J-series Acura/Honda

[Fred]

If you mean something like this: https://fbcdn-sphotos-a.akamaihd.net/hp ... 1130_n.jpg

Then the flattish bits are the gaps between the teeth. The teeth produce a nice sudden change. As I explained to you yesterday, with a missing tooth setup, you get the "gap" split in two if you wire it backwards, however on an even teeth setup (such as your crank) it's still important to wire it correctly because if you don't the timing will vary with RPM for the reason you mention. When it's correctly wired that doesn't happen, though.

The issue with your wheel is uneven flux. All missing tooth setups suffer from it to some extent and various designs compensate in various ways.

http://www.crespocams.com/TrigWheels/TWbig.jpg
http://vagner.jeje.je/bilder/albums/alb ... _wheel.jpg

See how its raised compared to the other gaps? It's to maintain the average magnetic flux such that the quantity and timing of the next flux change is correct. It helps with balance too.

This dude has it setup wrongly and will get varying timing as the slope of the between teeth section changes with RPM:

http://i142.photobucket.com/albums/r116 ... k00005.jpg

Toyota dizzy with single tooth and long ramp to position flux correctly for maximally steep voltage change:

http://i260.photobucket.com/albums/ii15 ... 0_7212.jpg
http://i260.photobucket.com/albums/ii15 ... 0_7215.jpg
http://i260.photobucket.com/albums/ii15 ... 0_7218.jpg

There are a few aspects to that that may not be easy to see:

1) A steep drop on one side of the tooth
2) A horizontal taper from high to low
3) A vertical taper from narrow to wide

In any case, the issue with yours is related to the missing teeth, and maybe that missing piece in the middle too. They cause the flux change to happen slightly differently than if the pattern had been 100% consistent. I'll take a look now and see what I can guess for you. You'll have to test, though, and it may only improve things in some areas, though likely it will help, to a varying degree, everywhere.

Fred.

[Fred]

I just pushed some updated angles to that temp branch. Build that and try it on the car. If it's wired up to run, so be it, it should still fire up even if I made it worse. Get a good log of the start up, and a video too :-) If you can be on a skype call to me when you hit the key, bonus points :-)

[Peter]

I'm in class right now, so I'll work on it when I get back. You've kind of got me thinking along the lines of the system having critical or tighter dampening. So that it doesn't overshoot the zero boundry when it comes up to the missing teeth. My biggest problem is I don't understand how that would be a problem with an optical distributor.

[Fred]

It's not about damping, it's about energy, and energy expended. There is energy generated by the coil due to the magnet and the passing iron body. That energy has no where to go, so the voltage shoots up, and then dissipates the energy into whatever is connected. The trouble is that the dissipation before the close together teeth is less than that during the large empty section, so when the tooth comes after that, you get a different part of the wave form going through zero, and thus it's a bit earlier/later. The other trouble is that that missing bit generates a voltage too, though weaker due to the distance from the sensor, both positive and negative depending on which end you're at. Don't over think it :-)

The angles I provided should be decent. If they're not, we'll revert to normal or try some others. In a few weeks we'll ditch this idea all together anyway. It'd be nice if your logs were smoother as the engine will run better due to a steadier RPM signal. That's the thing, not the angles. We're only talking about 1% or so anyway, I'm just a perfectionist ;-)

Fred.

[Fred]

And the optical is a different issue, it's diffraction, so the slot is wider than it measures because the light creeps around the corner and switches on the receptor earlier than the edge and keeps it on later than the other edge. The similarity was confined to expected angles vs. reality, that's all :-)

[Peter]

I don't know how to stress this enough. DON'T USE A BDM TO POWER YOUR TA CARD WHILE BURNING THE FLASH!!!!! It doesn't work. Fred and I just spent hours trying to figure out what the problem was. http://youtu.be/DPxbxBzXwMY

[Fred]

I would like to stress, in addition to that, do not use MTXLoader to do your loads unless you KNOW your serial shit is good, as it does NOT verify what it loads.

CONGRATULATIONS PETER!

Fred, going to bed.

[Preston]

I would like to stress, in addition to that, do not use MTXLoader to do your loads unless you KNOW your serial shit is good, as it does NOT verify what it loads.

CONGRATULATIONS PETER!

Congratulations Peter, Nice WORK!

+1 on MTXLoader comment, even if your serial doesn’t work, it still appears to load sweet (WTF really? why.. why?). SeanK's Loader pointed this out for me thanks Sean for a meaningful "Serial comms error" message.

Preston.

[Peter]

Congratulations Peter, Nice WORK!

CONGRATULATIONS PETER!

Thanks Guys!
Another short video of driving it on a beer run. http://youtu.be/eUWLKzSib_4

[Fred]

You're a legend! Just don't blow the fucking thing up, testing is good! :-)