14.08 is only 12% from 12.55 but 9.85 to 12.55 is 27% from 9.85 vs 32% with 13.08.
14.08 is only 12% from 12.55 but 9.85 to 12.55 is 27% from 9.85 vs 32% with 13.08.
I don't follow how the trims could ever be that large with flex fuel enabled and functioning correctly. STFTs may be high during the learning process, especially for a major fuel change but should be normal once learning completes.Originally Posted by metroplex
Looking at one of my logs for a full tank of "E85" after a flash/KAM reset, the flex fuel logic starts at the Stoich AFR Init value of 14.08 (9.8% alc in log). This resulted in an STFT of 38% at startup but quickly went down to 10% and less within 1-2 minutes with no codes stored. Learning completed in about 11 minutes with matured alcohol value at 76.5%, STFTs 0-4%.
My understanding is the flex fuel logic uses the alcohol learning process to infer the stoich to use from the mature point to the next refuel. Based on this inferred value found during the alcohol learn, flex fuel logic determines the alcohol percentage according to the Ethanol % vs AFR table.
For the values I have on my flex fuel tab, alcohol percentage in the fuel corresponding to an AFR numerically above 10.69 (E70) and below 13.08 (E28) should result in the computer using that learned AFR for stoich. STFT/LTFTs should be theoretically zero for any value that has matured within that range.
Between 9.85 and 10.69: Stoich will be 9.85; fuel trims 9% worst case.
Between 13.08 and 14.08: Stoich will be 14.08; fuel trims under 8% worst case.
I'm just thinking the 13.08 works for Ford's OEM setting of 10.26 as the lowest stoich (27% delta). But if you lower that floor to 9.85, then 13.08 becomes 32.5% delta. So if you're at 12.99 or something then it won't lock to the 14.08 end point because it hasn't gone above 13.08 and will stay with the 10.26 end point. That's just how I am interpreting the descriptions for those 2 settings.
This of course still doesn't resolve why my LTFTs aren't learning. HPTuners couldn't help, so I'm back to square one.
Flex Fuel Diagram.JPG
I think this explains what I am picturing in my head a lot better than going back and forth in text form.
So if we kept the 13.08 High AFR range, it would be stretching that threshold for the fuel trims using a 9.85 low end point vice the OEM 10.26 setting. That's why I was thinking 12.5-12.55 would be better as it'd keep the thresholds within 25% or so.
I loaded the new channels and started logging (it was raining non-stop for the past 2 days with 6 inches of rain in the area). It's Actively Learning (status) and Refueling (source).
With the new stoich range increased to 9.85, I see my stoich while driving is 9.80 and inferred alcohol is 84.7% - however when I am idling, it trends back towards 10.60 and 70.2%
Still no LTFT data.
I'm not interpreting High and Low Range with respect to fuel trims like you are. The definitions in my VCM editor read:
[ECM] 21349 - Flex Fuel High AFR Range: Above this inferred AFR, the stoich AFR can be locked to the high AFR endpoint.
[ECM] 21352 - Flex Fuel Low AFR Range: Below this inferred AFR, the stoich AFR can be locked to the low AFR endpoint.
Flex fuel logic seems to learn between these range values and will mature to an inferred stoich value between them, or lock to one of the endpoints if the inferred stoich reaches one of the range values. I don't think the High AFR Range will come into play until you fill up with E10 again.
We had the rain last weekend and earlier this week. Nothing but sunshine now.
Now if that Actively Learning would just switch to Matured, you will have LTFTs.
It's good you're seeing the Actively Learning status. When my Mustang is Actively Learning, I tend to drive it trying to keep the Mass Airflow between Airmass Min and Airmass Max for at least 10 seconds at a time (Min Time to Infer) and do this several times over a 10-15 minute period to reach the Minimum EGO Switches required to infer. Really, just normal driving should accomplish this.
You probably still need the maximum update time parameter set to .01 to allow it to mature and see LTFTs:
[ECM] 21364 - Flex Fuel Update Rate Limit (IPC): This is the rate at which flex fuel may update for use with the plausibility checker. This is measured in AFR/s.
If I'm correct, learning is taking place but your car is stuck learning without recording/manipulating the values to establish a final answer for that tank of fuel.
That missing parameter is certainly possible but I just learned that HPTuners is down to 1 engineer (Eric) so it may be months before he is able to get to it.
it would not hurt to go back to step 1. Turn off flex fuel logic once u have 100% 91 fuel so u can adjust the MAF vs period and re adj fuel trims as close to zero as possible, then Turned the flex fuel logic back on, now LTFTs will come on alot faster and it learns alcohol % quicker and accurately. PS I would turn off endpoint locking and allow the ECM to go off the true alcohol % untill you get things sorted.
Last edited by iceman321; 09-26-2021 at 03:27 AM.
I'm still burning out the E85, filled up like 13 gal of 93 octane but its still inferred at like 40% alcohol. Michigan has some strong E85.
On my actual Ford flex fuel vehicle with 87 octane gas, it's stable at 9.8% alcohol with stoich 14.00 and within 15-20 min of driving (after a new tune flash) it showed status as Matured, and Fault Exit. Once it showed Matured, LTFT data started to appear. On my other vehicle I'll have to keep burning off the E85 to see if it matures with straight gas.
I think the issue is that I'm stuck in "Refueling" for some reason. I can drive for 4 or 5 stays without a refuel and it'd still be on Refueling. This car has an issue with the fuel level fluctuating all the time but not beyond the 10% threshold.
On the factory FFV, the AFR Source is Fault Exit (at least after a flash) and the AFR Status goes to Matured within 1 drive cycle.
After spending a better part of an hour playing around with various settings, I think I figured it out.
On the FFV Ford, inferred source is Fault Exit after a tune flash but the Inferred Status will go from Actively Learning to Matured within the same drive cycle (just drive around for 20 min or so). On the non-FFV Ford that had Flex Fuel enabled, the inferred source was stuck on Refueling. I played with the refill threshold, NVRAM setting, etc... and finally found that if I disabled Flex Fuel and let it idle until LTFT started switching, I'd then re-enable Flex Fuel and now inferred source is Fault Exit. I'd see the Ethanol % slowly rise up from the default but I didn't get a chance to drive this vehicle long enough to reach matured status - but I think this solved it.
The problems all started when I reset all PCM adaptive data via FORScan on this particular vehicle. For some reason when I did the same thing on the FFV Ford, it learned just fine. It's a good possibility that this FFV Ford had the NVRAM setting by default but HPTuners didn't define this switch so there's no way to tell. But on all the F-150 5.0s and the Mustang 5.0s, they had this as the default. This non-FFV defaulted to using the Init Value. This is just a theory of course.
BTW if you look at the Fuel Type SAE PID, it will read Gas if flex fuel is disabled, but will always read Ethanol with flex fuel enabled - even if you're running straight gas. On the FFV Explorer, I've run nothing but straight gas all this year and its inferred alcohol content is 9.8% with a stoich of 14.00 which sounds about right for Michigan's E10 mandated fuel.
I went to get gas (just straight 93 octane) and now it went back to Refuel status. According to Ford's principles of operation, after refueling it stops learning LTFT and waits for the inferred stoich to stabilize before it starts learning again. I have a feeling that if I put in any amount of E85 over the standard E10 amount, it will just keep fluctuating and not reset the Refuel status unless I disable Flex Fuel. I will swings of about 0.95 AFR which whatever I have in the tank now. So it goes as low as 12.5 and as high as 13.5 and never stabilizes.
I tried increasing the window for AFR and lambda for stabilization but it doesn't do anything. Very odd behavior.
I feel the Max Update Rate that's missing from my strategy is what's preventing it from learning. If I do the workaround to get the AFR Source to exit out of Refueling, the next time I refuel it goes back to this. According to Ford, after a refuel, it stops the LTFT learning and purge valve until it stabilizes and learns/matures. The non-FFV vehicles probably all have the update rate set to 0 by default but because its not defined by HPTuners, I can't change this setting.
Sounds like you've done everything you can with the parameters you have available at the moment.
I'm not convinced that's the issue since the inferred ethanol is actually changing. I lean towards the stability threshold being too narrow or your air meter (MAF or MAP) not being accurate enough and causing large swings in inferred ethanol.
Does the inferred ethanol actually mature to ~e10 when flex fuel is enabled and on pump gas?
I'm on my way, but it's taking forever to get rid of the extra ethanol in the fuel. I need to run my tank near empty this go around. It's been about almost 2 tank fulls of pump 93 octane and I'm down to about 27%.
The problem is that it is supposed to infer the AFR with various mixes of E85 if it is supposed to be flex fuel. I tweaked the stability threshold without any positive results. As soon as I refuel, it goes to AFR Source Refueling and never leaves no matter what i do.
I'm pretty sure the non-FFV Fords have the Max Update Rate default to 0. What happens if you set your Max Update Rate to 0 on a Mustang GT and try to fill with E85?
Ha, I just did that this afternoon as a matter of fact. Everything else as shown on my flex fuel tab earlier in this thread. It did mature, but it took 45 minutes of driving. Way longer than usual. I had 93 octane in the tank. Just returned from a trip where E85 was not available. 14.5 gallons of E85 into the tank and it matured at 69.8%.
Do you have the stability threshold that Witt mentioned? I could set that to zero and refuel for a test as well.
Can you post a screenshot of your flex fuel tab?
Stability Threshold is stock 0.10 which is what the FFV Fords have as well. I've increased that value as well and it didn't do anything.
Regarding the fuel trims, my FFV Ford had really wonky fuel trims (Speed Density/MAP, no MAF from the factory and it is the N/A 3.5 V6) and it will stabilize very quickly on pump gas.
My non-FFV Ford would fluctuate the inferred AFR / ethanol content even on full E85.
Here's a screenshot of my current tab. It is still stuck in Refueling. I have experimented with increasing the stability threshold out to 0.90 as well as increasing the lambda max min, Air mass min (lowering it), lowering/increasing the switching/time to no avail.
I disabled Flex Fuel, flashed, restarted, and then re-enabled Flex Fuel to bump the AFR Source back to Fault Exit in order for LTFT to start learning. But once I refilled with gas, it switched to Refueling and has been stuck there since the fillup about 100+ miles ago. The only thing I have left to experiment is to try and get rid of all the ethanol in the tank to run straight E10 gas.
Flex Fuel Tab 1.JPG
Here's a screenshot of my blend curves that you have also, blindly copied from a 2020 F150 file: flex blend curves.JPG
I'm sure you've compared plenty of blend curves, but thought I'd throw that in for reference.
Other than the Max Update Rate missing as you already know, maybe something behind the scenes we're not able to see, or we're seeing something on your flex fuel tab that isn't actually active in the tune.
I didn't have a 2020 F150 tune file but copied the blend tables from a 2018 Coyote F150. I'll take a look at the ones you posted and perhaps that might do the trick.
For what it is worth the inferred AFR will swing about 0.7 to 0.8 max in a given drive cycle. The warmer the engine is at start, the less it swings.
Reply With Quote