A somewhat odd project.

[LAW88]

Hi all, I like to ride motorbikes fast, one unique advantage of a bike is the ability to shift gears without the clutch. Simply 'flick' the throttle closed for a split second to unload the gearbox and engage the next cog, all this is very fast but easy to get wrong and do damage to the bike or cause a crash.

A very simple (but expensive) option is the "quickshifter", basically it senses when the rider selects the next gear (by a micro on the gear shaft) and cuts the throttle for a few thousanths of a second. This means that the rider can hold full throttle and click up through the box very quickly.

I have a micro (taken from the redundant brake light pickup in the brake lever) which I can mount on the gearbox but what sort of IC would be capable of cutting the ignition circuit for a short (and adjustible) period of time. My initial thought was simply to use a capacitor with a rheostat but as it's effectively an AC circuit I would simply be supplying a delayed current to the spark plug (not good).

I only have a highschool understanding of electronics and that was a few years ago now so small words are best

Thanks everyone.

 

[bychon]

A simple way would be to wire the micro as a kill switch. It would only kill as long as your foot holds the shift lever off it, way less than a second. The other way is to use the micro to start a timer that triggers a kill switch for a tiny part of a second, and you better finish the shift in that amount of time. Is this what you're after?

 

[LAW88]

yeah, I was trying to avoid using the lever as a killswitch directly as this sort of defeats the purpose. some sort of solid state IC or relay with an adjustable killtime would be perfect

 

[LAW88]

around 50-100 milliseconds is good.

 

[bychon]

There are haffa dozen ways to do this. The timer is the easy part. How to kill the spark is the detail that is the devil. What kind of ignition circuit? Ballast resistor to coil? Capacitive discharge? MSD system? Timing pulse from magnetic gear sensor? Dual point distributor? Opto-coupler in distributor? Must have diagram to design interface.

 

[LAW88]

Does this help?

**broken link removed**

I'm thinking of breaking the circuit on the low side of the coil.

 

[bychon]

Do the L or G terminals kill the spark? That is, does putting it in neutral or using the side stand kill the spark?

If they do, that gets us out of the high voltage, high current, inductive reactance area and life gets simpler. Slapping a ground on a pin that is meant to be grounded is easier than getting into the coil section. If it has to be stopped at the low side of the coil, the stopper will have to survive the full voltage at pins B/Y and W/L because it will come right through the coil primaries and try to get through the stopper. I've read somewhere that a CDI can hit a coil with 400 volts. That makes building the coils easier, but it makes stopping them harder.

 

[LAW88]

yep G is a killswitch, brilliant, I'd never thought of using it

 

[bychon]

OK. What you need is a 555 configured as a one shot timer. You have to set up a microswitch that is ready to switch as soon as the lever moves. That change is going to be fed through a capacitor to a 555 timer configured as a one shot oscillator. The capacitor is so the signal doesn't keep on happening for as long as the shift lever is up (or down).

1)The switch goes "on".
2)The capacitor goes "pulse".
3)The timer goes, "kill"..."resume"..wait.
4)When the lever returns to rest position, the capacitor empties before the next shift.
5) go to 1 and start over.

From here, it's which polarity, how much time to kill, how much time to empty the capacitor, what voltage and current are going through the side stand switch, what transistor will handle that current, how will the 555 timer drive the transistor.

Are you taking notes?

 

[LAW88]

ok...I think I know what you did there. But mainly not.

The sparkys at work don't usually do this sort of work but maybe if I give them your instructions they could build this for me. Thanks heaps, big help

 

[shortbus=]

bychon, one thing you forgot was switch debouncing. With out it you could kill the ignition for longer than you wanted, from oscillation of the timer.

 

[crutschow]

If the 555 is configured to not re-trigger, then the bounce should be over by the time the 555 has timed out.

Question: If the microswitch detects shifter movement, hasn't the shift already happened by the time the microswitch detects it?

 

[nickelflippr]

How about a force sensitive resistor divider similar to this one? If you could sneak one each under the shift peg rubber (top and bottom), you could add some gear change triggering intelligence.

 

[bychon]

I'm kinda waiting to see if OP has locals that solder b4 I get into debouncing...and I was trying to remember if a 555 is debouncable as a one shot. (Darn, I'm getting old!)

The shift lever has most of an inch of movement before it touches a gear...probably milliseconds. Then it starts shoving gears around inside the transmission. The best part is that it only has to work during upshifts because you're trying to slow down (off the throttle) anyway when downshifting...I hope.

 

[bychon]

I am not OK with trying to fit force sensing resistors under the rubber right now. The rubber is supposed to have some elasticity to it, so measuring how much the rubber affects the sensor, designing a threshold detector (which might be in the 555 chip), etc. This OP can just about solder. I don't want to teach him how to operate an R&D lab.

Notice that I changed a couple of words in post #9

 

[LAW88]

yeah, shift lever has a fair amount of play so it's not hard to rig the micro to trip at the appropriate time. and it is only for up shifts. for clutchless downshifts we 'blip' the throttle to bring the engine revs close to the wheel revs which makes it all mesh smoothly, but that requires practice and bike specific knowledge so I don't think a basic circuit could help.

Thank you Bychon, I appreciate all the help.

This guy's throttle sensor shows what I'm trying to achieve.

 

[LAW88]

ok, so I've done some research (10mins on wiki) and I think I can see where you're going with this.

**broken link removed**

would a Rheostat at "R" make the "killtime" adjustable?

also are you suggesting making the pulse trip a CB on the killswitch? Sorry to drag this out.

 

[bychon]

Yes, the kill time will be adjusted with R1. But next you need to measure the side stand condition.
1) is G shorted to ground to run or to kill?
2) when the connection to ground is made, how much current flows to ground?

You see, the circuit is actually designed by thinking backwards. First, what has to be controlled, then how do we control it, then how do we make sure the nasty, noisy, motorcycle generator voltages don't eat the brain chip. After "what gets controlled" and "what will control it" gets done, the answers are apparent to the question, "how to safely feed the controller how much voltage and how much current".

I'm not talking about the shift switch and capacitor yet because they are easy. Feeding the G pin to ground is what I don't know.

 

[bychon]

When G is not grounded, how much voltage is trying to get to ground?

Sorry, forgot to ask that.

 

[bychon]

I just figured out, "trip a CB on the kill switch". Circuit breakers are way too slow, but the controller that replaces the side stand switch will break the circuit and then reset...really fast.