Firmware r023

Starts and runs! Idles fairly smoothly with only set-screw holding throttle body at 4-5% open. Drove around a little bit, runs very well and all parameters look good. Accelerates much harder than when K-Jet was running the show, until it gets to the sections of fuel map I haven't tuned. I'll be reviewing logs today, updating tables and doing more testing, provided the rain stays away. Thank you, everyone, for your help!

* Notes:

* * VE table values had to be doubled even though req_fuel was set at value recommended by req_fuel calculators.

* * Top dead center after trigger was retarded by 30 crank degrees.

* * Use MegaTune to enable/disable features (misc_out, boost control, iac, etc). Disable values in config.txt inconsistent between feature channel outputs.

* * LCD display is cutting out, displaying with strange offsets, otherwise acting funny. Vibration getting to headers or wiring elsewhere? Will test.

----

Progress 9 July 2006:

Firmware r045, Megatune for 1.0.36

On conversion from firmware r023 several options were left unset in config which threw everything off quite a bit. Be sure to review all pages and set config accordingly. Enabling antilag and launch by accident will prevent engine from running.

* Fuel cut was defaulted to 0kPa which confused me for a few minutes when it would fire while cranking but never start.

* Autotune still wants to drag my VE values down to zero, even though they VE figures are fairly accurate. Anyone else experiencing similar problems?

* Spark cut idle control works nicely, but now there are several pages for idle settings. Consolidation possible to avoid confusion?

* Having trouble spinning motor faster than 6000 rpm. Can someone review my config (good03.msq) and recent datalogs and give some comments please?

**We have this problem too with 1.0.36 firmware: when reaching something like 6000rpm, the engine shuts itself off. And then turning the starter, acts normal until you rev it more than 6000.... PeepPaadam

----

Originally tdc_after_trigger was set to 60 deg but engine wouldn't fire because of incorrect trigger_tooth was being used. First test why engine wouldn't fire revealed:

* * Top dead center after trigger was retarded by 30 crank degrees.

tdc_after_trigger was adjusted without changing trigger_tooth. This will be remedied in next test session. (14 Oct 2006). trigger_tooth should be adjusted by 3 teeth.

*Firmware r045

*Config at http://proton.shadowflux.com/vems/config/megasquirt200607291807.msq

*config_20060729.txt and tables_20060729.txt at http://proton.shadowflux.com/vems/config

* generate-config.bat misses a few config variables including

** toothrel_normal

** toothrel_missing

** fuelcut_min_kpa

** fuelcut_max_kpa

** tach_channel

** tach_div

** shiftcut_conf

** shiftcut_channel

** shiftcut_time

** softrpm_rpm good candidate to limit RPM

*** softrpm_rpm is set to 0xFF

** crank_minper can also limit RPM if too high (use 0x30 = decimal 48)

*** crank_minper is set to 0x50 currently.

*** dammit, this is too high. incorrect calculation early on resulted in this. It appears I have been limiting revs to 5859.375 rpm. This is pretty embarressing. thanks for your help everyone. Further changes will be made to experiment with the maps and other settings.

* I believe the IRC channel has come upon a concensus that passing the max MAP or RPM bin will result in odd behavior. I will resample MAP bins and note behavior.

** there is no odd behaviour if the bins are monotonous as they should. Eg if you have 220kPa in the highest MAP bin, above 220kPa you'll have the same VE and ignadv. If you have bad, non-monotonous values like (in this example, the RPM bin) 5000, 7000, 9000, 1400, 2000 than above 9000 there might be strange behavior. Correct it to monotonous 5000,7000,9000,9000,9000 (or ...,9000,9100,9200) sequence.

Maybe one of these vars has something to do with my artificial rpm limit at 6000 rpm? Engine will NOT rev above 6000 rpm. Seems like fuel is being cut, but I can't tell for sure.

Was having trouble with ego correction going wild. Fix was to increase injector opening times and reduce req_fuel from 9.2ms to 6.6ms. Now getting realistic ego correction and ve learning is working as it should.

In light of my recent crank_minper mistake I decided to create a chart showing the graphical representation of max rpm vs. crank_minper values. The Excel based chart is located at http://proton.shadowflux.com/vems/config/crank_minper.xls

A copy of this chart will be reproduced in the manaul.

----