Monday 25 May 2015

MEMS 1.9 ECU basic tuning / adjustments / road dyno - updated

A new feature (tuneview tab) has been added to the MEMS_DIAG android app (full version only)




The "tuneview" tab should be self-explanatory, one selects the tunable from the drop-down list and then uses the increment or decrement button to change the value of currently selected item (of course the connection must be first established on the main connection tab). The effect should be seen in the mini "Live data" section below the inc/dec buttons.

So far it has been only attempted against a MEMS 1.9 system, but in general both 1.6 and 1.9 ECUs are very similar.


The following "tunables" are available:

a) Idle target rpm - sets the desired idle engine speed (warm engine, but should also influence cold/fast-idle). Granularity seems to be 25 rpm ( +1 raw == +25 rpm) .

b) Idle fueling trim 1 - sets the base idle fueling trim, seems to be an offset from the main idle map. One should increment or decrement so that the FuelTrim value in the "live section" equals 100 (meaning no additional feedback is needed from the lambda sensor). Sometimes might be useful to set it a little bit on the rich side (FuelTrim <100 in the live data section)

c) Idle fueling trim 2 - in case of MEMS 1.9 it's identical to "b". But perhaps with some other variants it might have a different effect.

d) Unknown (ignition advance offset ? / acceleration (pump) enrichment) ? - This is currently unknown, seems to be somehow related to ignition advance offset or a kind of acceleration enrichment, if the value is DECREASED the throttle response seems to be better. Try min/max values to see if there is any effect.

(update: 2015-06-10)

Default setting: 128
Min value: 116 (best performance)
Max value: 140 (poor performance)

It could be a kind of ignition retard offset for markets with poor fuel quality (to avoid knocking). Apparently, there have been rumors a similar option is available with rover testbook.

There is no easy way to measure the effective ignition angle (map +/- offset), but knowing the vehicle weight, gear ratios (3rd gear, IRD, main diff.), cx + frontal area, rpm_over_time, we can create a rough estimation of engine torque/power delivered to the wheels.

Here is the result, torque & power as seen the the wheels (two runs, effect of min/max settings of the "unknown tunable", drivetrain losses not included) :


It does seem to confirm the subjective feeling, but of course more tests are needed.

And as a reference, "official" torque & power graph of a 1.8 K-Series engine (as seen at the crank):



e) Idle recovery - decay rate - Seems to influence the "return-to-idle" rpm speed. Its effect is best visible during gear change, the engine will maintain the revs a bit longer before it returns to idle (or shorter, depending if increased or decreased the value)

Important:

- it's an experimental version so use it at your own risk !

- Note down (and keep in a safe place) the initial values of each tunable (visible after 1st inc/dec change/refresh). You might need them to go back to the original values in case a rollback is needed.