MWMIata

1999 NB, MSII, AdapterCard, 10psi.


Found a strange issue when I was at the dyno tuning the car.

The rev limiter comes on at 6500 RPM no matter what I set the Hard / Soft RL at( I had mine at 7000 / 7250). Anyone had this issue?

Also, somewhat unrelated. What resistor / pot is everyone using the limit the sensitivity of the knocksense. I tuned to where I knew there was no knock but it still gives me false signals.

Thank You

Reverant

iTrader: (10)

I had this very same issue happen on a friend's MS2/V3, only it was happening at 5200rpm, it was losing sync. We could never figure it out, so we upgraded to a DIYPNP.

Dimitris

Ben

iTrader: (33)

Did you take a datalog of this? If so, post it up with a copy of your current tune and we'll have a look.

MWMIata

Datalog and msq attached.

Settings seem to have no effect on the limiter. I gotta be missing something!

Thanks for Looking

Wanna hit the track tomorrow after work, need to get the rev limiter working correctly!


PGillum

Ben

iTrader: (33)

You're not hitting the rev limit; at 6200 rpm you're losing sync. Does the car rev up properly with the stock ECU? How are you handling your cam and crank inputs?

MWMIata

Hey Ben,

I have Abe's Adapteroard 99-00 NB.

Yeah I reviewed and you are correct error 31 / 32 cam sync loss.

Car rev's fine on stock ECU last I checked.

How do I fix??

Marc D

iTrader: (7)

Did you do the cap mod? The caps that Abe originally sent were the wrong spec, it would cause severe sync loss.

AbeFM

iTrader: (3)

I had an issue where I would lose sync at 5,200 - and it took a long time to track down - turned out to be the lack of a protection diode on VICS - that's a big relay, and when it switches, it dumped noise/current into the MS and I would get all hell breaking loose.

My fix at the time was:


But my plug-and-play-ish boards already have diodes everywhere, so I don't think this is his problem.


Ah, yes, make sure you don't have coolant dependant rev limit on. I've had issues there. In fact, far and away the wierdest - I had it limiting ONLY at high throttle. Specifically, what will happen is you have a TPS-bypass. That will give you a different rev limit than you have everywhere else. That gave me SO many problems. There's no software to check for this, you just have to know that ALL rev limits are set to the same value, or understand what each and every one does and how they interact.

I've also had issues where normal engine cuts (overboost, rev limit, etc) trigger the knock sensor and since these aren't masked in that situation you will end up with massive spark retard for the first couple seconds of the next gear. Both are relatively easy to fix in software. Perhaps now that MS-IIe software is pretty much static it'd be worth going in and making all these fixes.



I pulled the back case off, and you can see the screw through it. I don't remember on the MS version if you need to open the case, I think you might. Try it - it's obvious, big, blue, and you tune it with a screw driver. It's very easy, turn it all the way to one side, and the light should come on, turn it the other and it'll go off. Then free rev the motor and turn it towards to "on" side (CW I think) as much as you can where it JUST won't flash before you hit redline. If you're really clever, you set it for a few hundred before redline, then tell the software to ignore it that high - and you can have it a bit more sensitive in the midrange.

Personally, though, I don't do that and I've been running 15 psi for over a year on stock internals. The motor's getting a bit tired, but it's not bad.


I have a feeling you're right here. I don't think there's anything in the board which really points to this, I believe I had this card on the sim and it rev'ed fine. And I believe his sensor gaps have been looked at.

So, I'm a bit stumped. I guess you could try to lessen the filtering of the input signal a bit. What caps do you have in there? Either less cap or changing the resistors should tweak the noise filtering, but I haven't seen anyone else have these issues (not to say they aren't real but it's worth being careful and seeing what else might be causing it.)

AbeFM

iTrader: (3)

I'm suddenly very keen on this, a friend's vovlo is ALSO falling flat on it's face inexplicably at 6,200 rpm. I've yet to see his logs. I seem to remember him saying if he's not in boost it doesn't happen, but since he's got a utterly different triggering going on, it's worth looking at. I wonder if it's just a firmware bug or something - though I would imagine more folks would have seen it.

Doing a compare to my own MSQ, I don't see a lot of differences. I see, compared to mine:
boost control is on for you
Skip pulses is 3 (vs 1) which might make a longer pause after a lost sync? Not an issue.
You have idle advance on, really doubt this is it either.
Knock Control is off, duh.
You have no spark latency in hardware, I'd recommend about 6 usec, then you might need a slight tweak on the dyno (on order of 3-4 degrees)
Your crank to run taper is SIX seconds? Wow. Not the problem, but interesting. Mine's 1. 1.5 would be ideal.


Shoot, only the first 12 differences are reported. I don't know what the other 11 are. I don't know what 'use cam pulse if available' is, but I have it to, so I guess that's not it. :-)

Oh, and the overrun fuel cut, I much prefer to see it down in the 1% or less range, mine's 0.9%, as my car makes significant power at 2% throttle, and I hate having it NOT go because of it.

Still looking like hardware, which I'm not happy about either. What caps DO you have on there?

Reverant

iTrader: (10)

My friend has a 2004, so no VICS (or VTCS, as it is a European spec Miata). I'm pretty sure that something is wrong with the circuit I made, where I didn't use the LM339 that I normally use, but some other IC.

Dimitris

AbeFM

iTrader: (3)

It's a more general thing, the solenoid, could happen on any of them - but I'll admit, the 5200 number rings a bell with me because that's where VICS switches. The basic point is beware of noise.


So I'm looking at a log of my friend's, the volvo, and he's getting his cut out in a similar spot (6270 and 6374 are the two peaks before the fall)... the only thing I'm seeing is he ALSO has boost control he's not using on. I can't see that causing it, but I'm pretty stumped. He thinks it's a bad bracket on his position sensor, but Paul's is stock.

I assume it's well gapped, paul? That, and for the third time: What capacitors do you have on your input circuit? If they are too big, you WILL hit an RPM wall.

Ben

iTrader: (33)

This is along the lines of what I was thinking as well.

MWMIata

Thanks Guys,

Sorry for the delay. In position C1,C2 I have what appears as PN HC 103

As for the KS The big blue pot, I have it on the least sensetive setting and it is seeing false knock. I assume I could decrease the sensetivity adding another pot but then is there a point to having the sensor?
Either way the RPM is the problem now.

Turned off the boost control still getting the lost sync at 6300 rpm.

AbeFM

iTrader: (3)

Heh, assuming I remember this right, 103 = 10x10^3 = 10,000pf = 10 nf. Yeah, you want 1 nf.

The quick test, clip the caps. Just cut one leg and bend it away enough it doesn't touch. That should fix you up. To get the best noise rejection, it's good to have some in there, so get some 102 caps and replace the cut ones with it.

MWMIata

Works with the caps cut.

Thanks Guys!

Ben

iTrader: (33)

Yeah, that makes sense. You may find that you need to add some filter back to the circuit, but see what happens. MS2 does have a pretty good software noise filter.

AbeFM

iTrader: (3)

The circuit I use for input is naturally very good at filtering, even without caps. But I don't like the idea, and I recommend putting them in. As I understand it, he is already in the process of getting them. It all stems from a very old typo on my part. :-)