Sequential COP questions & research
 

Yes I did a search, and no, the answer is not clear.

I was wondering if you guys who are running COP with MS1 on an NA, run it in sequential mode, or in wasted-spark mode? I read this and its not clear whether this is possible with the CAS?

http://www.msextra.com/manuals/MS_Ex...l.htm#twotrigs

(scroll down to sequential COP).

I think everyone here runs wasted spark with dreams and hopes to be able to figure out how to run sequential.
If you figure it out, make sure you share asap :)

MS1/Extra can only run them in wasted spark mode with the stock CAS; you'd have to make some creative mods to the CAS to get them to fire sequentially. The MS2/Extra 2.0 beta code can fire them sequentially.

or, maybe using a crank wheel with the existing 2nd trigger input?

on the 1.8 magnetic CAS the mods should be easy enough. there's two sets of sensors. And the plates are both independent so you can just goof around with them until the pulses are in the right place. Unless of course the pulses from the two signal out wires are phased properly already.

http://members.aol.com/solomiata2/Miata18CAS2.jpg

However you set it up, for COP with the MS1/Extra code you'd have to get 1 and only 1 pulse on the 2nd trigger for each cam revolution.

I havent seen the inside of a CAS but from the DIY website install article, the inner ring of holes has only one hole. So wouldnt that make it one CMP pulse per 720 deg. crank rotation?

On both the magnetic and optical NA CAS, the CMP signal is one pulse per crank rev, two per cam rev. The two CMP pulses are not of equal duration however, one is high during both the leading and falling edges of the corresponding CKP pulse, the other goes low before the falling edge of the corresponding CKP pulse. (or maybe it was the leading edges of CKP... I've got the printout at home.)

Either way, what you'd need to do is to modify the CAS wheel to eliminate one of the two CMP pulses. I believe you'd want to end up with the one that occurs just before #1 TDC. On the magnetic version, you could simply cut off one of the two teeth. On the optical version, you'd need to fill in one of the slits.

Here's what the inside of the optical CAS looks like:
http://members.aol.com/solomiata2/Miata16CAS.jpg

you can either cut off the teeth or build a circuit to "ignore" the second pulse. say with a T flip-flop.

Since its an optical sensor on the 90-93 and it doesnt care about proximity, maybe there is enough space to solder or maybe glue a tiny piece of aluminum sheet on the backside of the existing plate?

I need to get a spare CAS and try this myself.

On the MS two additional outputs need to be brought out to the harness.

you can get thin aluminum tape. home depot.

But the flip-flop had better know which pulse to ignore, or your coils will only fire correctly half the time you crank up the car.

My only concern with these mods is losing the balancing of the wheel. Either via adding weight (tape on the opto CAS) or losing weight (sawing off a flux shield on the magnetic CAS).

Jim

I posted this on the Extra forum today:

http://www.msextra.com/viewtopic.php?t=28252

Maybe I should've asked here first.

I think sawing off the magnetic CAS seems like a more reliable option, PROVIDED 4 pulses are enough on the CKP signal.

If an escort GT distributor is similar to a 94-97 CAS I'm gonna find me one in the yard today.

Ok I went to the yard and picked up a CAS. I have some interesting findings for you guys:

1) A Ford Festiva has the exact same optical CAS as the 90-93 miata, except that there is only one inner slot. It looks the same as the pic Joe posted, except that the inner lower slot isn't there. This would give the required 4 + 1 pulse.

2) I picked up a Ford GT distributor which is the exact same magnetic CAS as the 94-97 miata, except that there is only one inner "ear". This would give a 4 + 1 pulse as well.

So all those who're looking for a cheap CAS go to the yard and find yourself a Ford Festiva. Nobody wants parts off of those. Got the CAS for $20.

Sequential COP, here I come!!

escort gt distributor...

http://i6.photobucket.com/albums/y24...s1/ltsdist.jpg

There are two types of escort GT distributor, the one above and the magnetic type posted by y8s, except with one inner ear.

The point is, now that the requisite signal is in place, how to get 4 outputs out of the MS. I'll be researching and implementing this in the coming week.

This is the distributor I have:

http://img507.imageshack.us/img507/6507/dsc00398ti4.jpg

http://img87.imageshack.us/img87/6708/dsc00397tv1.jpg

I plan to install it this weekend, run with the stock coils for now, and then upgrade to COP.

that's pretty awesome. saves you having to mod anything!

Pipe, seriously...you think this will work?
I am very ignorant when it comes to electronics, but I do inderstand when someone explains something.
So, are you saying that, if I get an Escort GT CAS (I have a 94 1.8) that it will fit right in place of my CAS and i'll be able to run two extra wires (which you will explain I am sure) from my MS (standalone) to my coils and have sequential ignition?
Will this be possible on 94+ cars?
Will this work on parallel installs? I am asking for others...
Will this work on NBs? Also for others...

Thanks,
Dan

Now this is a solution! Instead of buying a ubber expensive Miata CAS we just but an escort dizzy! Please tell me you tested this and it fits between the engine and firewall!

Not tested yet, but the height of both CAS is equal. But it likely won't work with the stock ECU.

Dan, I'm no electronics guru either, but its easy enough to make it work. You need to run extra wires from the MS, enable Spark Outputs C and D, and reconfigure your wheel decoder settings, but I still need to figure that part out. Will document everything as I go along. I don't see how things are different for 94-97, or on 99+ except for the latter's different crank angle signals.

Wheel decoder settings are pretty straightforward: Set it to a 4 tooth wheel, trigger A to 1, B 2, C 3, D 4, and set the returns to zero. If running wasted spark, set it to dual-distributor mode.

Matt, shouldn't it be:

Base Teeth = 4
Trigger Pos

A = 2
B = 1 (These are your PNP settings)
C = 4
D = 3

Dual Dizzy mode will fire A,C on one coil, B,D on the other coil. Assuming all stock wiring, ignitors and coils.

Ok, I'm still hanging in there so far. The new wires I'm running from MS are trigger wires, correct? This would not be an issue for me. The issue is now, here, figuring out what goes where and what order I have them fire, but I am sure you will figure that out way before I start comprehending this. So, we are on the verge of breaking seq. ign. done pretty easily and cheaply? Awesome, I feel enlightened to be able to read this thread, let alone to contribute to it.

What Matt is saying, I think, is that since you have separate trigger wires for each individual coil (COPs) you can run it that order. 1-2-3-4 since you control each coil separately and are not running one coil per 2 cyl.

Right. The two wires are trigger outputs, which will require you building a pair of extra output circuits substantially identical to your current two. If you're using the "book" mods, this would be a pair of 2N3904s wired up just like the LED drivers. (you can omit the actual LEDs)

I'm pretty sure that the wheel decoder would be set as Matt suggests, to base = 4, A1 B2 C3 D4, with second / none missing.
Reference: http://www.msextra.com/manuals/MS_Ex...l.htm#twotrigs and scroll down to
"Sequential COP
4cy ONLY"

Note that this example applies to a four tooth crank wheel with a separate cam sensor, so it assumes four pulses per crank rev.

Dope. :)
What year festiva?

A1 B2 C3 D4 is right for seq. COP, but I was talking about triggering the stock coils in wasted spark mode with the 4/1 CAS. Only reason I thought its A2 B1 C4 D3 was because the PNP settings are that way, just that you need to fire A & C, B & D together before the trigger counter resets.

Actually, I think it's going to ultimately be determined by the mechanical constraints of the sensor. You can only mount it on the head one way, and with a certain amount of adjustment room. Seems to me that wherever the reset pulse winds up being relative to the engine at #1 TDC is going to dictate the decoder order.

Right, that's what I was thinking too.

Not sure about the year. I think any year would work.

Msns-extra doesn't offer up the ability to dial each cylinder's tune in does it?

Basically going this route would be to just take an extra firing outta the ignition system right? ie: removing the "waste" in wasted spark....

No and yes.

Basically, yeah, replace our CAS with Escort CAS since you need each cyl. to fire only once per revolution vs. our wasted spark of twice per revolution. Ignition pretty much stays the same, it only reads each revolution once now instead of twice, thats why we need two more trigger wires/circuits to fire ea. cyl. (cause now you have 4 individual trigger circuits/wires, one for ea. cyl.) once only per revolution.

Geez, I hope I make any sense here.

Wait, so ms1-extra code has the ability to tweak spark per cylinder once in sequential mode?

If so... that is usefull for tuning with! Do some egt logging per cylinder and even them out with spark!

No it does not have THAT capability. In other words, you cannot trim spark and fuel on a per-cylinder basis.

You misinterpreted my answer. You asked two questions: The answer "No and Yes" would be interpreted as:

what does straying away from wasted spark net you?

colder operating temps of the coils thus better efficiency

And, as to why this is needed, you discharge the coils half as often compared to the present wasted spark setup so:

1) Your coils would last twice as longer (theoretically speaking).

2) You would have a lot more spark energy available and therefore, all of the spark energy available due to capacitive discharge can be exploited. In essence this would mean you can open up that spark plug gap even more.

Hooray for not even being able to handle reading my own pseudo double negative question properly! :p

1) I kinda doubt that.
2) No, it's not gonna be a lot more. The coils charge and discharge. I don't have my books and formulas, but doubling charge time doesn't double spark. If the coil gets 99.5% charged before you spark, and then you give it twice as much time to charge, your working on getting that last 0.5%. Granted as RPM increase the time the coils have to charge decreases, but that doesn't mean spark power drops linearly with time.

But get er dun. I'm MS II so If I ever do COPs I wanna do them sequential, since I already have a crank wheel and plenty of outputs. :)

I'll buy the increased longevity concept. Fewer firings, less heat. Maybe.

Not sure about the "Spark Energy" concept, though. That is almost entirely a function of dwell time, and we're not running into a dwell limitation. For an engine turning 8,000 RPM, running wasted spark, there are 133 ignition events per coil per second, or 7.5 milliseconds between each ignition event. With the stock coils, we typically run ~5ms dwell, and with the COPs we run even less (somebody help me out here- 3ms? 3.5? Something like that)

Point is that once the primary winding is fully saturated, increasing the dwell doesn't get you anything except a damaged primary winding. And with wasted-spark, there is more than adequate time to reach full saturation.

Ok- to be purely academic, if the coil fires half as often, then the windings will run *slightly* cooler, the DC resistance on the primary will be *slightly* lower, and therefore the saturated charge current will be *slightly* higher. But you'd need a pretty accurate scope to notice.

One might make the argument that extra power could come from less voltage drop, since the dwell time is fixed off batt power not off voltage at the coils themselves.

This is true. The supply wire running to the coils is kinda puny, hence the "Cap mod" that folks are talking about now. Charging one coil vs. two means less drop across the V+ supply wire. Increasing the voltage available at the coil would raise the saturation point.

OEMs run sequential COP because it extends the lean misfire limit. Guess we dont really need to run that lean, except off-boost.

Anyway, I'm going to compare both and see if there's a difference or not. Sometimes things don't always follow theory.



Couple of quick questions for you guys:

1) If the 94+ CAS is magnetic (VR), then how can it give the same signal as the 90-93 optical (Hall) CAS?

2) The COPs that you guys run now, is it inductive coil or CDI?

First, Hall and Optical are not the same thing. From a pure physics standpoint, Hall has more in common with VR than it does with optical. Though the more I think about it, I believe (purely from looking at the rotor) that the late-NA sensor may in fact be VR rather than Hall as is commonly thought.

The signals are the same between the two CASs because they both contain an active circuit which provides a squarewave output via an open-collector driver (that's why they require an external +12 supply to operate). By contrast, most VR sensors that we see in Ford EDIS applications or DIY crank triggers do not contain an active driver- they emit a raw AC waveform and thus require an external circuit to derive a TTL-compatible squarewave from them.

Anyone done this? Do you have to fit the Festiva disc into the 1.6 CAS?

No, I believe The Pipefathers observation was that the Ford CAS is a direct fit, both electrically and mechanically, into the Miata. Thus, no need to dismantle anything, just remove the Miata 4-2 CAS and install the Ford 4-1 unit.

But no Festiva came with a plain CAS. They came with distributors with an optical sensor in place of points.

Two optical sensors, as far as I'm aware. Which, when you discard the cap and rotor, is pretty much equivalent to a Miata CAS, in a 4/1 configuration.

totally off-topic, but modernbeat, are you the same one who had the black B-mod Locost?

I know I'm digging this up from the past but does anyone know if there need to be any modifications done internally to the megasquirt to enable sequential?

for sequential spark, you need to free up some outputs and run some additional wires both internally as externally.
For sequential fuel you can install Jean Bélanger's board without having to give up outputs. Again, modifications inside and additional wiring on your '91.

Once you've dealt with the matter of having one cam pulse per cam revolution, all you need to do is enable SparkC and SparkD as outputs in the software, and the build a pair of spark output drivers to mirror what you've done on the primary (D14/D16) drivers.

36-1. :drool:

A 36-1 wheel, by itself, is not sufficient. You must have a cam pulse in order to run full sequential, and it must be exactly one pulse per cam revolution.

Already have a Festiva CAS on the way so that much is covered. I know next to nothing about designing circuits so bare with me here... So I basically need to make a second set of spark out put drivers in the proto area of the board. What areas of the board do I need to jump from and to, to make that work?

Also I remember when building the harness that your supposed to pair pin 32 and 33 together as well as 34 and 35 for the spark out. In theory couldn't you just keep them separate and run the triggers individually off each pin? Again I know nothing about circuit design.

SparkC is taken from D15 (top of R27) and is thus identical to your existing SparkA and SparkB. It can be build on the back of the board just like the others.

Spark D is a little weirder. You remove R1, and wire to the bottom pad of where it came from. Then build (in the proto area) and circuit which replicates the LED driver circuis (you can omit the LED itself) and connect R1 through a 1k resistor to the base of the new transistor.



http://www.msextra.com/manuals/MS_Ex...v3Multiple.gif



Those pins are for the fuel injectors, not the ignition. And they're paralleled just to increase their current-handling ability. For spark, just pick any four pins that aren't otherwise in use.

Thanks for the help guys, can't wait for this weekend to make this happen.

MS2-Extra Hardware Manual

Sequential Injection Code for MS2

MS2-Extra Ignition Hardware Manual

Frank's Westfield MX5 - Megasquirt

I've been looking to do the same stuff and these pages are all you need.