Non-Linear Flex Fuel Ignition Timing
 

After a fair bit of research, I was surprised to discover an interesting fact about ethanol mixes: knock resistance (partially octane but also other factors) does not increase linearly as ethanol concentration increases. Instead, knock resistance rapidly increases up to about 50% ethanol content and gains are only marginal past that point.

With this in mind, isn't it reasonable to assume that a blend curve unique to timing may be advantageous for ethanol flex fuel applications? That is, being able to ramp up to full timing by 50% ethanol content while keeping a separate blend curve for all of the fueling changes. I currently have an Rev MS3 Basic and I am almost certain it doesn't support something like this. In fact, I am unsure of whether there are any PnP ECUs for Miata's that support this. Is this due to some oversight on my part or are the advantages too negligible to matter in most cases?

Another note -- I live in Iowa. E85 is abundant but actual content varies wildly based on the season. Being able to run peak timing advance on everything from 50% to 85% ethanol would be a major bonus for me. Folks in California and other states with more consistent ethanol content may not need to worry about this at all.

Blended tables based on ethanol content is very common and a standard feature on MS3PRO ECUs. I would be surpised if it wasn't supported on all MS3 variants, as I believe they all use the same code base.

Edit: obviously REV has his own FW, most likely uses the 1.4 code base. Not sure when blended tables was added to the source code.

MS3 certainly supports it, but there appears to only be a single “global” blend table for VE, Timing, enrichments, boost, etc, as opposed to separate tables. Perhaps I am just missing something here?

8 blend curves available in 1.5.2. You can choose to blend with flex linearly, using a flex curve, individually with flex % as the independent variable, or a selection of other independent variables. For example, I use 4 VE tables, RPM switched at 3250 RPM and flex blended with the following curve

https://cimg5.ibsrv.net/gimg/www.mia...f352ec1fff.jpg

Ignition advance OTOH is fully table 1 by 63% eth and dictated by this curve

https://cimg5.ibsrv.net/gimg/www.mia...855e459336.jpg

FANTASTIC info, just had t trouble getting all of the settings correct for my MS3. For future readers, the "table choices" section is more complicated than it appears. When doing duel fuel, you can choose "flex blend" which blends ALL tables via a single curve by default. However, disabling the "Alt. XXX" field will allow you to select a separate blend table for VE, spark, AFR, etc.

A few extra questions:

1. How critical is it to have different blend tables for fueling? Is it reasonable to use a single blend table for VE, accel enrich, warmup enrich, after-start enrich, and cranking?
2. Is a blend table for AFR really necessary with a lambda sensor? From my understanding, most sensors output raw lambda and ECUs translate that to AFR based on whatever stoich you are using. If that is the case, is there an advantage to running different lambdas for ethanol vs gasoline?
3. How do you tune these things? My current plan is to get a tank as close to 85% ethanol as possible, tune the secondary tables, and let it rip. Is that safe?

1 - as you can see I use global blend for fuel related things, so that should communicate my thoughts there. I use tables 1 for eth because the startup curves for 2 are jacked and starting on eth is way more sensitive than gasoline. The jacked tables 2 are still adequate for startup with gasoline under all conditions.
2 - more power is available with a slightly leaner mixture (12.5 or so using lambda =1 @ 14.7 reference scale) which can be safe with more advance at higher octane. I don't want to run that lean on gasoline. Personal preference.
3 - all you man, it took a very very long time for me to tune this car. I did several passes both up and down on the eth scale over a period of many months.