DIYEFI.org Forum

Today I finally got the break in the weather to try out my modifications to the V3.00 motherboard.
As I posted in another thread here: http://www.diyefi.org/forum/viewtopic.p ... 5&start=10 where I gave a summary of what i had done to remove all of the power devices from the digital and signal rails.

Since writing that and testing again on the car for noise on the rails and sensor inputs I have made some more changes:

Changed the IRFZ34N COP Mosfets upgraded to 55Amp 60Volt P55NF06FP devices and brought them inboard to the MS2V3.00 board on the new power ground heatsink bar.

Stopped using the MS2 stepper chip to drive the mosfets for PWM idle and VANOS. I found that when not loaded just right i.e. with an inductance the PWM current limiting feature self oscillates, causing 500mV of 25Khz ripple on the ground return from the MS2 daughter board. This throws the CPU into disarray, having loaded the MS2 stepper with several stepper motors I have seen the noise reduce but not below 200mV. Which on microcontroller power rails is exactly what you don't want!

Used 2 P55NF06FP MOSFET's to drive the push pull 3 wire IACV it's a ridiculous overkill I know but I have a surplus of these! I used a simple common emitter NPN BC547 to provide the inverted signal for one of the MOSFET's.

Fitted Varistor surge suppressors between the lowside of the COP and power 12V to kill any nasty spikey cack from the Coils.

Rearranged the connections to the power 12V and the new power ground heatsink bar to aid simple construction and extra MOSFET's.

Used a direct battery feed into a relay controlled by the ignition key to get a straight from source clean power feed for the MS PCB.

Kept all of the dirty noisy stuff on the main relay controlled +12V BMW circuit.

The results.......Well astounding, with an auto generated map it started from cold within 5 seconds and idled smoothly and comfortably at 1200RPM then after much playing and amusement settled at 950 rpm smooth as silk, the battery monitoring ADC channel no longer picking up Radio one was behaving beautifully.

I have driven it, and as predicted needs some work as its a bit spluttery around high Kpa and low RPM, but that always comes with the tune.

All together a very productive day and some really valuable lessons learned to carry over into the FreeEMS design.

I am going to make some high res video images and post them on Youtube for you all to see with an update to this thread to give everyone a detailed and tested plan for proper and relatively easy results.

Before: Noticeable instability and noise on battery voltage corresponing with injector PWM.
http://www.youtube.com/watch?v=evd6msCcWZ0

After ground separation: Totally stable and drive-able!
http://www.youtube.com/watch?v=nVRwZ-N1auI

Glad to hear you got it running. I've been following your threads, and I think many or most of the issues you posted about are non existent in freeEMS, but time will tell what strengths and weaknesses it has. Keep up the good work.

That's awesome Dave :-)

Only one small thing, I thought you said you had high Z injectors? If so, you should not be PWMing them at all.

Glad to hear it is running well for you!

Fred.

Fred wrote:That's awesome Dave

Only one small thing, I thought you said you had high Z injectors? If so, you should not be PWMing them at all.

Glad to hear it is running well for you!

Fred.

I do have Hi Z at the moment but with a supercharger bolted on i may need to swap the injectors out and i don't know yet what flow rate i will require or whether they will be High or Low Z.

So 1:
keeping options open

and 2:
How could i test the integrity of my modifications on a set of low Z injectors without fitting them?

and 3:
If only I benefited from this work then thats a bit selfish and you guys should gain from this too.

plus its not as if I need the space that the few components take up, or possibly weight saving about 50 grams!
i'll just take a pee before setting out if I'm that concerned with weight savings!
So really there are more genuine reasons for just leaving the parts in the board than not.

My high z driving board has all the parts too. I was concerned about potential issues with duration and closing etc due to lack of current, however if the duty is high, then it is much the same. I assume you are running something like 90% rather than the usual 25% ?

Fred.

Well the manual said 75% but i went for 90% seems to work ok