J-series Acura/Honda

[TonyS]

Hi Fred / Peter,

As I am both interested and ignorant regarding the interfacing of cam and crank gears to VR and Hall sensors. Can you please explain -

Why was the gear cut down (effectively broadening the tooth face)? Was the original tooth width outside the specifications of the Hall sensor that was picked? If one chooses to not use original equipment, are there certain gear(s) / sensors / relative position combinations that are known to work well?

Why so many teeth? What is the primary purpose of the cam signal?
I thought that it was to determine when you were either on a compression or exhaust cycle of a reference cylinder in order to synchronize the information coming from the crank position sensor (the primary timing source)?

I underand (I think) the relationship between a VR sensor signal and physical gear tooth location. Is the Hall type sensor less / more forgiving of distance / specification mis-matches than VR types? What does a "perfect" Hall sensor / gear tooth signal look like? Is it a brief negative spike at the exact center of the tooth or is it more broad with software determining tooth "center"?

There is so much information on this forum (which may already contain answers to my questions), but it does take some time to "search" for individual topics of interest. Having a DIYEFI-manual which addresses individual topics and consolidates information I think would be nice to have sometime in the future.

Thanks,
Huff

[Fred]

Got google talk or skype account? Be good to hear your voice! Can answer all of that quickly in voice. Not typing it, though.

EDIT: For Tony: viewtopic.php?f=54&t=1523

[Peter]

Why was the gear cut down (effectively broadening the tooth face)?

Mostly because Fred told me it would work better. I'm sure there is an ideal tooth width for a given sensor. The point of the tooth effectively doesn't allow enough metal close enough to the sensor to change the magnetic flux enough for the sensor to work correctly.

Was the original tooth width outside the specifications of the Hall sensor that was picked? Why so many teeth?

They really don't give much in the way of specifications. I bought this trigger wheel and hall sensor before I knew any better when I was still trying to build my own open source engine management system.

If one chooses to not use original equipment, are there certain gear(s) / sensors / relative position combinations that are known to work well?

I would strongly recommend oem hardware if it's capable of doing what you want. The Honda has a 6-2 Cam trigger with 12 even teeth on the crank. Right now there isn't a decoder that will run that setup. So instead of coding a new decoder I decided that I could test the board on my truck since I already have all the hardware from my attempt at an engine management system to put it on there. So the only reason I'm using an aftermarket trigger is because there isn't an oem one. At least not in the fuel injection sense. The 6-2 setup is far too course for ignition timing, but it runs the fuel just fine. 60-2 needs optimized in the FreeEMS code so that it can rev higher. So probably somewhere between 12 and 24 teeth is optimal for this system right now depending on what you're trying to do. I've switched from the hall sensor to an old VR sensor from one of my previous Honda engines so that the hardware doesn't have to change when I put it on my Honda.

What is the primary purpose of the cam signal?

You're right except for there being no crank sensor on my truck for the primary timing.

Is the Hall type sensor less / more forgiving of distance / specification mis-matches than VR types?

I honestly have no idea. One of the reasons I'm not using the Hall Sensor anymore is because I was getting some really strange oscilloscope traces from it. I'm planning on making a thread in the pit lane when I get done setting this up today.

What does a "perfect" Hall sensor / gear tooth signal look like?

A 5V peak to peak square wave.

Is it a brief negative spike at the exact center of the tooth or is it more broad with software determining tooth "center"?

It really doesn't need to be the center of the tooth. Call the rising edge the "center" and offset the sensor 2 degrees to match that. In other words use a timing light to account for the sensor offset and any lag in the hardware. VR sensors are the ones that look for the center of the tooth. Which happens when the magnetic flux against the tooth is the same on both sides of the sensor. The important thing is that the teeth are very uniform so that the sensor gives you a consistent uniform output.

There is so much information on this forum

I know, the learning curve seemed relatively sharp for me. I think it's to keep the hand holding to a minimum until the code base is ready for every Tom, Dick, and Harry to be able to use the system. Occasionally I consider opening new accounts to PM Fred really noob questions like: Where can I buy a plug and play board for a 2013 Mustang?, Is this just like MegaSquirt only cheaper, and if so where can I get a board?, or Will this code run on a MegaSquirt board? Just to see if I can piss him off. lol

[Fred]

ROFL!!! :-)

Two factors: 1 what you said 2 more important things to do.

Perfect hall is square/flat waves, but can be 12v or 5v or any other pull up V

VR is the better choice with an appropriate input conditioner.

Spoke to Huff on skype earlier. Thanks for your post!

Fred.

[TonyS]

Thank you very much Fred and Peter. I now how have a much better understanding of the relationship between the hardware (gear / sensor) and software (especially the "decoder" term).
Huff

[Fred]

NP! :-)

Peter: viewtopic.php?f=8&t=1708 sorry!