MS2extra manual says the signal voltage from the KnockSenseMS is too high for the processor port directly and they suggest using an input circuit with a zener and resistor. Is this not necessary then?

 

Hi Ben,

you wrote in one of the posts, that you have extra ground running to the DB15 - how did you wire that inside the DIYPNP - or how would you do?

Greets

 

That circuit is for the GM knock sensor module; the KnockSenseMS is a different device that puts out the voltage the MS needs already.

 

The mainboard has redundant access to the power ground plane. Not much to it.

 

Sorry Ben, my English is leaving me here. Could you please explain a little further?

 

He means that there's enough ground points on the board that you can use. Just use the one you like best.

 

Ha yes, of course, silly me .

 

Ben, how do you like yours so far?

Jim

 

It worked great. I pulled it 2 weeks ago, to mod it for direct coil, sequential fuel, and a couple other things such that it would play nice in parallel with the stock computer.
Have not had time to complete due to house repairs caused by the big floods we had several weeks back.

 

Why in parallel?

 

To pass OBDII plug in emissions.

 

Doh - at least its cool that the stock ECU doesn't fall flat on its face because you share the CKP/CMP sensors.

Jim

 

And if it did, one could always add a NA CAS.

I have a 60 tooth trigger wheel on a custom insert to fit on the crank pulley. I'm tempted to just put it on and be done.

 

That was my solution when I had the MS1 in parallel - NA CAS from a 96.

Jim

 

Mine is also assembled and flashed - the mainboard. Sooo coool.

 

Zaphod, du kannst dein Setup mal im Turboforum vorstellen wenn du möchstet.
Wäre cool

Just a littel note to Zapod about his building.

 

Gonna do that...

 

Been reading up, gathering info for my next build .
I really wonder why the µs inputs work fine and the ms inputs don't.
The opto-circuits are pretty much identical (apart from a schottky diode instead of 2 normal diodes).
Opto µs
opto ms

The ms VR circuit is very similar to the µs circuit and can be easily modified
VR µs
VR ms

I'm tempted to convert the ms opto and VR circuits to those of the µs in my next ms build.

OTOH, I can build Jason's low battery circuit in the opto area and both of Abe's input circuits in the VR area.
Should I convert to the µs circuits, I don't have room for the low battery circuit (proto area is completely full already with other ciruicts).

Bah, choices.

 

Hey, hadn't seen this thread before! Congrats on getting it working. Looks like there's a lot of wiring, but it makes sense, it IS a general purpose board, and for that it works really well. Yes mine has (so far?) the odds and ends all taken care of (alternator, AC, VICS, etc etc), but admittedly this is less I/O limited, by using the microsquirt it's more approachable and flexible than any MS-II based board would be.

I forget exactly what I did, but the AC switch is something pathetically easy, like a single diode and some wire. It's in the AC thread, at any rate, but you don't want to go wasting MS-I/O unless you're also going to try to use it for anti-stall.

How's the idle??


As I got a better tune, I would have overboost issues. Adding extra timing in helped keep it under control.

Noise on the inputs, or noise on the general lines?

That's REALLY interesting the uS is ok with the inputs that way. It's the same exact processor as the MS-II, or some variation? It doesn't make a lot of sense to me at first glance, unless it's just getting away from the.... 'less adequate' MS3.0pcb

Someone may have answered this, but you should be fine. the KnocksenseMS is called that because its tuned and set up for the MS.

*IF* you tell Boris you have a 99 miata, he'll set it up special to use the OEM sensor. I sent him one long ago (lent me from someone, maybe you or olderguy?), and he developed a tune for it. Then I take the wire from the OEM harness, run it out to the knocksense, then back in and into the CPU


OH GOD NO. Do not run a cas. You might as well run MS-I. You get horrible timing, like, 5 degrees of error.

A 60 tooth is a fine idea, though.


Shoot, there was one more point, but I forgot it. C'est la vie.

Subscribed.

 

I agree, I didn't want to waste the I/O either. Matt Cramer says to run it through one of the inputs.
Considering that because we didn't scope the idle valve frequency before yanking the computer out so we unplugged the valve and turned the bleed screw to idle at 1k rpm, I'd say it idled pretty damn good.

I was actually running what looked like a lot of timing on paper. I didn't want to push it until putting hte car on a dyno.

Occasional spike of all input parameters. Saw the same thing on my MSI, and was cured with additional grounding.

Don't know why myself, but I'm told that it's common for the uS to be happy with noiser triggers than the MS2 is.


That's freaking cool. I already have a knocksense, FM mount, and GM sensor though. Oh well. The FM mount is probably in a better location anyway. Maybe.
Yeah I'd run the 60 tooth, because I have it. It came off Frank's car (had a MoTec I think?) when he put in the V8. Most people would find it easier to put in a CAS. Which makes me think... I based my new timing table off my 91's dyno tuned tabe vs 1.6 stock compared against a stock 99 table. Makes me wonder about error in the old style CAS.


Answers are a little ambiguous because once I got the car running, it only stayed that way for a few days. I've pulled the DIYPNP out of the car and reconfigured it for parallel operation. I've almost finished the wiring harness. A trip to Frys for some more wire and heatshrink will get it finished. Can't believe I used a 100' spool of wire!!

Sure hope I can get CEL-free parallel operation.