Winter Upgrade - Megasquirt

[Luce]

My understanding is the hall puts out enough of a signal to directly drive the optoisolator without the VR conditioner circuit. The VR conditioner greatly amplifies the signal, and the induced noise as well, to the point that MS thinks the noise is a trigger. But it's where MS knows there shouldn't be a signal, so it loses sync. That's why when the engine was sitting still, you were getting little blips of tach readings.

I don't know where the idea to run the hall through the VR circuits came from other than someone realized it would work (sometimes) and made it easier to market the MS as a one size fits all kit. I see no other reason for it. Since you opted for the assembled and tested ecu, I'm assuming you're not comfortable with changing the jumpers, but that would be my first attempt at a fix. If you dig into this, make sure the optoisolator, U3 is present as well as D1, D2, R12 R13, C12 and C30 are present. Also make sure you have all 3 jumpers in place.
To try this approach-
TACHSELECT needs to be disconnected from VRIN and connected to OPTOIN.
XG1 and XG2 need to be connected.
TSELECT needs to be disconnected from VROUT and connected to OPTOOUT.
Be sure you can locate all 3 of these sets of jumpers and that you do have a quality low wattage soldering iron and the skills to use it. This is a fairly densely populated multi layered board, and you can ruin the whole thing.

If you still have noise issues, disconnect XG1 and XG2 and use a vacant DB37 pin and run a wire from XG1 to that vacant connector pin in MS box, and run the wire to the ground on the hall sensor. If possible twist the hall signal and ground wire together. This eliminates most of the induced noise.

In my opinion (again, take it for what it's worth...the musings of just another internet troll) using the VR circuits for anything other than a VR pickup is like hooking the speaker output on a home stereo to the microphone input on a tape deck to copy your buddies albums to tape. Yes, if you turn the stereo volume down to the absolute minimum while recording, you may be able to make out what song you're hearing, but it will be far short of enjoyable music.

If someone can present an argument for the merits of using the VR circuit with a hall sensor, I would love to hear it.

[bobcowan]

Originally Posted by vector1

might be noise off your alternator, you have it hooked directly to the battery and all other precautions taken with the wiring?

Good thought. 2 weeks ago I removed the belt to stop the alternator from spinning, and ran off battery only. Didn't help.

Originally Posted by Luce

My understanding is the hall puts out enough of a signal to directly drive the optoisolator without the VR conditioner circuit. The VR conditioner greatly amplifies the signal, and the induced noise as well, to the point that MS thinks the noise is a trigger. But it's where MS knows there shouldn't be a signal, so it loses sync. That's why when the engine was sitting still, you were getting little blips of tach readings.

I don't know where the idea to run the hall through the VR circuits came from other than someone realized it would work (sometimes) and made it easier to market the MS as a one size fits all kit. I see no other reason for it. Since you opted for the assembled and tested ecu, I'm assuming you're not comfortable with changing the jumpers, but that would be my first attempt at a fix. If you dig into this, make sure the optoisolator, U3 is present as well as D1, D2, R12 R13, C12 and C30 are present. Also make sure you have all 3 jumpers in place.
To try this approach-
TACHSELECT needs to be disconnected from VRIN and connected to OPTOIN.
XG1 and XG2 need to be connected.
TSELECT needs to be disconnected from VROUT and connected to OPTOOUT.
Be sure you can locate all 3 of these sets of jumpers and that you do have a quality low wattage soldering iron and the skills to use it. This is a fairly densely populated multi layered board, and you can ruin the whole thing.

If you still have noise issues, disconnect XG1 and XG2 and use a vacant DB37 pin and run a wire from XG1 to that vacant connector pin in MS box, and run the wire to the ground on the hall sensor. If possible twist the hall signal and ground wire together. This eliminates most of the induced noise.

In my opinion (again, take it for what it's worth...the musings of just another internet troll) using the VR circuits for anything other than a VR pickup is like hooking the speaker output on a home stereo to the microphone input on a tape deck to copy your buddies albums to tape. Yes, if you turn the stereo volume down to the absolute minimum while recording, you may be able to make out what song you're hearing, but it will be far short of enjoyable music.

If someone can present an argument for the merits of using the VR circuit with a hall sensor, I would love to hear it.

More than one person recommended the VR circuit and the hall sensor. There are a lot of ECU's out there set up that way, and they seem to be working fine. I only know enough to be dangerous, so I take their word for it.

Speaking of dangerous, me getting anywhere near that board with a hot iron would be dangerous. I can move pins around, solder wires together, etc. But my skills are not near good enough to come near this SMT board. That's why I had them build it.

The hall sensor should not need a pull up resister. But, out of desperation, I put one in anyway. More than one person recommended it, including an exclusive builder. It seemed to make it better. We'll see this evening.

[Luce]

Another option would be try using a dizzy with a VR sensor. A stock one should work, and let MS run the coil from the single coil driver that should be installed. Don't forget to set the options in the setup menu so MS isn't looking for the missing tooth.

You need something to drive the oil pump anyways, right? Unless you're dry sumped.

I'm running an MDS distributor that has a VR triggering the MS. I use the coil driver to drive a 6AL box, and I have a bypass plug so I can make the coil driver directly drive the coil if the 6AL crapps out. So I have the advance curve of the dizzy, but I can add or subtract a few degrees with the MS. But my setup can not run individual coils because MS has no idea what plug to fire.

Another set up I've read people successfully use it to run a full factory edis system, and let MS control the advance through the SAW and PIP signals. I think this is as good as the MS 4 coil setup I think you were after performance wise.