-------------------------------------------------------------------
This is the Start of Part III - the tuning of the MegaSquirt.
-------------------------------------------------------------------
The section is to capture my attempts on tuning my car that has recently been fitted with a MegaSquirt (MS) ECU and throttle bottle injection (TBI). It is not a "how to" even though it could be seen to be one!
My cars specification is:
1968 Morris Mini Deluxe
1310 A+ engine (1986 Mini Metro)
A standard Mini Metro camshaft
MG Metro Turbo cylinder head
Maniflow LCB & RC40 twin box
HIF44 SU fitted with a Patton Machine Products TBI carburettor adaptor
Running distributor less with a Ford EDIS coil pack in direct coil mode, i.e the MS is driving it directly without the EDIS module.
The Rochester 5235277 Fuel Injector flows either 48lb/hr at 12psi or 41.6lb/hr at 13 psi or 58lb/hr at 10.1 PSI from three different sources.
Cooling is by standard Mini Radiator, but I've replaced the engine fan with an electric fan which is controlled by the MS
Since last Friday afternoon when the car burst into life, I've been diligently both data logging and also auto tuning using the fantastic piece of software called
TunerStudio (TS).
I logged locally and also both ways on my trip to the Hanging Rock car show. Had a bit of time last night to look at the data logs and it appeared that I was getting many out of range air fuel ratio (AFR) readings.
Edit: Here is the legend for the different graphs:
RPM - Rev's Per Minute
MAP - Manifold Absolute Pressure
TP - Throttle Position
DutyCycle1 - Duty Cycle of Injector (0 - 100% of time that injector is open. Shouldn't go over ~85%)
AFR - Air to Fuel Ratio
PW - Pulse Width of injector (Milliseconds that the injector is open)
CLT - CooLant Temp
MAT - Manifold Air Temp
Batt V - Battery Voltage
The following graph shows the AFR jumping up to 25:1 at cruise... it' not supposed to do this!
Scratched my head a few times over this, but to cut a long story short what I did was connect my netbook to both the MS and also to the TechEdge (TE) wideband controller. A then ran AFR gauges connected to both sources and let the car idle. The TE showed around 13.5:1 the MS showed 25:1 (the TE won't read up to 25:1 it's range is 9-19:1) so this meant the MS was going out of range.
Then the penny clicked, my son whilst looking at all the pretty gauges whilst we zoomed along the highway on Saturday, said to me that the TS software that was auto tuning, was displaying "out of rang error" every now and then.
The TE outputs the AFR as a analog voltage 0-5V which corresponds to 9-19 AFR. It was obvious that the MS was getting this very wrong. I then found a setting called "Calibrate AFR table...", mmm sounds good to me. I selected "TechEdge Linear" for the EGO sensor drop down box and then clicked the "Write To Controller" button. Bingo both AFR gauges now read within 0.1-0.2 AFR of each other!
So I decided to go for a bit of a drive (went to the You Yangs via highway, then back to Werribee via Little River so as to vary the logging and auto tune.
Here is a graph that now shows roughly the right AFR at cruise:
Looks a lot better doesn't it!
Some other graphs that have to be created and or modified are:
Ignition advance table:
Here is my AFR table:
Volumetric Efficiency (VE) table. This is what the auto tune function changes in TS. I have roughly highlighted the area that it has changed on today's drive:
And finally here is it idling followed by a period of acceleration through the gears:
The car is driving better than it was yesterday, so I consider that as a positive.
