"base ethanol%" setting issue

[piledriver]

My current baseline blend is set at 10%, ~7% actual according to sensor
VEAL run tuned everything off target richer.
Was tuning vetable1.

Ve table was previously giving me ~exact targeted AFR even in open loop.

Could be malpractice on my part, first time i have tried it.

ms3-veal-e85blendissues.zip

Edited subject as its not really a TS issue.

Mods: Perhaps move to MS3 devel?

[piledriver]

Loaded a tune from last April and the "off target by 1" problem went away, still trying to find the exact setting that hoses it up.

[piledriver]

The offending setting appears to be "baseline ethanol%"

My local fuel has been "E10" for ages, so my baseline vetable is on ~6-8% ethanol (according to the sensor).

Unfortunately setting that makes VEAL tune to the wrong AFR, and the base tune becomes ~worthless.
Setting it to zero everything is peachy...,\Could be malpractice on my part.

When is James coming back? Perhaps a better explanation of what that is supposed to be for would help.

[LT401Vette]

There are a number of problems with VEAL and table blending...
VEAL doesn't work any differently while table blending is on and isn't really aware how much input is coming from each of the tables or which table is wrong.

For it to work really you need to have it all running to 1 condition and tune the one table, then all the other table condition and let it tune.

[jsmcortina]

When is James coming back?

Where did I go?

James

[piledriver]

When is James coming back?

Where did I go?

James

Thought you were traveling for awhile.

In any case, setting "baseline ethanol%" to 8% makes VEAL pull the AFR one full point off target.
Also causes the existing table (learned on E10) to run lean by the same amount.

Tooltip indicates it allows you to enable the blends w/o a retune but it doesn't seem to work in a obvious way.

What am I doing wrong?

[jsmcortina]

In any case, setting "baseline ethanol%" to 8% makes VEAL pull the AFR one full point off target.
Also causes the existing table (learned on E10) to run lean by the same amount.

Learned E10 with Flex enabled or disabled?

If the fuel table was tuned for E10 with Flex disabled, then enabling it and setting the baseline to 10% should give you the same fuelling. The setting is saying "my base fuel table is for 10% (or 8% or whatever.)

With flex enable and the baseline setting at zero (the original way), it will be assuming the fuel table is for 0% ethanol and making a correction (adding fuel) when you actually have E10.

James

[piledriver]

Learned disabled, works perfect set to zero or with flex disabled

[jsmcortina]

Learned disabled, works perfect set to zero or with flex disabled

That's rather terse, can you explain in more detail with MSQs and datalogs that illustrate it.

James

[piledriver]

Basically it runs ~exactly on AFR target with flex disabled or enabled with baseline ethanol set to zero.

Set baseline ethanol to ~8% (typical fuel here. and what VEAL tuned on before I had the sensor) and it goes all pear shaped.
Veal tunes vetable 1 full point below setpoint, and the fueling is calculated wrong. (lean) by about the same.

I'm under 10% ethanol, so it shouldn't (and isn't) doing any correction based on the blend tables, they start at 10%.

The only thing that does not work as I understand is the baseline ethanol, which seems to break things when set to other than zero.

The rest of the blending appears to work well AFAICT, although I have only been able to score E70 around here so far.

Will grab a short log and re upload, will change that one setting mid-log.

[piledriver]

Sorry, work got in the way.

Heres a breif log/msq at idle: starts with baseline ethanol% set at 0, then changed to 7%, then back to zero.
i was surprised it even idled at ~19:1 afr, but i digress.(it was warmed up a bit)

What is this setting for/supposed to do, and how is it meant to be used?
AFAICT the tooltip info doesn't match its operation, or I'm not grokking what its supposed to be for.

ms3-veal-e85blendissues-2.zip

[piledriver]

I currently have an MS2 cpu running my MS3/MS3x box, and have been running E85 (well, ~E65 actually, can't find much better) and I tried an experiment on MS2e and at least the normal flex fuel setup does work well AFAICT.

MS2e seems to lacks all the lovely table blending MS3 has, but it does support the sensor etc.
I ran VEAL on ~E65 and got things well tuned.

set baseline ethanol to 10%, as thats where it was (should it have been set to 65%?)
.
I don't have table swicthing etc on, just enabled the sensor etc with the defaults)

Filled up on gasoline (E10) and it went from E65 to ~E20 and it calculated the fueling quite well.

I'm still not sure how the baseline ethanol setting is supposed to work//be used, but the basic flex fuel functionality seems to work as expected using a single VE table on MS2E.

[jsmcortina]

The baseline number is supposed to match the Exx number you were using if you tuned your fuel and spark tables with Flex turned OFF. If you then turn Flex ON, the fuel and spark should remain the same.

Without the "baseline" feature, the code seemed to assume that the base tune was for 0% ethanol (which isn't likely in the USA) so even with regular fuel, the tune would have been immediately different.

James

[piledriver]

The baseline number is supposed to match the Exx number you were using if you tuned your fuel and spark tables with Flex turned OFF. If you then turn Flex ON, the fuel and spark should remain the same.

Without the "baseline" feature, the code seemed to assume that the base tune was for 0% ethanol (which isn't likely in the USA) so even with regular fuel, the tune would have been immediately different.

James

That's pretty much how I assumed it should work, but it didn't seem to want to cooperate.
(see first post etc)
Seems to work OK if I use it wrong though.