yes exactly, also it will only switch to the safe map if it was on the aggressive map by-way of the 4.7v+ beforehand.

I haven't thought too much about the logic, and there is probably a hole in there someplace :)

no worries MrFred, the change from 9417xxxx is quite complex anyways due to the extra 4 maps... I need to triple check everything :)

edit - in fact I should have gone from 88580014(myrom) to 88590015... would have been easier, duh!!! /smacks self!

ok something in the logic might be a problem.

when going from aggresive -> safe, I do a check to ensure the previous ADC value was 4.7v+.

Is there any chance that the methbox will transmit a value in between 4.7v and 0v?

ie if my code sees:
5v.... 5v.... 5v.... 4.2v... 3v.... 0v... 0v... 0v...

Then it will stay on aggressive...

This may seem counter-intuitive, but if I go straight to safe when I see 0v then that will break nonMeth+ICS equipped evo's...

It is quite simple to generate a 5-0V decay on a failsafe output.

Adding a small hysteresis would be nice between switch voltage points.

aggressive map >2.5V
safe Map <2.5V
Hysteresis ~1V would be nice to start with.

well I don't want a decay because that will break the code :)

Can you explain the Hysteresis more? I know what it means, but not in your context!

Cheers
D.

I know you know this tephra, but the ECU has some ready made code in the TPS hysteresis IC spray control in auto mode. Of course voltage hysteresis is no good for debouncing.

Are we not better using a digital input for switching and analog one for scaling?

mmm, i can add in debounce logic pretty easy, just a combo of counters and nested if's :P

re two inputs, i was thinking about this today, 1 cable is less wiring and we can use the extents (0v and 5v) as off/on - digital switching.

Analog inputs are scarce, best read with interrupts disabled, susceptible to noise, don't always read 0 or 255/1023 etc when fed 0 or 5V because of analog grounds and Vcc etc. Digital inputs require less processing overhead, have the option of being interrupt driven, often have built in Schmitt triggers etc. If we don't have or can't find a digital input then we don't have the option, but if we can use one and leave the analog one for sensors that require it I think it would work out better all round.

Deboucing circuitry many not be as necessary if you have two quite well defined voltages. Debouncing is more important for toggle/flipflop single hit switching.

As far as we are trying to achieve now is a reliable transition switch between the two maps. We can almost leave the analogue part out.

Having a switch point of 2.5v with a certain amount of hysteresis will ensure a one-shot cross-over point.

Having a dead band of 1V well allow 0-5V positive transition, system treat it as "1" when it crossover the 2.5V. It has a noise immunity of 0.99V during the transition. I am just trying to terat the ADC as a Digital input for the timing being.

Reserve the ADC capability for the future to perform, perhaps close-loop fuel/alcohol ratio control.

tephra,

You have a full inbox. I have one more off-topic question for you please.

Ok, just got in from work. Trying to catch up on all the transitioning talk,lol. So Tephra, with the ROM you emailed me, it will only switch properly if it see's an instant voltage difference of 5V? I'll see what I can test tonight but its dark and the temps are dipping into the single digits already :burr:

I got the first set of pins I ordered and they are definitely for the larger of the 2 pins. I should have the correct size in a few days. For now I may just steal a few of the ICS wires for testing.

Richard - fixed.

JoT - yup basically

John - I see your point, I will leave it upto the group to decide - Hijack the AutoICS digital input OR continue to use the ADC.

Since I have no idea what the true differences are to these options, I'll leave it up to you guys to explain at some point,lol. I just finished swapping over my ROM info so I can possibly load it tonight.

Jamie,

I think we are try to emulate the ADC input as a digital input for the time being until one day it can be used as a true Analogue input for close loop control such as duel fuel E85/normal gas etc.

I am sorry it is getting so cold for you to do the work.

are you going to be using the pin#42 that you found and tested..or we sticking with one of the ICS pins?

thats the option yes, however if we are to use the ICS pins for meth failsafe then the code will need to be changed and this will conflict with current ICS altmap users...