: |
|
|
I really don't see that as being a big problem if your running the stock base fuel pressure of 43 psi then the pump relief valve is irrelevent up until you hit manifold boost pressures of 37 psi or so.
Not many folks play up in that range, and even if they do its only for a couple seconds at a time.
Even at a base fuel pressure of 55-60 psi you can still run manifold boost pressures up in the low 20 psi range before it becomes an issue.
Clearly fuel flow is worth monitoring, as well as fuel pressure, and it is really good to have the objective data.
Thanks to AMS for going to the trouble of running the tests and being willing to share it with the community.
mtbottle:
Thanks for the compliment!
It is really just the result of some methodical testing.
By the way to all concerned, after several full tanks of E85 on the setup I am still getting the behavior I expected, with my idle fuel trims stabilized at +14.8% on 100% E85 and 55 psi base fuel pressure.
Fuel milage is a bit lower than I wanted at high mid 18 mpg - low 20 mpg depending on how I drive. This may have more to do with a heavy foot rather than a solid comment on the setup.
I will be dialing back the fuel pressure a bit and see if that helps the fuel milage by leaning out the open loop a bit.
Larry |
|
| Back to top |
|
 |
hunterk2001
Joined: 13 Apr 2007 Posts: 2
|
Posted: Fri Apr 13, 2007 8:35 am Post subject: |
|
|
This sounds really interesting to me and am seriously thinking about converting my truck to run on E85, but I'm a little confused. From what I've read, doesn't the ECU try to maintain a stoich of 14.7 in closed loop? If so, doesn't E85 require a stoich of around 10.6?
So I guess what I'm confused about is when you are running straight E85, your ECU is adjusting fuel trims for a stoich of 14.7 in closed loop, so wouldn't that be really lean? What does your WB show in closed loop driving?
I'm not saying what you are doing is not working, I'm just trying to understand it. |
|
| Back to top |
|
 |
hotrod
Joined: 19 Apr 2005 Posts: 872 Location: Colorado
|
Posted: Fri Apr 13, 2007 11:04 pm Post subject: |
|
|
| Quote: | | So I guess what I'm confused about is when you are running straight E85, your ECU is adjusting fuel trims for a stoich of 14.7 in closed loop, so wouldn't that be really lean? What does your WB show in closed loop driving? |
It is not a problem. The O2 sensor and the close loop fuel trims do not try to maintain a 14.7:1 air fuel ratio, they try to maintain a lambda of 1 which is stoichometric for any fuel. The 14.7 is the AFR that happens to reach lambda 1 on common gasoline. On E85 or any other fuel blend the AFR will be different but the ECU does not care what the true AFR is only the fact that the O2 sensor is showing near zero free oxygen in the exhaust gasses. If it has enough fuel trim it will hit what ever AFR is required for any fuel mix you can get the engine to run on.
Your question is very common, the concept is actually so simple people tend to over complicate it in discussion because most of our tuning references use AFR numbers rather than lambda.
In closed loop my wide band always shows exactly the same information regardless of fuel mix! It is usually between lambda 0.99 and 1.02 in cruise.
On most wideband systems the displayed AFR assumes you are using gasoline so the displayed AFR is typically the equivalent gasoline AFR for the real exhaust gas readings. On my wideband it displays both lambda and AFR on the same screen. I usually completely ignore the displayed AFR. --- as I recall it usually is in the 14.6 - 15.1 range on my meter.
Larry
Last edited by hotrod on Mon Apr 16, 2007 3:34 am; edited 1 time in total |
|
| Back to top |
|
 |
hunterk2001
Joined: 13 Apr 2007 Posts: 2
|
Posted: Sat Apr 14, 2007 10:07 am Post subject: |
|
|
DOH!
Makes absolute perfect sense, can't believe I missed it. Like you said, all my tuning references were based on AFR numbers.
Thanks |
|
| Back to top |
|
 |
|