After putting a bunch of time into cleaning up this drawing for a thread over at msextra.com, it occurs to me that it never did get stickied here. Prolem solved: Note that it's been suggested that the TL082 opamp may yield better performance than the LM393. The two are pin-compatible, so if possible you should attempt to substitute a TL082 in this circuit instead. The TL082 is available from RadioShack as P/N 276-1715, so this is one more thing in its favor. As drawn, this is for MS2. You could use it for MS1 as well, by changing JS10 to JS8 on the B-side output.

__________________

Noob alert: Is this made on a separate board (like a breadboard)? This may be harder going MS-II than I thought.

If you are careful about your layout, it will easily fit into the proto area of the 3.0, with plenty of space left over for a couple of relay drivers. Abe was good enough to do a nifty 3d layout of the assembled circuit:




Of course, you can if you wish to it on a seperate board, as etnad is doing:

(note that I do not condone this hideous mess of resistors. Are you listening, etnad? Clean up your goddamn wiring!)


When I abandoned the CAS and went to a wheel, I actually built the new decoder circuit in such a way that it is completely external to the MS, which made it easier to fine-tune. It's in a small external box, which tethers to my MS via a 3 wire cable as seen here: http://www.miataturbo.net/forum/show...884#post251884

__________________

Ah! That is a mess! Lol. I rather build it in the proto area so everything can fit onto just the one board. You only had to build the separate box since you had a custon wheel right?

Yeah. As I explained in the writeup, when I assembled my MS I figured I'd never need the VR circuit, so I left those components out and glued a very large capacitor down to that empty space on the board. So in the end, when I wound up needing the VR circuit after all, I had no choice but to build it on a separate board. (my proto area was already packed with other circuits, including my previous dual pickup circuit)

I decided to put it in an external box since I has just been through a very frustrating experiance with my previous ignition trigger circuits, and I wanted something that was easy to remove, test and modify if need be. As it turns out, it failed horribly the first time, but worked perfectly thereafter, once I flipped one connection around and re-installed it.

But the vast majority of folks build their pickup circuit right on the main board, regardless of what type of circuit it is.

__________________

isn't his last name Mara?

did abe do the bottom one?

Uhm... ****. I honestly don't know- I never bothered to steal his wallet and check his drivers license.

Abe, have I been mis-spelling your name all this time?

you think I read?! you know me better than that. I only asked because I've seen some of abe's circuits....this etnad character must have been his padiwan.

Oh, I was responding top Y8s' question about the spelling of his name, You snuck in right before I hit send.

What do you mean by "the bottom one"? The 3d rendering?

Tee hee hee

http://abefm.smugmug.com/gallery/262...38616327_GWT5v

Oh yeah. I remember that day well. We were coming around a big, fast turn, and I was trail-braking. Initially started heading for that wall, then over-corrected, spun 180, and exited the road backwards into those trees which, thankfully, stopped me without doing too much damage to the car.

Shortly thereafter I became a bit more serious about my tires and suspension. Needless to say, the T1Ss went away, as did the AGXs, the big rear bar, etc...

You can cut the parts counts drastically by using a 74C14 (Schmitt trigger input inverter).

And your heavy noise rejection filter will delay the signal by 4° at 7200 RPM. Not good.

Or a CD40106 (which is very similar to a 74C14)

Like so.

You can increase the 470Ω resistor to ~1kΩ and the delay will be <0.5° at 7200 RPM.

Attached Images

input.gif (2.7 KB, 917 views)

This circuit was the result of Abe reverse-engineering the input circuits in his OEM '99 ECU. It is not an attempt at optimization, merely an exact duplicate of the factory circuit.

Delay due to the input filtration can be negated in software, as there is a specific field for entering the hardware latency of the trigger input.

but isn't the delay frequency (rpm) dependent iow, a fixed latency has no(t a lot of) use?

I'm a bit at a loss here. As I understand it, the MS2 has the necessary software for the NB sensors, using this circuit.
Can the NB sensors be used on MS1 as well or does its software not allow this? It prefer MS1 over MS2 if possible (because of better idle and WI control).

Well, I grant you that as f approaches ∞ there will be a phase shift in the voltage across the capacitor, however in practice this seems to be relatively small by way of comparison to the fixed latency of the circuit.

MS1 software doesn't support the NB sensors.

The MS1 software cannot be made to understand the NB's unevenly-spaced crankwheel. One could make it work by trimming the two TDC teeth off of the wheel.

It's also been postulated that, even for an NA, this circuit is likely more reliable than the ones commonly used for 4G63 input. A number of NA owners (myself included) have found that CASs which worked perfectly on the stock ECU exhibited severe problems when used with a MegaSquirt. I'd like to see the next person who runs into that problem try this circuit.