Phil's missing teeth decoder dev thread

[johnsop1]

Hmmmm,

My LPT port sym. must not be doing a good job simulating a 36-1 signal.

Any chance you could post a picture of a scope trace off the 36-1 waveform from the jimstim?
Then I could fix up my LPT port sym. (I am too cheap to bye my own jimstim.)

[johnsop1]

It crossed my mind that I may have set the polarity of the missing tooth pulses wrong. I have pushed a version to git that syncs to the opposite polarity.

If you get a chance give it a try.

[johnsop1]

People writing other decoders may want to reuse code for the NipponDenso.c decoder. If the ignition and injection code were functions each decoder could just make a call to generic ignition and injection functions. I realize that we do not want the overhead of function calls in the ISRs, that is why they would be defined inline.

[Fred]

I have the polarity of my rig setup wrongly, but it should have worked when I inverted it in that case. I'll try it now and see what it does, then I'll crack out the soldering iron AGAIN and give it another whirl. I believe I may have scoped the output some time ago, let me check.

http://i260.photobucket.com/albums/ii15 ... rigger.png

Note that I had the settings on the JimStim inverted for that and that is why the long part of the second trigger is high and the pulse low. Same for the missing tooth. It should be low, not high.

Fred.

[Fred]

People writing other decoders may want to reuse code for the NipponDenso.c decoder. If the ignition and injection code were functions each decoder could just make a call to generic ignition and injection functions. I realize that we do not want the overhead of function calls in the ISRs, that is why they would be defined inline.

Yep, 100% right, in fact, it will be mandatory to use some shared interface like that. I just haven't got around to it just yet. I'll get around to it soon. In the mean time, feel free to dirty yourself by copying and pasting the appropriate parts :-)

Fred.

[johnsop1]

I believe I may have scoped the output some time ago, let me check.

That waveform looks funny to me. I expected to see the missing tooth portion to consist of a high and low pulse equal to the period of the normal tooth.

http://www.megamanual.com/ms2/wheel.htm

Maybe someone out there has a scope trace from a real wheel.

[Fred]

Same results exactly. I couldn't pick a significant difference between inverted and right way up, but something tells me right way up (positive pulses to pin = 1) syncs slightly more often than the opposite. Bear in mind, both are just flashing a bit. I could take a video, but it's barely worth it.

I await your future pushes :-)

BTW, you don't have to use the pull request feature, any form of communication is sufficient, this thread, email, pm, pull request, phone call, text message etc none are better than any other.

BTW2, you blew away 2 github repos AFTER I pulled from them! This is generally a big no-no. Once you publicise a change you have to live with it unless for some reason you are sure no one pulled. It doesn't really matter, except that if you look back at the history you can see some of your commits multiple times and it makes the graph look funky. Best avoided in future ;-)

Fred.

[Fred]

That waveform looks funny to me. I expected to see the missing tooth portion to consist of a high and low pulse equal to the period of the normal tooth.

http://www.megamanual.com/ms2/VR_missing_tooth.gif

That red line is with the polarity incorrect. Such a signal would be near impossible to sync to during accel and decel due to the dynamic change in tooth width.

If you look at the bottom of it, the wide part (also marked with red lines, but thinner) is where the two leading edges would appear when processed by a circuit like the B&G one or lm1815.

Check this out : http://trigger-wheels.com/store/index1.html

I hope that clears it up :-)

Fred.

[johnsop1]

I guess my point is I expected to see two wide pulses. The picture that you posted has only one.

My LPT sym looks like this.

[Fred]

The signal you show is what you get when you hook up your VR with the polarity back to front. It would be extremely hard to detect that at speed even if you meant to :-)

Fred.