3. Get someone off your ass by thinking you are braking a tad bit earlier then normal.

 

true, I know plenty of racers that will turn them on 100ft before they get to the braking zone just to get you to brake early.

 

Damn Justin. Thanks man.

It works in drag-racing

(I'm learning a lot as usual Mike)

 

They're used all the time for shift cuts across a huge range of motorsport, and are just as reliable as any other method.. Like anything else, they're only as reliable as the implementation and the interface hardware (and software). Switches work also, but have problems with consistancy and mechanical hysteresis. But, they don't require an interface amplifier or any signal conditioning like a strain gauge would. So switches are a lot cheaper. The really nice thing about a strain gauge is that the shift detection point can be changed in software because the strain output doesn't have a binary threshold like a switch does. It's just like the difference between using a temp sensor for temp detection instead of a thermal switch.

Load cells are used for that too, but they're usually just strain gauges already integrated into a mechanical housing. There are some cool ones on the market that install inline with the shifter, so you just cut the shifter (and remove the length of the load cell) and either slip the load cell over the shifter sections (for the clamp type) or thread the shifter sections. Then you have a imperceptible hinge point in the middle of the shifter with a strain gauge on it.


About the brakes: besides clearing pad knockback, there are a few other reasons for a quick dab before the corner. Sometimes it's just done to get a tiny extra bit of surface temp into the pad for better initial bite with race compounds, sometimes to clean dust, sometimes to clear moisture (under the obvious conditions).



I love this thread, it's full of things I like to talk about.

 

Ok I like where your heads at lol! Help me out here though:

I like the idea of having the strain gauge/load cell inline with the shift lever. Not sure how else you would do it mechanically? But how do you get the strain cage to just trigger the ignition cut for a very short period of time, so that the ignition is not off the entire time your pulling/pushing on the lever? The reason being that you wouldn't want the ignition to be cut while your trying to blip the throttle on downshifts. I would assume you would just figure out the force required to pull/push the lever out of gear and use that force as your force required to trigger the cut. Assuming the force is highest when your first push or pull on the lever. Does the strain gauge just splice into the coilpacks or do you need to use something that would allow you to utilize some simple logic? In the end all you want to do is release the torque on the mainshaft to allow the dog to slide on it's spline and engage the other gear.

Thanks for the input man.

 

I do speak up on occasion

 

The other ways to do it mechanically are to use a switch (multiple switches actually) that the shift lever would contact based on the direction it's moving. A strain gauge inherently measures strain, so you can bond one to a shift lever and measure strain when the shift lever bends (when force is applied). That's how strain gauges attached to suspension members or driveshafts are normally used. You'd probably want to make a shift lever with a little bit of compliance in it just to make things a little easier, but that type of bonded-on strain gauge is a lot less straightforward to design and install than the pre-packaged load cell shifter setups so I wouldn't consider them.

Normally you'd interface the strain gauge with the ECU, and usually the strain gauge would need it's own electronics for amplification which would then connect to an ECU input. If it's a flexible motorsport ECU, you can put the voltage thresholds into the ECU calibration, along with any logic that would be needed to do what you need to. For example, you could have different strain thresholds for different gears, or have different strain values depending on whether you were downshifting or upshifting. That's pretty easy with a sequential transmission because you might already have a position encoder or potentiometer on the shift selector drum so the ECU can detect the gear position, or be able to detect if you were moving the shift lever forward or backward.
An ECU that can do gear detection can usually be programmed to change the amount of ignition energy cut depending on the gear or the throttle position (which is a rough approximation of engine load) or RPM. So you'd set it up to have a certain ignition energy reduction based on the strain threshold (input signal voltage), RPM, throttle position, gear, etc. It's not just one fixed amount of cut energy, it would obviously have to be a vector or a map if you wanted to take those parameters into account.

The Ferrari, etc., paddle shift transmission controllers have entire 3D maps to change the ignition energy, ignition timing, and fuel depending on those parameters, because those parameters have a huge effect on shift harshness (NVH, jerk, etc) and drivetrain life. That's all the way at one extreme though, and not really necessary for a track car with a dog box. You just need something to take the load off of the dogs (like you said) so it doesn't have to have lots of bells and whistles. Gear-based ignition cut is a very common thing to do with motorsport ECUs though, so many ECUs already have some functions built in right off the shelf.

Anyway, this is something that you need a decent ECU for if you have a fuel injected engine, or at least a fairly sophisticated ignition controller if you have a carb'd engine. There are some piggyback shift cut systems that go inline with the coil drive signal, but they're obviously limited in their sophistication compared to doing it directly with the ECU. Obviously the more signals you have access to, the more complex management strategy you can create.

If you're using the stock LSx ECU, you can probably find a piggyback that attaches in series with the 8 low voltage logic signals that go to the coils (those coils have the coil driver circuit built in IIRC) and reduces the duty cycle based on whatever parameters you can set. Basically the same thing those shitty Power Commanders do with the fuel injector signals on a motorcycle.

What are you using for an ECU? I haven't read back through the rest of this thread after the 1 year revival, so you might have already said..

I've done enough brain dump for today, I'm going to bed.

 

Definitely using the stock LSX ECU for now. A piggy back setup sounds interesting.

 

Sent some info, check PM..

 

roger that i'll reply back to you there after I have a chance to read your PM's.

Got the car in the garage today so I can't start mocking up the engine mounts. Also got my long shifter in today for the Tex T101. Looks like a nice piece.