https://datazap.me/u/dven11/log-1764964854?log=0&data=1-2-3-4-5-6-8-9-11-14-15-18-20-25-26-27-28-29
| duty_15 | duty_20 | duty_25 | duty_30 | duty_35 | duty_40 | duty_45 | duty_50 | duty_55 | duty_60 | duty_65 | duty_70 |
32 | 32 | 32 | 32 | 35 | 35 | 33 | 43 | 50 | 52 | 58 | 60 |
Great data. Yes, I've seen that high-elevation cars on 91 struggle on timing even on stock map. I likely wouldn't recommend JB4 to someone in that situation unless they have access to ethanol like you.Compiled the data from a bunch of logs on different E content to capture timing convergence in ign_1, which seems to be my best indicator of how happy the car is with the octane.
Map 0: great correlation in my E20 and E25 runs, but there's a big step between E10 and E20. Need to try E15 to fill out the data and find the actual point of convergence.
Map 1: improvement between E25 and E29, but the E29 timing looks near perfect. Will try E35ish to confirm. Interesting that timing here is a fair bit higher than the Map 0 E25 pull. The DME appears to favor advancing timing vs adding boost. The second graph shows boost stuck on a second y axis. A bit more cluttered to look at, but it demonstrates that the response to insufficient octane is to pull a bunch of timing and add boost, possibly to chase torque targets?
Map 2: also looks pretty dang good on E29, although timing is a bit lower than the Map 1 E29 pull <5000rpm, also interested to see if any changes are observed bumping up E a bit higher.
So, what does this say about octane needs? Effective octane has a logarithmic relationship with ethanol content; you get a bigger jump in octane going from E10 to E20 than you do going from E30 to E40. So calculating as if the relationship were linear, as I have done in the past, might get you a decent estimate but not all the way there.
source: https://ethanolrfa.org/file/2967/Octane Value of Ethanol_Hoekstra_2025-06-24.pdf
With my 91 octane being e10, the base gasoline octane is likely ~87 AKI (R+M)/2. So, in my case:
E10 ~91oct
E20 ~93.5oct
E25 ~94.5oct
E30 ~95.5oct
E35 ~96oct
E40 ~96.5oct
Stock, the car appears to want ~92oct. Map 1 wants ~94, and Map 2 wants ~95.5. Couple things to note here: octane ratings at the pump are minimum values. If you have "good" gas, it's likely that your octane is a bit higher than what is shown at the pump. If you have shit gas, it's probably right at that value. Additionally, I'm at 5000ft+ elevation. Compressors (the intake side of your turbo) operate on pressure ratios, NOT nominal boost values or even absolute pressure. Pressure ratio is absolute pressure/ambient pressure, or (boost + ambient pressure)/ambient pressure. I wish we had a published compressor efficiency map for the stock turbo, as it could provide some good insight into its operation and also better flag when the turbo is getting pushed outside of it's efficiency range. All else equal, my turbo's limits are lower than at sea level, and it'll push into heat generation instead of air density increase at lower boost levels - which increase propensity for knock.
So don't take these values to be absolute truth. They are applicable to MY car, in MY ambient conditions. MT cars in Colorado will likely see similar results. Everyone else, consider these just reference values.
Logs
Map 1, E29 (91oct mix): https://datazap.me/u/nick470/map-1-e29?log=0&data=1-4
Map 2, E29 (91 oct mix): https://datazap.me/u/nick470/map-2-e28-0?log=0&data=1-2-3-4-11-15&zoom=8-72
So, my takeaways from this are that I've found decent octane requirements for my car, in my ambient environment, for Maps 0-2. A little more testing to be done, primarily with E15 and E35 blends to fully flesh out my ignition convergence.
What I'm starting to focus on now is cleaning up the spool up region. I'm going to start this by just fiddling with duty bias in the OTS maps. If this is unable to sufficiently clean those up, I'll shift to a boost target approach on Map 6, dropping target below even Map 1 <3500rpm. In this region, I'm getting spikes in calc_tq, trims, throttle, and just generally a lot more noise in both boost and ign_1 than across the rest of the pull.
Once I clean up Map 1 and Map 2 to my liking, I'll record single gear pulls on Dragy, pick out a metric that offers the best repeatability (ie 40-90 or something around there), and calculate average acceleration to get a realistic delta between maps. Without compressor map data, this can be used to back my way into identifying when I'm at the edge of the compressor efficiency; if boost goes up, timing is the same, and no other quirks appear in the logs, but the car isn't getting faster, I know that I'm no longer adding more oxygen molecules into the combustion chamber. Then, I can start to build out a boost curve in Map 6.
Yes, you can run up to E35 or so with the fuel pressure connector installed. The user input is FOL (fuel open loop). Recommended between 60-70. Just raises fuel pressure a little which reduces trims, injector pulse width.
Dme_bt has the same wiggle, so boost is following what the dme wanted. The same wave is seen in the speed sensor, so it was probably traction control. (keep it on, if you spin you'd need to lift anyway so it's a garbage log either way).
Timing is also bad, but this is probably a separate issue. You have lots of trim headroom if you want to add e85.
Bingo.
| map6_15 | map6_20 | map6_25 | map6_30 | map6_35 | map6_40 | map6_45 | map6_50 | map6_55 | map6_60 | map6_65 | map6_70 |
1.5 | 2 | 2.5 | 3 | 3.5 | 4 | 4.5 | 5 | 6 | 6 | 5 | 4.5 |
Its just a voltage offset for wastegate position, essentially the JB4 WGDC variable is meaningless other than 50 being neutral. <50 is lower duty cycle, >50 is higher.
What you are saying is something I recently tested with Terry. It may still be enabled on newest firmware (34+). Under E85 settings in user adjustment, try raising the "min gear" variable to like 5. If that feature is still implemented, that will delay the JB4 WGDC control on spool. Its adjustable too.
Give it a shot.
Nah it can go way more than 6, but PID is irrelevant since it is off on additive maps now. WGDC=FF.
Alright so I have a list of things to try out, any suggestions on order of operations?Its just a voltage offset for wastegate position, essentially the JB4 WGDC variable is meaningless other than 50 being neutral. <50 is lower duty cycle, >50 is higher.
What you are saying is something I recently tested with Terry. It may still be enabled on newest firmware (34+). Under E85 settings in user adjustment, try raising the "min gear" variable to like 5. If that feature is still implemented, that will delay the JB4 WGDC control on spool. Its adjustable too.
One other thing you should try is enabling Bit 4 and Bit 6. Terry said Bit 4 (320i mode) gives better WGDC control (resolution?) at lower boost, and Bit 6 (BEF) increases the voltage bias for WGDC control (ie, lets the JB4 have more EWG "gain").
All I know is that it helped on mine
| map6_15 | map6_20 | map6_25 | map6_30 | map6_35 | map6_40 | map6_45 | map6_50 | map6_55 | map6_60 | map6_65 | map6_70 |
| 1 | 1.6 | 2.2 | 2.8 | 3.4 | 4 | 4.5 | 4.5 | 4.5 | 4.5 | 4.5 | 4 |
That's kind of a bummer he would just hardcode it off now, additive with small PID corrections for weather was working well on my car
He changed the way FF is targeted, but I can't remember exactly what. It works just fine, just relies more on the DME PID. Absolute maps work same as always.
