To start... I'm neither evil or hard to get information from. I'm a pragmatist with a twisted funny bone.
I could go into a long diatribe discussing how some bikes used high and low speed ignition windings on the stator and two other windings to generate CDI signals, relying on the phasing between the various AC signals to control when spark occurs... but that's not relevant here.
One of the keys to understanding TCI and other similar systems comes from realizing the tooth on the flywheel is wider than the sensors detecting it. This means the controller gets one pulse when the tooth enters the sensor's area and another pulse, albeit opposite in polarity to the first pulse, when the tooth leaves the sensor's area. The first pulse lets the controller know an ignition event is about to take place; the second pulse is a "drop dead" signal that fires a coil if it hasn't already fired, and should coincide with the static timing of the ignition system. What happens inside the TCI module is a bit more convoluted, but essentially it is using RPM to figure out when to fire a coil.
One possible method is for it to use the voltage generated by the sensor to determine what the advance should be: as RPM increases so does the voltage output from the sensor. A higher voltage, representative of higher RPM, would therefore trigger the ignition system to fire earlier thus causing advance. And, in kind, as the voltage drops, the system would trigger later, thus retarding the ignition, up to the point where the tooth itself leaving the sensor's window would cause a coil to fire. Likely, the VR's signal is "clamped" to some level such that maximum advance occurs at some RPM above which, although the sensor's output voltage is increasing with RPM, no additional advance would occur. I'll leave it to the EE's to contemplate the necessary circuitry, but having opened up the ignition module on my motorcycle, I can say that although transistors imply digital operation there's very much an analog component to the system.
Another method, whilst not being likely to have been used at the time, is to have a microprocessor simply get the signals, run a timer and do the math to work out position, then fire the respective coil. The go-to 8051 or comparable at the time would have been more than capable of doing this. Freescale has an EVM for small engines that goes a little bit into this theory, and other vendors such as Ignitech have developed fully digital systems that emulate CDI and are fully programmable.
Unfortunately, because the tooth is wider than the sensor, the sensor's signal isn't a reliable input for a VR chip like the MAX992x as they require a fairly clean positive-going-negative transition through zero volts to give an output. The stock VR output looks somewhat like the following, but please note the signal could be inverted depending on how the sensor itself is wired, and is probably not sharp lines and instead rather curved like a sinusoidal wave:
Code: Select all
/\
___/ \____ ____________________________
\ /
\/
... instead of:
Code: Select all
/\
___/ \ ________________________________
\ /
\/
... which is what the MAX chip would expect.
So, I would have reservations using this to trigger fuel injection only without some sort of conditioning circuitry.
Triggering off of a coil signal is a possibility, but we'd need to figure out exactly how your coils are triggered- i.e. CDI-like pulse to the (+) terminal or switched (-) terminal. If you have a higher-resolution image of the bike's wiring, you can email it to me at
abecedarian01@gmail.com. Alternatively, it might be possible to trigger off of a "tach" signal feeding your cluster, but then again we'd need to know the voltage levels present on that wire.
You can lead the horticulture but you can't make them think.
