Megasquirt tach signal is erratic

FINALLY got the duster out for a good long test run. 28 miles or so.
Ms2 3.57 running batch fire fuel only on a 1970 Plymouth duster. Ignition is an msd billet distributor and streetfire box feeding the tach signal to the megasquirt. 

Issues: the tach signal is erratic.  But its odd. Only does it at light throttle at operating temperature.  Stand on the throttle and it goes away.  The erraticness does not seem to coincide with anything like bumps or electrical components cycling.

Any ideas?

Where are you picking up the signal?  Off the coil or off the tach line from the MS?

Tach input signal to megasquirt is wired directly to tach output from the msd box. Box triggered by the msd distributor. 

Is this the car's actual tachometer, or the software tach while tuning?

I think the doctor's question (and I agree) is about what the tach is getting its signal from; the Megasquirt, or the MSD box?

Is your setup Distributor->MSD Box->Megasquirt->tach?

And it's not running oddly, right? It's just the tachometer that's misbehaving? That would mean that the Megasquirt is probably getting all the right info, but that info isn't reaching the tach.

Just for clarification; is the tachometer acting up or are you seeing an erratic signal into the MS?  If it's the signal into the MS how are you determining that it's erratic?

Tunerstudio shows it as erratic, spiling to 11k and flashing red. Car breaks up and runs like ass at that point.

I had thought it may be being caused by the tach as that was the only recent electrical change, so i disconnected it. So, inly way to know the signal is erratic is in tunerstudio and by ear.

Okay, this is archaic, but since it sort of hits your specific "noisy tach signal to MS with MSD box", I found this link.

So the simplest question is whether the Megasquirt and MSD box are grounded at the same location. The other salient bit got into adjusting capacitor values...

The other thing Teh Googles popped up pretty quickly is the MSD box's tach signal being wonky if it doesn't have a really solid power supply. And that one specifically follows your part-throttle issue.

Hmmm....

Me tions grounding msd to the same place megasquirt is grounded.

My ground for msd goes straight to the battery. Megasquirt is passengers side cylender head. 

However, i have a 4 gauge from drivers side head to the battery in the trunk.

I think step one should be check the battery terminals. 

However, pmease keep this coming!

I feel lime i am awfully close to a sorted Install after a few years of struggling. 

You do have to be careful with MSD and tach stuff.  If it is firing multiple times, you know, like Multiple Spark Discharge, that will confuse a lot of things.  Is the signal coming from the MSD thingie one spike per ignition event?  MSD makes a "tach adapter" for the 6 series to fix just that problem, but I don't know exactly what you're using.

In reply to Dusterbd13 :

I had a similar problem with my Ms on an lt1. The instructions say to ground Ms to engine block and this is a big mistake. The reason is you ignition which is thousands of vokts grounds to the cylinder head. The signal going to the ECU are either 0-5 volts or 0-12 volts. The high voltage being grounded to the same place causes the ground reference for your ECU to change.

Take the grounds off the cylinder head and run a test wire directly to the ground terminal on battery. See if that changes the behavior. Look closely at some of you other sensor inputs and I think you will find tgey all get a bit cleaner.

That change fixed a bunch of small gremlins for me.

In reply to Fladiver64 :

Ill definately try that. Got nothing to lose but a few minutes.

Keep suggesting stuff y'all! Im excited to button this thing up and actually drive it around. 

throw the megasquirt in the trash - thats my idea. 

In reply to fidelity101 :

You have no idea how close I've come to that.....

No direct experience, but the grounding idea sounds promising. 

I do know that when I was messing with the MS on my build a while back, the wired going to the ignition module were shielded. Pretty sure the tach did not come off of those wires, but I could see interference with something not being shielded that needs to be causing the erratic behavior. So it's probably worth double checking if any of your wiring needs to be shielded. 

Did you read the MSD power supply link? There's more there than "voltmeter reads good", and the symptoms and conditions mirror your own really closely.

Dusterbd13 said:

In reply to fidelity101 :

You have no idea how close I've come to that.....

you have no idea how good it felt to smash mine with a hammer so nobody would ever have to face that garbage again.

In reply to fidelity101 :

Not sure what problems you had.  I've done 3 and have another planned.

So, current plan is to try moving all grounds to the same bolt on the head based upon the threads. 

But....

There us a 4 gauge cable from engine to battery negative. The msd is on battery negative. Would it potentially cause issues to leave the msd at battery, put all megasquirt grounds and 4 gauge under the same bolt, and have them share a ground path that way? 

Also, per the voltage supply side of the msd: heavy red straight to the battery, switched is through a relay. Near as i can tell, per gauge on dash, system voltage stays constant regardless of throttle position or engine load. Is there a way to watch what voltage megasquirt is seeing so i can compare? 

In reply to Dusterbd13 :

It's the MSD voltage that would be giving issues according to that thread. I think there should be a voltage gauge in the software somewhere (though I'm way out of touch), but I'm not sure it'll tell you anything meaningful about the MSD's power supply. If you have another relay handy, maybe try swapping that out, assuming you've already rechecked/cleaned the connections?

If you can put a voltmeter on that switched wire on the MSD side of the relay, that might tell you something.

In reply to Ransom :

Ooh. I have a spare voltage gauge!

Hopefully we can figure out if the software has a voltage gauge as well. Maybe be able to confirm or deny an issue there....

Ideally all the grounds should be run to one location as close to the battery as possible.  When I wire a car (or more often a dyno cell) from scratch I install a ground junction block right near the battery and run everything to that.

Will it act up without a load?  It would be interesting to see if the timing jumps around when it's acting up.  You could be having an issue with the signal to the MSD box  or the box itself could be bad.

I cant seem to get it to act up without a load. Yet.

And this is the first wiring job ive ever done withb issues. Megasquirt is proving very, very finicky in this project. Not even the early fast ezefi systems had this much trouble with tach signals and interference. 

Is your tach signal shielded and the shielding grounded on one end?

Can you send me a tooth log of the erratic tach signal to support@diyautotune.com with a copy of the tune file? This will give me a closer look at the signal.

In reply to Dr. Hess :

Most is. Theres about 18 inches of wire from the msd box that is not.