MS3X on Sbc 383. 36-1 trigger wheel with hall sensor on my own bracket. Have my MSD 6a running ignition module and Vortec Distributor as cam sensor. I thought it was all good and was tracking another problem when I stumbled on this. Finally took it out for a drive after re-sealing manifold for oil leak. The ignition started cutting in and out at about 2500 rpm. Timing was rock solid at idle and little revs. to 2000 rpm. I did a tooth logger and it just stops whenever the ignition cuts out. The hall sensor is less than .5 mm from the gear and doesn't have runout from what I can see. I'm thinking I have a setting wrong somewhere. it was even bucking a bit coming off of the problem and dash started to go nuts and jump around. so I limped back to the house.
as an odd coincidence or possibly affect of the problem my digital dash (1989 corvette) spazzes out and flickers all weird numbers and randomness on it. I thought maybe I was a tooth off of my distributor and there was a big gap being bridged causing huge rf interference. I tried moving tooth over but that just made things seem to be worse with the flickering. When I place the rotor pointed at cyl 1 it is actually a few degrees before the cyl 1 prong. To me that is bad and would explain timing dropping out as it advances to like 25 degrees. However the tooth logger shouldn't drop out should it?
I also suspect my pots might be set wrong. I set r56 to 2.5 volts just like the manual said. Maybe that is the wrong part of the manual? Can someone look at my settings and see if they see something blatant? the AFR starts leaning out super hard revving the engine too, not driving until I hit ~2800rpm then it leans out really fast. you can see in datalog it jumps all around.
This is the picture of the distributor orientation. This is not the current orientation, I have it set one tooth back so it is actually pointed at the corner of the fuel rail. Really this setting shown seemed to be worse for the dash flickering. The actual #1 prong is lined up where the little cast #8 symbol is.
Indeed, that rotor moving away from the cap tower when MS advances the spark is not a good thing. Back when that dizzy had centrifugal advance it wasn't an issue, as the "timing trigger" (points cam or VR reluctor) were integral with the rotor mount, they moved together and the rotor was always pointing right (ignoring vacuum advance, for now!).
Since you now have a 36-1 crank wheel and MS3X, why not ditch the dizzy and single coil, go to 4-coil wasted-spark or CNP? Then, you can eliminate any "who's pointing where" (cross-fire) or too-short dwell problems.
I'm not against moving to that, however the bill on all this keeps rising. I was told that the easiest way to convert to a sequential injection was to use the vortec distributor as the cam sensor. I feel like I should be able to get this right though. I'm more interested in how to fix this, because the problem might not be in the rotor alignment.... I mean I've already invested in the Vortec metal base distributor, and new wires cut custom for this setup... I mean I'm already 10k into this engine on a teachers salary, why not just dump another $400? I'd like to try to get this working first though.
Any ideas on why this is happening? See anything wrong with my setup settings?
$400 is kind of pessimistic. You could build the coil driver for $40 from parts, or pay $90 for a completed unit. I like the style coil similar to Standard Motor Products #DR39. You would need four, and new price ranges from $10-50 ea, depending on brand and source. However, that coil was used in a wide variety of GMs from '86-09, so should be cheap and easy-to-find in a junk-yard. $100-200 for the conversion, depending on how you go about it.
I browsed this thread again. Why do you think the ignition is cutting out, that it isn't a misfire due to some other reason? Why are you using no "vacuum advance"? The last log posted shows no sync-loss.
Edit: Also, at about 49.3 sec in the log, there is an odd-looking jump in several parameters. Notably, the MAT jumps 14F in 1/10 second, that can't be real. There could be a clue in there about a wiring problem.
In the tooth logger at time index 7372 all tooth logging drops to 0. Then it comes back in at few seconds later. There are huge variances temporarily all through out the log that to me seem to line up with me trying to rev the engine. Am I seeing things differently?. If the ecu wasn't registering teeth, that wouldn't mean a sync error necessarily correct? It is getting the rpm from the tooth logger right?
This engine and all components are brand spanking new. This engine is totally electronic ignition controlled. I didn't even know Vortecs came with vacuum advance ever. Vacuum advance is old tech.... Why would I want to use it? This is a road race engine. This thing needs to be rock solid reliable as I beat on it for years to come. Even my stock 1989 HEI didn't use vacuum advance. I have not ever had good luck junkyard surfing. Parts are very unreliable here in the south and I basically live in a swamp so 100% humidity doesn't work for electronics. and I don't want to chase more ghosts than I already have with questionable parts. Maybe for brackets and simple pieces, but not ignition parts, no way. I have no aversion to eventually moving to a different coil setup, but right now this setup should work. I want to get it working first before I do changing another thing.
I see the MAT Jump. That is interesting and I will check the wiring again. I've really taken so much care to make as close to perfect wiring harness as possible. I have chased so many electrical gremlins in my life I didn't want to do it with this engine so I took my time. The thing was running with my HEI unit and no cylinder identification and no problems other than poor fuel mixture. All I changed was the sequential parts.
I'm hoping I can get some more suggestions before I get to work on it Sunday or I'll just be chasing my tail after I check the wiring of the MAT sensor.
Yes, I saw the glitch in the tooth-log, but I don't think it means much. I think there have been many logs posted here showing such that were deemed to be "OK". Your log looked OK to me, other than that one page.
Nearly all engines use some form of what I loosely called "vacuum advance". That function is what sets the "load" axis of the VE and spark maps. I feel kind of at a loss as to how to explain this better, maybe some body else will join in. Yes, you may not need that for racing, but lack of it could explain why there is some misfiring when load is light like that; lowered compression pressure results in slower burning and you may not have enough advance at those low loads. Certainly that stock '89 engine used "vacuum advance", but it was built into the PCM's software, not a can hanging on the dizzy.
Thank you for brain storming with me. I hate troubleshooting and it gets me fired up cause i cant see the problem like a blown piston.
Yes I use a manifold air pressure sensor. So I see what you are saying now. I'm not using like alpha-n or anything. It adjusts to my manifold pressure. I thought you were talking about on a mechanical distributor. The 89 vette uses a maf instead of map.
The problem I had wasn't a misfire, it was a drop totally out of any firing. I know what a few misfires feel like or even bad timing. This was the same result as switching off the ecu and back on for a split second. The spike in lean mixture is what prompted me to think it was the ecu not getting triggers. If the ignition wasn't firing but injectors still fires, my Afr gauge would have gone super rich. It might have even been the injectors not firing but those connections are rock solid. I checked them first.
This problem is somewhat predictable. It doesn't happen at idle or low rpm. And it isn't just a bad mixture or timing is a bit off. It is either timing is jumping to a different cylinder or injectors are shutting off, or ecu isn't sending a signal .
If the ECU loses sync, the crank signal, both the spark and injector pulses cease, as well as fuel pump power. Regardless, a misfire usually results in a lean AFR reading no matter what the fuel mixture is. The WB O2 sensor is reading how much O2 is there after the supposed combustion cycle. With a misfire there is poor combustion so little O2 is used up and the reading is "lean".
I kind of doubt cross-firing in the dizzy would cause an abrupt engine cut sensation. The big unknown here, for me, is the MSD box. I have no clue what it's rev-limiting may be set to.
Msd box has a plug for rev limiter. It is a 7000 rpm plug. The cut out point isn't nearly perfect enough for that but the box could be suspect. I'm going to see if I can bench test it somehow but once again that seems unlikely.
I looked at your datalog.
1. You are not seeing any sync-loss. This is good - no need to worry about tooth logs etc. because it is working.
2. It looks like your computer is losing communication with the ECU because "SecL" has steps in it with 4 seconds or so of missing data.
3. I think your VE table is too lean where you are running lean and misfiring.
See at around 130 in the datalog. You get on the throttle, the RPMs climb and the MAP drops (this is normal.) You end up with the "dot" on the VE table around 1800-2000RPM and ~40kPa. In this area your VE table is between 40 and 45%. Bump that up to 50% as a start.
4. Do check that your distributor is phased correctly. This is covered in the distributor section of the Hardware Manual.
well I got a better log of when the problem was happening. It is definitely loosing sync and lined up perfectly with the very violent surging. I included my MSQ again becasue I changed the tooth where it gets phasing from to 29 (perfectly in middle of cam trigger). I also made sure distributor was phased perfectly. What is weird is when I first started the car after rechecking all the phasing, I revved the car all the way to 4000 rpm no problem. Everything looked good and I thought that was going to be it. Then I went for a drive and same thing. About halfway around the block it started doing the surge misfire.
The sync problem happened at all different rpms this time so it wasn't like last time where it happened in the same place. It was also more spread out and happened at random intervals, not just when getting in higher rpm. I think this is all coincidence.
I'm seeing a variance in the rpm before the sync loss, and it reads "cranking" when the sync is lost.
And here is quick video on the spastic dash. not sure if this has anything to do with the problem. It was doing this the whole time even when I unplugged the rpm input so I double this has anything to do with the actual problem. Think this may be EMI from the 6a unit.
Well, there is plenty of sync-loss in that log for us to sink our teeth into! There are a lot of questions to be asked about the CKP sensor, wheel, bracket, shielding, pull-up, pot adjustment, etc; but how about disconnecting the MSD ignition box and using the standard MS single-coil driver, see how that works? That should cost near-nothing to try. If your coil is non-standard, let us know what it is. Maybe the standard MS coil driver can't handle it...
I like that idea. That would show me if I'm getting EMI from the MSD Box also. Do I have to mod the board to use the coil driver or can it be done with temp wiring? I use a single AC Delco coil off a divorced HEI. Like on a 86 or so camaro. I hope you guys can see the difficulty I had with designing the wheel. The problem I had was how far the stock pulley goes off from the engine block. this is the only way I was able to make this work. I think there is a chance that some vibration may be getting to the hall sensor and when it is revved in neutral, the engine is smoother so no misfire, but when there is load then the engine prob has more vibration. These pics are when I was mocking it up on my old engine block. It is such a tight fit and had to make it easy to disassemble so I hope you can see my design. I can re-enforce structural integrity if needed but in my hand it feels super solid. As for shielding, the only thing shielded is the Hall return sensor wire which is the wiring harness I got from DIY. So the Ground and Power wire follow that. There is about 3 inches of wire that isn't shielded at the sensor with the pull up resistor. It is pretty far away from the coil and box through. Like as far as possible, opposite side of engine compartment.
You do have to install the components and make connections on the mainboard for the driver, but those won't interfere with using the MS3X spark outputs later, you can leave the standard single-coil driver in place. I built my V3.0 board that way. It normally is used on a wasted-spark V-8, but occasionally on a single-coil setup.
I think most of us put the crank trigger between the dampener and pulley. Yes, that may offset the pulley a bit, but alignment on all the driven pulleys isn't perfect anyway. Or, better yet, the (drive) pulley can be modded so the offset is near-zero.
That long bracket worries me, I can imagine it fluttering at certain resonant frequencies. But, I would think that would result in some "rubbing" against the wheel and would be pretty obvious. If that bracket/sensor is fluttering, it may be easier to see shining your timing light at it. Try using the light on enough cylinders so you see at least every 90 deg. of crank rotation. Actually, using various amount of "fixed" timing, you should be able to check just about every crank position.
It sounds like the shielding is OK, are you connecting the shield to an MS ground pin?
Is that a DIYautotune wheel and sensor? What voltage are you powering it with, what pull-up to what voltage?
Have you measured axial and radial run-out of that wheel? I think you said gap is 1/2mm, so I'm guessing the radial can't be much.
Modding my board again? Uhg. What about if I used a ignition module from my HEI 8 pin module to fire the coil? Like just rig it up attaching the wires to it instead of splicing wires as messing too much up?
My initial plan was to put it behind the pulley but it was throwing a lot of pulleys out of alignment. I would have had to space them all out and the ac compressor was the hardest. The solution was to get extended bolts and space all the brackets from the engine block mounting holes. The run-out on the wheel to sensor was less than .25 of mm. Very small. It was as good as I could get without precision machinery
I'm almost leaning towards the 6a box being source of dash spaz and vibration leading to sync loss. It makes a lot of sense. A redesign would be a pain but not the end of the world. iirc I put the 1k resistor maybe a 10k? as the pull up jumper for the hall sensor. It came with the package.
The reason I doubt it is that is that there was no problem revving the engine. Only while driving. .
It's not much of a mod. What mainboard do you have, the V3.0 or V3.57? If you built a V3.0, installing Q16 and R43 at that time, you are already most of the way there. I'm not sure, but the V3.57 might be even easier. Just a couple of jumpers? Yeah, the HEI module could work, but that would keep an additional unknown in there and some confusion about using the "GM bypass" or whatever it is.
If you are careful to avoid shorts, you could mount a BIP373 externally on a piece of aluminium strip as heatsink for testing and wire it to one of the MS3X outputs with a 220R resistor.
I'll call diy tomorrow and see how quickly I can get the part and look at the process of modding the board. Pretty sure I have the 3.57 board. I'm dealing with a bit of oil in cylinders also from what I think is a bad break in on the rings when I was having ignition problems a year back and flooded the engine several times on initial startup. So I might have to pull the engine over Xmas break (I'm a teacher) so I want to have a working system so I don't kill a new set of rings again.
Have you looked to see if Q16 is already installed? If it's there, you just need a small resistor, and those are often available locally and are very inexpensive.
