if you want to ditch the stock ignitor you will need to build an IGf generator circuit so the stock ECU doesn't throw CEL. There's info on the yahoo emanage group on how they make the IGf circuit when emanages are used on the 7mgte dli setup.
To help mount a 20V in a RWD I'm trying to remove the dizzy.
This will require some electronics to produce signals to (in this case) 4 coils instead of the one coil used from factory. I was hoping to intercept the IGT signal, and when it goes high, trigger one of 4 attached ignitors. And when it drops, drop the signal to the ignitor, creating a spark. After one ignitor receives a signal, the electronics will step to the next ignitor and send the next "spark" to it. Basically the electronics would send the IGT signal to each ignitor in turn.
I'd also intercept the G1 or G2 signal (one falling edge pulse per CAM revolution) which could work as an absolute reference for which cylinder should be sparking next. This would be needed at startup and in case the counting of the IGT signal got out of step (due to noise/programming error/etc...).
Now given that the blacktop ignitor does dwell internally (see here) I think we will need to keep the stock ignitor. And instead of intercepting IGT we could intercept the coil negative output of the ignitor.
The next problem is that I was planning on using a 5V microprocessor (Atmel MEGA8). Essentially this is because my uncle (electronics guru) recommended it, and he has stacksof them and can give me the hardware and software needed to program them. These can output a decent current, but only 5V. And so I was wondering if normal dumb ignitors would work correctly on a 5V input (though thinking about it more it would make sense if they did).
The clock speed of the chip is 8MHz from memory, so a clock cycle time of 125ns. 9000rpm is 54000deg/sec, so 1 degree is 18.5us, meaning around 150 clock cycles per degree of crankshaft revolution. I think this should be enough time to do the simple sequencing required, meaning timing should differ from factory by less than a degree.
So the overall plan for the system is:
And the questions I still have are:
- Will an Microtech X4 box, or other dumb ignitor work with a 5V input?
- Could I ditch the stock ignitor and replace the X4 box with 4 ignitors with internal dwell (eg Bosch 008)?
- Are there other chips I should consider which would do the job?
- Have I overlooked any obvious problems?
Does anyone have any comments? Or advice about this plan?
Thanks
Hen
if you want to ditch the stock ignitor you will need to build an IGf generator circuit so the stock ECU doesn't throw CEL. There's info on the yahoo emanage group on how they make the IGf circuit when emanages are used on the 7mgte dli setup.
Use a couple of LM1815s for interfacing G1 and G2. The rest seems pretty right I think. Could use whatever ignitors you want even if the input is 12v, just use some transistors to bump up the output of the Atmel. Alternatively you could use the micro to control dwell also, or use 4 smart ignitors.
Bumping up the output would take a few components per output, so I'd prefer not to, though it is an option. I'd also consider building my own ignitor using 4 logic level IGBTs (or whatever they are, afraid I'm a bit hazy on that stuff).
And Viper, what are LM1815s? (actually, I'll google it when I have more than 2 mins). We were thinking of interfacing through a capacitor and a pull up and pull down resistor.
As for using smart ignitors, would using 4 smart ignitors each seeing 1/4 of the sparks cause any problems with dwell timing?
Also, JP I've seen something to create an IGf signal to keep the ECU happy, so I'd ignored the problem hoping it'd be easy enough to simulate later.
Hen
Hen,
I have also tried to make contact with some people who are running the old dizzy on the front of the engine trick. This could solve all problems.... such as this guy
Thanks for your help smart ones.
R.I.P Peter Brock.
Bookmarks