acedeuce802

iTrader: (1)

Just talked with a few guys about this. The best option would be to experiment on a dyno at a few different fuel mixtures, obviously, because there is a lot to take into account. Too much to estimate the proper blend curves. There are some general trends to follow that will get a solid tune at least.

Assuming the same lambda is kept between E10 and E85, the chart I posted above would be a pretty good curve. E85 responds to AFR for knock resistance more drastically than E10, so enriching a bit can net an effective RON of up to 130. It's common to run a blend (say E30 or E40), run it a little richer than typical, and run spark values similar to a typical E85 setup.

E85 burns quicker than E10, so MBT can actually be slightly retarded from E10. So, vacuum and low boost where you're not knock limited may actually benefit from few degrees less timing. That's for a given lambda, though. E85 can be run leaner due to it's cooling and knock resistance, and leaner mixtures desire more advanced timing. Luckily lower load stuff can be easier tuned on the road without blowing stuff up, so that may be an experiment to see what E85 needs after leaning out in cruise and low load.

Schuyler

iTrader: (1)

Does anyone have any experience with how their tune changes from summer to winter E85? My tune random went crazy lean recently, and has stayed that way, and I'm kind of curious if this is to blame.

18psi

iTrader: (76)

yes, as ethanol content goes up, you will lean out. as it goes down, you'll richen up.

a few times I've seen it jump a full 1.7AFR

your station probably switched to the proper blend now.

pdexta

iTrader: (17)

Great find, thanks for posting. I really didn't like the single blend table for fuel and spark, but I thought that was the only way to do it.

pdexta

iTrader: (17)

I'm having an issue with the multiple blend tables that I can't figure out and wanted to see if you guys had ideas. My car runs great when I turn "Dual Fuel" on and let it use "Flex blend" to blend VE and Spark tables 1 and 3 at the same rate based on my flex sensor.

If I disable Dual Fuel and use Table Switching so I can create 2 different blend maps for Fuel and Spark it doesn't blend at all and just uses VE and Spark map 1. My flex sensor is reading about 80% ethanol now, runs great on Dual Fuel, I switch to Table switching and it goes off the wideband lean. These are the settings that don't seem to blend at all:




Is there something I'm overlooking or did I screw up something?

aidandj

iTrader: (5)

What is the blend table x axis

aidandj

iTrader: (5)

Also post log.

pdexta

iTrader: (17)

XAxis is set to "Flex%" on both tables.

Took a quick log at idle of both tunes, Dual Fuel (18-51-44) then Table Switching (18-53-02).

aidandj

iTrader: (5)

Somewhat stumped. I would post on msextra. What does Flex% read when running (can you make a guage for it?) and also where does the indicator sit on the tables when running.

aidandj

iTrader: (5)

I started a thread about it over here. Megasquirt Support Forum (MSEXTRA) ? Flex% same as Ethanol percentage? (View topic)

Mind if i post up your logs?

hector

Alright bringing this back from the dead to see if you have had any results you would like to share.

I am currently running e85 in my car for the first time and it will be using the flex function as I don't plan on running e85 all the time. The plan is to tune it with available e85 but run on e10 all the time except at autox events where I will plan ahead and have only 1-2 gallons of gas when I arrive at the event. Empty out the remaining e10 fuel and put in e85 from a container. After event, put in all remaining gas and fill up with e10. I would love to get a drum of e100 or whatever so I can have consistent alcohol content but am really hesitant due to the minimal use I require and the hygroscopic nature of the fuel. BTW, car is NA VVT with MS3Pro.

Anyhow I was having a hard time understanding how to tune the VE3 table but posts 48 and 50 in this thread straightened me out. I haven't done any tuning as of yet. Just created a VE3 table from VE1 with 30% more fuel across the board. Set the table switching to blend curve 3 and let the ecu do its thing. I have attached a log of a quick drive around the block. I know you guys have upgraded everything but I'm living dangerously and just left everything I had before. Stock pump and fuel lines, 20k? mile fuel filter, NB FPR on injector rail with return, and factory 01 injectors. Duty cycle was up to 90% at 7800rpm but it appears to be quite rich. Of course this is without any tuning so who knows. And with the screens on the air horns which has proven to make the mixture rich on datalogs by about .5AFR.

Stupid question, would getting the spark table right affect the AFR reading? I ask since I haven't added any spark advance yet.

Also, the data shows an e85 fuel correction of ~146. Does this mean its blending VE1 and VE3 or is it just doing the flex blend which you have no control of? I don't have the flex blend on.

And the ethanol percentage varied from 68 to 75 on the datalog. Is that normal? And it hit 75% when I was doing the pull to redline. Like it was following it. Is that weird or is there an explanation for that?

Thoughts?

Savington

iTrader: (31)

You don't need to do this. MS3 will add a fuel multiplier to Req.Fuel, so your VE1 table as-is is a good starting point for VE3. You will need to lean everything out, low load more than high load, so start over on VE3.

This means it's adding a 46% fuel multiplier. The blending is a different setting

Mine is very consistent. It should not vary at all under load.

hector

Thanks for the reply. I find it odd that the ethanol content varied. Must be a ground/noise issue.

hector

Ok, I thought I had a good understanding of what to do but not really. This is my new thinking and it might be right or not, and a revelation or just regurgitation.

As mentioned before in this thread, what is trying to be accomplished is to have two fuel and spark maps. One map for e10 and one for e85 and have the ms3 blend the two maps depending on ethanol content. This is the more satisfactory means of using the flex sensor. You can just let the ms3 do its own gross compensation and extrapolation of fuel and timing but it will simply take the info from a min and max addition table and extrapolate by ethanol content. This will work good enough and if all you want is the flexibility to run between the two fuels, that's fine. But if you want to extract all the power potential out of e85 and still be able to be safe switching/blending the two fuels then you have to go in another direction.

So in order to do that you have to disable the auto correction of the ecu by setting the table for fuel multiplier/timing addition to 0 or 100% which means no additional correction by the ms3. Then make a VE3 map and Spark 3 map for e85. Tune on e85 until its perfect. then set the ms3 to do table switching on the blend curves. Lastly, tune the blend curves with different ethanol content which is probably the most difficult/tedious part.

Here's what my settings were that were causing my fuel to be always rich no matter what I did. Had I kept tuning VE3 like this I would have ended up with a VE table for e85 that would be leaner than the e10 VE1 table. That's because the ms3 was adding 46% more fuel to whatever table it would use. Hence, to run e85, you don't even need to make a separate VE table if you don't want to. Just let the ms3 add a predetermined amount of additional fuel. Yes, I guess you could tune the VE3 table to be less than the VE1 table and have the blend function do its thing with the predetermined additional fuel but that just seems counterintuitive.




And here is a screenshot of my VE3 as just a straight copy of VE1 and you can see how rich it is.



I have not tried out anything as I came to this conclusion yesterday morning which was Fathers Day and I was planning on spending the entire day with my family enjoying food, drink, the pool, and the company.

So, anyone have any thoughts?

Savington

iTrader: (31)

To me, turning off the fundamental basis of flex fuel (a multiplier based on ethanol content) and then tuning it all back in yourself using VE3 is counterintuitive.

Let the MS3 do its thing by adding a global multiplier based on ethanol content. From there, you can trim VE3 as needed to ensure that lambda values stay where they should be. Yes, this will mean that most of your VE3 map is "leaner" than your VE1 map.

VE1 is sort of a similar "trim" map - there is a calculated fuel flow based on RPM and injector size, and then you "trim" with varying VE% figures in the fuel map. With the combination of the standard flex fuel correction and VE3, you basically do the same thing.

Savington

iTrader: (31)

Here's how I would set up a fresh MS3 flex fuel install.

1. Dial everything on gasoline and save your tune.
2. Install the flex fuel sensor and set Baseline Ethanol% to the value you get with "100% gas" (in CA this was ~8% for me)
3. Turn on flex fuel, but leave all the alternate maps off. This just leaves the global blend, but doesn't allow blending between VE1/VE3 (or Spark 1/Spark3, etc)
4. Drain the tank and fill it with E85. Tune the spark/fuel, AFR target, ASE/WUE/Accel maps, and all the other maps that have switching available in the software. Tuning all these maps on VE1 will let you use Autotune like you normally would. Put a few tanks of raw E85 through the car to get the content up as high as you would ever see it, and make sure the tune is dialed in.
5. Load the set of E85 maps into the alternate map positions, load your gas maps from step 1 into the primary positions, and turn table switching on. Set the flex blend curve so that the ECU switches over to the alternate maps by 80%EC
6. At this point, you're about 95% of the way there. The only setting left to dial in is the flex blending map, and that will happen by trial and error with varying levels of ethanol content in the fuel.

aidandj

iTrader: (5)

Spoke with a local MR2 tuner today (was at his dyno). He said he had the occasional signal noise issue. Until he started using shielded wire for the signal. Grounded the shield at the ECU, and it made quite a big difference. Definitely worth trying if you are getting signal noise.

hector

Had some time this weekend to play with the car. Did steps 1-4 as Sav recommended but am only concentrating on fuel so far. Will have to start spark tuning soon and then go back to fuel I imagine. The other settings shouldn't affect base fuel/spark so good enough on those is good enough.

BTW, how critical is the change in desired AFR? Is it like a whole lot different or just slightly different? I know, search search search, but you could throw me a bone here. It seems to run better lean than on gasoline.

And the flex signal is running through the factory o2 wire back to the ecu so it is shielded but not to the ecu. The shielded end is at the sensor side. The signal is rising with RPM by about 3% ethanol content. I have some work to do on that.

18psi

iTrader: (76)

You pretty much want 12afr at high load on e85

acedeuce802

iTrader: (1)

You could run the same AFR map as gasoline, but you can lean it out a bit at high load. You can safely run a lot leaner at low load. I'm cruising and idling at 1.25 lambda without any issue right now. I haven't checked MPG yet to see if it's actually worth it, but it doesn't feel like I need more throttle input than normal.