Alpha-N
- Note: low-boost is not supported. Always use high-boost, especially with alpha-N.
- Low-boost might not be supported in future at all
- in VemsTune before 2010-10-13, 220 kPa threshold had to be adjusted instead of 110 kPa - admittedly confusing, and fixed
The new method is quite simple to set up:
- TPS based VE * "map-correction" table
- "TPS lookup below 110 kPa (adjustable threshold), kPa lookup above for ignition and lambda"
Check VemsTune dialogs and help !
GenBoard/UnderDevelopment/AlphaN and Below is old, OBSOLETE information, the hybrid_rpm_... method is difficult to set up, not supported now (0,0 RPM or 25500,25500 RPM applies according to "load multiplication" setting). The new method is superior and simpler to set up.
TPS-MAP blending for hybrid_rpm_a < RPM < hybrid_rpm_m ; Load-multiplier - overview through examples
Traditionally, 3 modes are most practical:
- MAP-based, called speed density (for the whole RPM-range). The most common setup, maybe 95% of applications
- load-multiplier is enabled (config13 bit2=0)
- hybrid_rpm_a=00 and hybrid_rpm_m=00 (both 0 RPM so MAP is used above 0 RPM, that is always)
- TPS-based Alpha-N - under hybrid_rpm_a (which might be the whole RPM range if you like)
- load-multiplier is disabled (config13 bit2=1)
- hybrid_rpm_a=FF and hybrid_rpm_m=FF means both set to 25500 RPM, so TPS is used below 25500 RPM (that is: always)
- you can use TPS-MAP blending by setting eg. hybrid_rpm_a=14 and hybrid_rpm_m=1E (just example values for blending between 2000 and 3000 RPM ... other values might be used too), but DO NOT expect flat ve-table above hybrid_rpm_m (because load-multiplier is not applied, it will be steep)
- MAF-like direct mapping. MAF or other analog signal representing load is connected to MAP input (and scaled via kpafac, offsetted via kpaoffs)
- load-multiplier is disabled (config13 bit2=1)
- hybrid_rpm_a=00 and hybrid_rpm_m=00 (both 0 RPM so MAP is used above 0 RPM)
- TPS-based Alpha-N at low-RPM, with load multiplier applied - for turbo engines, the load-multipler is desired (MAP-signal necessary)
- load-multiplier is enabled (config13 bit2=0)
- hybrid_rpm_m=18 (2400 RPM, for example). Set below the RPM where your turbo starts to deliver boost
- hybrid_rpm_a=12 (1800 RPM, for example). Somewhat below hybrid_rpm_m of course. You want at least 500 RPM range for blending (2400-1800=600 in the example), but easier to map if you have 1000+ RPM range (possible if boost comes in at higher RPM, which is usually the case with agressive cams where TPS-based mapping makes sense at all)
- Please note that the calculation for this rare setup changed for RPM < hybrid_rpm_m 2007-04-14 in 1.0.73 and 1.1.20 firmware
- the VE (j[]) table will (need to be changed at low-RPM and) be monotonously increasing with respect to TPS (no longer have the low-VE strange shape at high TPS and low RPM). No need to retune VE-table for RPM > hybrid_rpm_m
- the change does not effect speed-density, where hybrid_rpm_m=00
- the change does not effect pure alpha-n where load-multiplier is disabled
- actually, we don't know anyone who used this setup. If hybrid_rpm_m > 0 and load multiplier is enabled, you are effected.
Tuning
[[Manual: Detailed.Control.Alpha.Config]]
Tuning is definitely more difficult than pure MAP based speed density.
Note that it is highly recommended to place an RPM bin (r table) on both hybrid_rpm_a and hybrid_rpm_m (3200 and 4800 RPM in the example) because the tables will look strange (and possibly steep) between these columns. The reason is the smooth, continuous change. The function that selects relevant table-lines blends smoothly from TPS to MAP as the RPM goes up from hybrid_rpm_a to hybrid_rpm_m.
Otherwise the TPS is highly nonlinear at low RPM, as an example, TPS=6% can result in 70% of the available power at 1000 RPM. Also note that any changes in the TPS sensor position (or calibration) will affect the VE table heavily in the alpha-n region.
Notes
- VE_learn currently works for speed-density only, not for Alpha-N
- make sure to adjust VE learning not to learn under hybrid_rpm_m (if using VE learning at all)
- MegaTune theoretically moves the dot according to kpa_tps_blend below hybrid_rpm_m (and according to high-resolution MAP above that RPM). However, be suspicious until you find it acting right.
- The highest "kpa" (which in fact load, not kpa under 2000 RPM) bin recommenended value is 200..255 because 255 means 100% TPS in the alpha-n low-RPM range. All the other load bins can be as normal, eg. from 10..100 in case of an NA)
- could it be changed so 0-100 kpa would mean 0..100% TPS?
- theoretically possible, but I don't see it justified:
- there are many bins, remaining 11 is more than enough (only those ECM-s need more kpa bins that have broken injector-opening model)
- the TPS -> MAP is not linear at low RPM anyway. There are only 2 points where TPS and load bins are equal (one is 0, of course). If the second point is MAP=40 kpa (40/256=16% TPS) or MAP=100kpa as you suggest: matters little. Actually, at TPS=39% (100/256) at low RPM the load is above 90%, so the current approach is better.
Who uses it?
I only know a racing team with ITB-s that uses VEMS with alpha-n on a 4-cyl. Unfortunately they don't publish their stuff (to protect their" work" ). I know they blend from alphan to MAP from 2000 to 3000 RPM.
Single throttle-bodys should not be a big problem either.
GRM team has ITBs and will be trying it. as always we'll post our configs and tables. 4 clyn, 16v, 2.0L huge cams.. nitrous.. if it ever works.
I'm (PeepPaadam) going to use ITB's (from GSXR750)on my Porsche 931 Turbo though I was wondering if I could get away with the speed density tuning only and not blending?
I (MembersPage/JanneMantyla) use it on my Honda 16V SOHC engines with CBR954RR 42mm ITBs. Daily driver, works very smoothly, easy to tune too and pretty good fuel economy.
[[Manual: Detailed.Control.Alpha.Operate]]
See also