If I understood well, you had to cut the 5 remaining teeth from their original width to something narrower. Is that correct?
Thanks for sharing your experiment anyway. This is so interesting
Sometimes I don't know how much technical information I should put into my posts, but if you're interested in hearing about it here's more...
Let me know when I reach the "Too Much Information" zone (obviously I find all of this extremely interesting or I wouldn't be doing it as a hobby).The Bandit's original equipment ignition system is a product of Japan's Denso Corporation (and its subsidiaries and suppliers). Denso's approach to ignition control during the '80s and '90s was a pretty unique and rapidly evolving one. The Bandit 400's system represents a sort of "half way" point in the automotive/motorcycle transition from the old-world of Mechanical Points and Condensers to the current world of fully digitized, ECU controlled ignition and fuel injection.
So when viewed from the perspective of today's world the Bandit's early '90s Denso ignition control system seems a bit strange or maybe just quaint. The part of this "strangeness" that was affecting me a couple days ago was the width of the teeth on the Denso trigger wheel.
Modern ignition control systems use one of three methods to sense a trigger wheel:
1. a VR sensor (VR stands for "Variable Reluctance"). A VR sensor is a simple "piece of metal passing by a magnet" type of sensor. A VR sensor is unpowered, it generates current for its signal simply by the metal trigger wheel teeth passing by the permanent magnet contained within the sensor.
2. a Hall Effect sensor, which is basically a more modern version of the VR sensor. It is a powered sensor and it has a built in circuit to interpret the magnetic effect of trigger wheel teeth passing by it, then its circuitry sends a simple "square-wave" signal (0 volts to 5 volts) off to the ECU,
3. an Optical sensor, which is exactly what it sounds like.
The Bandit 400's Denso system uses a VR sensor, which is what the overwhelming majority of today's systems use, but the circuitry in its CDI box was apparently designed to interpret the passing of the trigger wheel teeth in a very different way than what has become the norm (in current times, i.e. the year 2000 and onward).
Here's a breakdown of what happens between the trigger wheel teeth and the VR sensor:
As a tooth on the trigger wheel approaches the VR sensor it induces a positive (+) signal that becomes stronger until it peaks just as the leading edge of the tooth reaches the sensor, then the signal begins to descend until the tooth is centered across the sensor. It is exactly at this point, when the trigger wheel tooth is exactly centered with the sensor, that the signal "crosses zero" and enters the negative (-) range. As the tooth continues past the VR sensor the signal tails off to almost nothing until the next tooth arrives at the sensor. It is this "crossing zero" effect that the Microsquirt uses to count the tooth.
For those of you who appreciate the math behind things, here's a breakdown of the actual timing involved in my Bandit's trigger wheel:
The Bandit 400 can rev to 13,500 RPM. There are 60,000 milliseconds in a minute. So 60,000 divided by 13,500 equals 4.45 milliseconds for each 360 degrees of engine rotation. There are 6 teeth on my Bandit's trigger wheel (5 real teeth and 1 that's missing but the Microsquirt knows about) at redline engine speed (13,500 RPMs) there's a tooth passing by the VR sensor every .74 millisecond. Or instead of .74 milliseconds, since there's 1000 microseconds in a millisecond you can say a tooth passes by the sensor every 740 microseconds. Or you could go the other direction in magnitude and say that there's a tooth passing the sensor every .00074 seconds.
The thing that really amazes me is even at that speed my Microsquirt ECU isn't working hard. At those same RPMs it could easily handle a trigger wheel with a lot more teeth (I'm impressed, but I'm a complete nerd that way).
But, back to my problem with the Bandit's trigger wheel...
Today it's apparently "industry standard" to have trigger wheel teeth that are approximately the same width as the permanent magnet in the VR sensor. This matching of tooth-to-sensor sizes makes for a very clean and definite "tooth passing by" signal and this is the sort of signal that modern ECUs are designed to receive and interpret and, most importantly, this is what my Microsquirt's circuitry is expecting to see.
It appears that the Bandit's "state of the 1990s" Denso CDI box circuitry depends on other portions or events in the VR sensor's electrical output for information, or maybe it is simply immune to the possible side-effects of a really wide tooth.
During cranking (under the electrical starter-motor motivation) an engine doesn't maintain a constant speed. Instead, the compression stroke of individual cylinders causes it to go through a "slow-down-speed-up" cycle.
My Microsquirt was having trouble correctly interpreting the trigger wheel teeth during cranking, which is a complete show-stopper. The Microsquirt was double and triple counting the teeth that were passing by the VR sensor during the moments when the compression slowdown-speedup was happening. And with the Bandit's inline 4-cylinder engine a piston reaches TDC on a compression stroke every 180 degrees of engine rotation.
If the Microsquirt cannot properly interpret the trigger wheel teeth it won't be able to "achieve sync" which means it can't match the trigger wheel pattern it is sensing to the trigger wheel pattern definition that I've programmed into it. When this fails to happen the Microsquirt will not command the ignition coils to fire or the injectors to squirt fuel. The engine will not start.
Here's the part that I feel dumb about: The MegaManual, the main document which outlines the "how to" of do-it-yourself fuel injection, says that the trigger wheel teeth should be approximately the same width as the VR sensor. I don't know why I ignored this directive, it was right there in black and white. I'd read the MegaManual several times over so this information was in my head (somewhere) the whole time I was building my Bandit.
But for some unknown reason I ignored the information until I reached an impasse. Then a couple of days ago, when I was talking to my brother on the phone about the issue, I suddenly realized what I needed to do.
I went back down to the garage, took the left-side cover off of the Bandit and cut the teeth down from 13mm to 6mm without taking the trigger wheel off of the engine. I carefully covered everything with plastic wrap and aluminum foil so the grinding/cutting debris wouldn't get into the engine and oil. Then I put the cover back on and when I tried the starter the engine fired right up.
For the "visual learners" I'll put up a couple pictures of the trigger wheel, both before the tooth modification and after.
