When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
There's basically four possibilities are this point:
1) Your second cam sensor is bad
2) Your crank sensor is bad
3) Your wiring is bad
4) Your ECU is not properly filtering the signals.
The first seems unlikely, although it's conceivable that if the nubs on the cam gear were damaged then you could get intermittent signals from two good sensors. That seems unlikely though. The second is possible, although miata crank sensors seem to be more reliable than cam sensors. I assume you checked the gap and it's correct and the sensor isn't loose or anything, right?
Do you have the MS3 plugged directly into the factory harness, or do you have an extension harness? Theoretically an extension harness should use shielded wire for the cam & crank signals, although in practice many people get away without it. I think the most likely cause is bad filters inside the ECU, that was the cause of the problem when I was seeing it. What happens if you turn software filtering off entirely? Will it even start?
Losing spark can't cause a sync loss unless you lose it for so long that the engine stalls. If you're in gear and rolling this basically doesn't happen because the car's momentum keeps it going fast enough (otherwise you'd lose sync every time you lifted off the gas on the straight).
"sync" means that the computer is seeing the pulses it expects, in the order it expects. It's not dependent on anything the engine does except that it's turning and the timing belt is working correctly.
There's basically four possibilities are this point:
1) Your second cam sensor is bad
2) Your crank sensor is bad
3) Your wiring is bad
4) Your ECU is not properly filtering the signals.
The first seems unlikely, although it's conceivable that if the nubs on the cam gear were damaged then you could get intermittent signals from two good sensors. That seems unlikely though. The second is possible, although miata crank sensors seem to be more reliable than cam sensors. I assume you checked the gap and it's correct and the sensor isn't loose or anything, right?
Do you have the MS3 plugged directly into the factory harness, or do you have an extension harness? Theoretically an extension harness should use shielded wire for the cam & crank signals, although in practice many people get away without it. I think the most likely cause is bad filters inside the ECU, that was the cause of the problem when I was seeing it. What happens if you turn software filtering off entirely? Will it even start?
Losing spark can't cause a sync loss unless you lose it for so long that the engine stalls. If you're in gear and rolling this basically doesn't happen because the car's momentum keeps it going fast enough (otherwise you'd lose sync every time you lifted off the gas on the straight).
"sync" means that the computer is seeing the pulses it expects, in the order it expects. It's not dependent on anything the engine does except that it's turning and the timing belt is working correctly.
--Ian
MS3 is PnP, directly into harness.
I will try no filtering in the morning.
As I did R&R the mixing manifold, there is some possibility I damaged something on the Crank Sensor or it's wiring.
I checked the ground on both connectors and they are intact. Meaning "0" ohms between the ground pins on the harness to vs the Intake Manifold.
I have ordered a Crank Sensor.
Wiring is bad... that is why, after I get the Crank Sensor, if I still have problems, I will remove all wires from the DB37 except those being actually used: Power to CAN module and O2 Sensor Controller. Pin 2 ground to the Controller, and Pin 3 ground to the CAN; and the CAN module to MS3.
The only other thing I can think of is that the rear O2 sensor lead I just cut and did not put shrink on the ends. This was done 8K miles ago, however.
If we get to the internal hardware filters, would my source of help be you, Reverant, or both?
Wiring is bad... that is why, after I get the Crank Sensor, if I still have problems, I will remove all wires from the DB37 except those being actually used: Power to CAN module and O2 Sensor Controller. Pin 2 ground to the Controller, and Pin 3 ground to the CAN; and the CAN module to MS3.
The only other thing I can think of is that the rear O2 sensor lead I just cut and did not put shrink on the ends. This was done 8K miles ago, however.
If we get to the internal hardware filters, would my source of help be you, Reverant, or both?
If it's a wiring problem, it would be in the wires between the megasquirt and the sensor. With a PnP ECU, those are all factory wires, it's not on the db37 unless you've done something unusual in hooking it up.
The O2 sensor wires shouldn't impact it unless you're concerned that you might have nicked the cam/crank sensor wires while doing it.
For hardware filters you should talk to Rev. If you're curious about what I did, though, it's here:
Also, would it help if I tied the MS end of the shield to the ground rather than the sensor end? (I have not yet verified that the wires are shielded, I'm taking Ian's word for this). This would not be easy to accomplish, but just checking.
Have you talked to Rev yet if the sensitivity of your MS crank and cam inputs can be adjusted?
See discussion below. Reverant's response makes me think his ECUs are equipped with self-adjusting VR conditioner circuits that don't need trim pots.
Originally Posted by Oscar
Does everyone with an ms3 need to deal with this pot adjusting ------ry? I'm not looking forward to crack my case open every time something goes fucky. If my rev-built ms3xxx ever gets here of course.
Originally Posted by Matt Cramer
Not always - and if you start with an MS3-Pro, it uses a self adjusting circuit that doesn't need trim pots.
Originally Posted by Reverant
Neither the Enhanced MS2 nor the Basic MS3 need ANY kind of adjustment.
There are definitely opto-isolators in mine. Those are good at isolating the electrical noise from being able to damage the input circuitry, but according to a friend of mine (I'm a software guy, my relationship with electronics is still a bit fuzzy) those are actually somewhat less desirable than a simple RC filter from a signal integrity standpoint because they have a tendency to amplify noise. I dunno.
The RC filters solved my random glitch problem. Yes, I've been having some other sync errors recently, but the pattern is very different, they're correlated with each other and with engine response, rather than occuring randomly once-every-15-to-20-minuets. After the investigation I did into it over the weekend I'm pretty confident it's a failed cam sensor.
So did you replace the optos with the RC? As in, did you remove the optos?
We make embedded computers for our products so I also have access to help if I need it.
I did not physically remove them from the board, but I disconnected them.
In my MSlabs box, the cam & crank signals come in on the Mazda connector on the big board, go across the big board to the medium-size daughterboard (I think this is the board normally used for an MS2?), up through stack of connectors to that daughterboard, get filtered, then come back down that stack, cross the big board again, and up a different stack of connectors onto the small daughterboard (the one with the MS3 CPU on it). From there they go to the input pins on the HC11 micro controller.
To disconnect them, I pulled one of the connectors out of the middle of that second stack and used a soldering iron to melt the plastic so I could pull out the two pins that carried those two signals, and I soldered wires in their place. Next I went to my ECU extension harness (I mount the MS3 in the glove box, so I have a 4 foot extension harness to plug it into the factory connectors) and I moved the two wires carrying the cam & crank signals from the normal location on the ECU side of the harness to two different pins, pins that the MS3 was not currently using for anything. I then attached one side of my filter circuit to those previously-unused pins and the other side to the wires coming from the connector stack. There are photos in the build thread i linked above (also here: https://www.miataturbo.net/build-thr...2/#post1206290)
Again, I don't know if this is the right solution for your problem or not. Reverant's opto-isolators seem to work for most people who use these ECUs, I don't know why they didn't work on mine. Perhaps there's been a revision to the way the board works since I got mine, or perhaps I just got unlucky and have an opto-isolator on my board that's out of spec. Replacing the filters this way is a non-destructive mod (the only thing it damages is one of the connectors in the stack, they cost less than a dollar each and are trivial to replace, no solder required), so it's something we decided to try one Saturday afternoon. It worked, so I stuck with it. The difference was night and day -- before the mod, if I turned off the software filters in the MS3 then the car would barely even start. With the mod, it started and ran just fine with them disabled.