thegrapist

I'm finally pulling the trigger on a diypnp when I get paid next week. I'm doing a parallel install on my 99. Is it possible (or more specifically- not a total pain in the *****) to run a 12 tooth trigger wheel with a parallel setup without making the factory ecu **** its pants? Or am I stuck with timing float? I need the car to pass emissions here in jersey and they do a tailpipe sniffer and scan your obd2. So I can't have any cel's or non-functioning obd2 ports.

Thanks guys!

thirdgen

iTrader: (26)

In parallel, it will **** it's pants. Also all over anybody within 50 feet on the **** area.

thegrapist

Oh god. Everyone in south jersey will have messy pants. Has the internet tried any other hacks? two trigger wheels and sensors? Or any other clever nonsense?

tasty danish

iTrader: (9)

yeah, go standalone.

seriously, why parallel?

Joe Perez

iTrader: (8)

Because not everybody lives in Florida:

In many, if not most states which have emissions testing, '96 and later cars must pass a plugin OBD-II scan in addition to any visual or dyno testing that might be performed. Basically, they plug into the car's diagnostic port and interrogate it for both DTCs and readiness monitors. If the vehicle's computer does not respond back correctly, then even if the only thing coming out of your tailpipe is baby foxes, you will fail the emissions test.



thegrapist, have you already installed the 12 tooth wheel, or are you contemplating installing it in response to some perceived deficiency with the stock unit?

thegrapist

I have not installed it. But from the reports I've heard of the timing shift varying +/- 3 degrees, I'm just trying to minimize my chances of popping a motor. Do you think I'll be ok with the stock sensor? I plan on running a begi s3 with a GT2560R or FM's kit without electronics (but since I hate FM with a burning passion I'll probably get the S3 kit) and will be aiming for 250 whp.

As for Jersey's inspection, their obd2 testing is weird. They'll hook it up regardless, and you SHOULD fail if you have a readiness code (happened in the case of my civic) but I passed two turbo gti's and a jetta with cat efficiency readiness codes (literally cleared the code and crossed my fingers feet before the testing facility) and passed.

But any feedback or direction would be greatly appreciated

Joe Perez

iTrader: (8)

Well, there may very well be some timing deviation in the stock setup- it's a long way from the trigger point to actual TDC. A couple of thoughts occur to me, however.

1: It's fairly safe to assume that for any given combination of load and RPM, the amount of deviation (due to non-linear angular velocity of the crankshaft) would tend to be a constant, as compared to:

2: The stock NA setup, wherein the reference is taken not from the crankshaft, but from the far end of the intake cam, which gives you a TON of spark error due to belt flexation. I'm not talking about a constant deviation, but actual scatter. And,

3: Even with that setup, 250 HP isn't exactly running on the ragged edge.


In other words, if you were upgrading from a CAS-triggered setup, and you were shooting for 400 HP, then yeah, this would probably be a good idea. As it is, the 12-tooth wheel which is currently on the market is targeted more at folks with Adaptronic ECUs, as unlike the MS2 & MS3, those units can't read the stock NB trigger pattern.


Ok, that was... pointless. And deleted.

kenzo42

iTrader: (6)

So for a 200whp 1.6 on 91 octane, I assume wouldn't be crucial?

thegrapist

oh that sounds good. thanks for your help. I remember reading something on m.net that some guy popped his motor with a super conservative tune and only making 230 or so at the wheels, and was attributing it to thd timing drift. I'm just getting old and want to set everything up once and not have to deal with ghetto road side hacks just to limp home.

So then there's really no use for me to get an diypnp ms over a dynotronics reflash, is there? Aside from turboing my 99 the only thing I want to do electronically is install toyota COPS, but I can just make a circuit to reduce the dwell time. I won't have datalogging and won't be able to fine tune it, but once it's set up, I don't plan on ever touching it again except for maintenance stuff.

Braineack

iTrader: (62)

If you're going Parallel, ditch the DIYPNP and goes V3.0 with MS3 + MS3x

Joe Perez

iTrader: (8)

I personally would avoid any kind of "reflash" unless it was absolutely unavoidable, such as would be the case if you were racing in a limited class which restricted ECU modifications. A reflash isn't going to give your ECU the ability to read MAP, or extend the range of the MAF sensor, or give it the ability to read a wideband O2 sensor, or (...)


Here's the big question: Are you planning to go through inspection with the turbo in place, or do you intend to pull it all down to stock every other year?

If you want to try to get through with the turbo, then I highly recommend a parallel Megasquirt configuration using either an MS2 or MS3, rather than any of the PnP models.

If you're ok with swapping for the test, then I'd explore a DIYPnP.

MSPnP is not applicable to '99 and later cars at the moment, as it uses the MS1 CPU which, like the Adaptronic, cannot read the NB's crank and cam pattern.

C. Ludwig

There is nothing that precludes the stock cam sensor from making big power. Keep in mind this is the same sensor used on a long list of Mitsubishi 4G63 that have made up to twice the power of any Miata engine.

Does the MS not allow for variable trigger angle settings? Many ECUs allow you to program out any sort of consistent timing drift, which is common to a lot of sensors.

Braineack

iTrader: (62)

We aren't talking about the 4g63 here.

the 99 does not use this CAS.

C. Ludwig

My bad, missed the 99 part.

The question of variable trigger angle ECU setup is still valid.

Joe Perez

iTrader: (8)

Yes, the MS does have a correction for fixed latency. And obviously, if you are dyno tuning your spark table, then any error can be corrected for so long as that error is constant for a given cell.

There's nothing wrong with the NB crankwheel. The one sensor on these cars that I do have a serious problem with, ironically, is the 4G63-style cam sensor you mention, which was used on the '90-'97 Miatas. It's not that the sensor itself is a bad unit, but when it's driven by a cam that's driven by a belt, you are going to get spark scatter no matter what. On my '92, I observed as much as 10° of scatter at high RPM, and that was with a relatively new & tight belt. That's when I said "**** it" and built a 36-1 crankwheel setup.

Braineack

iTrader: (62)

I'm surprised how well my MS3 can keep the timing solid on my CAS. Much less scatter than with the OEM ECU. It's actually pretty solid.

thegrapist

There's so little documentation with the ms3's that I was just going to try ben's tried and true parallel with a diypnp, I didn't really think the MS3 was a better option for parallel installs.

While you're here I want to pick your brain: how many inputs does the ms3 board have? And would sequential injection be useful for me? The car is a DD for now (I'm buying an ms6 for snowy days and taking out skanks) I plan on running track days and autocross (maybe 8 a year tops) once I get the suspension sorted out. Nothing more than that. I will put the time into making sure it has a solid tune and will start under any temps, but I'd like reliability and not blowing up to be the top priorities for my build. So any tips would be awesome.

hells to the no. I'm leaving that turbo on until I pop the motor or get sick of the car. I'm lazy, but also realistic. I know for the first year or so I'll be hunting down bugs and fixing my hack work. But once I find the weak points and most of the problems; I don't want to do much else besides change fluids, spark plugs, and sensors. I think I'm going to go with an ms3 based on both your guys recommendations.

Thanks again for your help guys.

Braineack

iTrader: (62)

There's plenty of ms3 docs. It's not really any different than MS2, just better.

Inputs, more than I'll ever need. I can think of 6 generic inputs on the MS3x Board alone + 3 0-5v inputs for sensor logging + a few easy "triggers" (such a tableswitchnig or launch in). Not to mention all the extra inputs/outputs on the mainboard.

Yes, you want seq.

Plus with MS3 you can use the SD card logging when you are on track...

Joe Perez

iTrader: (8)

Parallel with a DIYPnP? Yeah, I guess you could make it work. Probably a lot more trouble than doing parallel with a standard (non-Pnp) MS. The PnP models are usually only used in non-parallel application, as their ability to plug into the unmodified factory harness is the whole point.


From a purely technical standpoint, the MS3 is better option for all installs. Not saying that the MS2 isn't entirely adequate, but the MS3 is definitely going to be the best-supported system going forward.



All of them.

Seriously, the I/O capacity on the MS3, especially if you have the MS3X expander, is insane. Even Abe wouldn't be able to complain about them.


Sequential injection, if configured properly, buys you slightly better emissions at idle and low load (especially with larger-than-stock injectors) which I gather is important in your application.

thegrapist

awesome thanks for all the info, guys. Looks like I'll be picking up an ms3 soon, not sleeping, and blowing up this forum with questions.