RomRaider

Is there a status update on where this project is at and what issues still need to be resolved? I am keen to make the jump to corn juice and really want to stay with RomRaider rather than switching to Ecutek. Many thanks to all those involved in this project.

Hey Merp,

Just out of curiosity, how much of a donation would you need to "accelerate" development of this patch ? I'm sure we could rustle up enough interest to make it worth your while...

Cheers,

Adrian

Donations are always appreciated and keep me motivated to work on this, but motivation isn't a big issue right now, just time. I'm busy with work and getting my car running again so I can perform live testing and do a complete debug of the code. I plan on having it running for testing in June, keep your eyes on the MerpMod Test Build thread for releases.

That sounds promising. I would love a working patch within 3 months. Good luck with the testing on your car.

Z0rr0 - did you manage to get your ECA using the Arduino working? I have been coding up flex fuel for the Nissan ECUs and have it running on my 200SX at the moment using the Zeitronix unit

I took your initial modified code but changed it to use digital input interrupts instead for the frequency collection (your posted code required analog inputs)

Current issue is the smoothed output from the PWM output (pin 11) to get the 0-5 volt range. Have currently tried to increase the frequency and using a low pass filter like this:
http://www.instructables.com/id/Analog-Output-Convert-PWM-to-Voltage/

However the resistor drops the voltage to smooth the spikes which results in a lower voltage range than then needed 0-5 volts

Did you manage to find a way around that part? I'm thinking an opamp or similar to multiply the output voltage to get the 0-5 volt range back but interested to see what you have done

I did not get it fully functioning.
I had a noisy environment, the output would be fine, until i started the car.. then it jumped all over the place.
I never did get around to fixing it.

with the key on and the engine not running it worked perfectly.

I still think an easier way to do this is using SSM...

Aside of accounting for the mixing of fuel whilst driving contributing to a small change in alcohol percentage, I do not see a need for the ECA to have an influence on the map blending ratio past the point that the car is initially started. I.e., you can't/shouldn't fill fuel with the engine running...

If the OBD port was used to change a value in RAM that related the blending ratio once on every startup, this should alleviate a couple of issues. Firstly, any issues relating to noisy inputs, also gets around the issue of the ECA being such a critical component of the car, i.e. if the ECA were to fail mid drive, it wouldn't matter if it was only being used at startup. Lastly, those of us who still have intact TGVs have a way to get the fuel content into the ECU in an automatic fashion.

I've got code thrown together on an arduino that can read an ECU. Should simply be a matter for a simple command change to write to the RAM area (with a patched ROM i'd imagine).

Problem with that is post startup blending especially after fueling. This could work with rough and fine correction but then why make it that complicated?

In current form, what I'm working on will allow you to lock the fueling, although it will take 2 flashes. Just set the blending input mode and default accordingly.

However, if I move:

into the pRamVariables struct (RAM) instead of 'progmem' (ROM), you could theoretically do it all without flashing, although you will lose your stored injector scaling after a reset.

Implementing this in the ECU would be straightforward. Start a counter at startup or use an existing one and about 4-5 lines of conditionals.

what do you mean by "lock the fueling" ?

i.e. prevent it from changing whilst the engine is running ?

Yup. Basically set blend mode to ECA input -> flash -> get data -> set mode to static input and set static value -> flash. With ram variables as mentioned, no need to flash, but you'll need some kind of interface to do generic ram writes, and run the risk of losing the static value upon reset.

I ended up moving to a PIC16 and 8 bit DAC connected to it currently constructed on veroboard. This avoids the problem of using PWM output and trying to smooth that out.

Seems to work okay but some extra work is required on the input as it is occasionally glitching when the car is running like what you are noticing. The input signal on the scope appears to be fairly stable

Looking at the Zeitronix ECA they use appear to use an analog comparison via two inputs, prior to converting it but I hope to be able to stick with triggering interrupts from the digital input using pull ups and filtering on the signal

You will find you will need to oversample the data. If one of the samples is glitchy then it will affect your output. Sample say 16x and then look at the average, discount any samples outside the average and you have a more stable output

Looked at the part# on the Zeitronix. Not using comparators but just an 2 x interrupt driven digital inputs. 10K pull up and 1K via diode to ground on the input signal plus 10uf decoupling caps. Just got to try and reduce the effect of noise in my circuit and should be good

I'm probably just going to use the ECA2 to make things simple, but can't you use a RC circuit in place of a microcontroller to convert the PWM to 0-5?

Please do - just bought one

Engine is in, so a couple more weeks and I can do some flex fuel testing for you.