Learn me Tracking down electrical noise

So, wes and i are both convinced that the last major issue with my megasquirt setup in the duster is noise. We were getting VOLTS of noise according to him. Im still not entirely sure what he saw, to be honest. Regardless, the rpm readout in tunerstudio jumps to insanity every now and then with a corresponding breakup in the engine. 

I want to go with the nuke it from orbit approach to finding every single source of noise and eliminating it. Every single one. Permanently. 

Problem is that i have no idea how to find it. After a rather exhaustive google binge, i have become more confused than when i started. 

Id like to be spoon fed in small, easy to understand words. I have a feeling that some of my knowledge is wrong on testing and measurement procedures, and is leading to my frustration and burnout. As well as my inability to actually fix this berkeleying thing.

So, step by step, from basic to Einstein, learn me testing and isolating sources of noise please.

Noise is like radio. You have one powered wire with a changing current, and that wire can then transmit and electric field that causes smaller currents in other wires.

I doubt noise is your problem though, maybe something with the crank sensor or wiring? How do you know it is noise?

How much wiring does this setup require? I'm not an electrical expert (so feel free to stop reading) but doesn't electrical degradation happen on a fairly slow slope? It's not like the sensors are super accurate, so any noise seems trivial.  Bad grounds or a poor wire connection...much more likely. 

Also, as I understand it, noise is like water in pipes.  Given a large enough system, say a city- you're going to lose some water.  That causes pressures to diminish near the end of the lines, sort of like noise at the end of a system.  This is one of the reasons that we now have fiber optic cables running all over the country.  Electrical noise is turbulence, it happens over distance.  I just though it took a fairly long distance if the wiring connections were kosher. 

High current and high voltage wires/cables should be ran away from sensor wiring.

So starter/alternator cables and ignition wires should be away from other wiring. Twisting wires also helps

How do you have it installed?  Is it plugged straight into the stock ECU connector, or is there an extension/adapter harness?  What kind of car is it?  Did you assemble the MS yourself, or is it a commercial one?  What version MS?

I spend a lot of time chasing noise in coax systems. While different hopefully some steps will help. First question is the noise repeatable. Is it something that happens every few minutes or is it like a once a week thing?  When it happens how long? You need repeatable or a long duration to effectively trouble shoot. If you have those then I would start by disconnecting stuff. Can you run off battery only to eliminate the alternator?  Try pulling fuses to eliminate other components. For example a friend had a problem with his engine shutting off. It would immediately start back up. I don’t know how he found it but pulling one fuse prevented it from happening. Then it’s a matter of looking at one small system instead of the whole car. 

I saw similar failure mode in my FiTech system with random impossibly high RPM spikes causing the system to stumble then catch itself.  I traced the problem to the tach signal wire from the MSD box to the throttle body.  Extended it to move it away from spark plug wires then wrapped it with aluminum tape for shielding and problem went away for me.  Frustrating as hell and glad it's behind me now.

In reply to wawazat :

That does bring up a good point.

Aluminum foil can be used to trouble shoot and isolate by wrapping it around wires. Try using it with the foil connected to ground at the sensor end or the ECU end (only one end of the foil), or floating (no ground connection to the foil).  Don’t connect both ends to ground or you can create new problems with ground loops.

I have had to do this and I always start with motors. Anything that has a motor I will disconnect (if you can) and see if it cleans things up. The order is up to you. One at a time or all at once. I like All at once because if the noise goes away then you know you are on the right track.  Then it is just a matter of hooking up each one until you find the culprit. 

Real shot in the dark here, and not what you asked for, but its all I got:

I have heard of MS, etc doing similar things because dedicated powers/grounds were needed. I.e. don't ground them all to the same bolt. Maybe run one power one side of the engine bay, the other opposite. Stay very clear of spark. Clearly, I am no expert.

So, i guess clarification is in order.

70 Plymouth duster. Megasquirt2 3.57, msd box, custom made wiring harness bumper to bumper. Using factory distributor, batch fire mpfi.

Weve been chasing issues related to ignition and tach signal. Every change has made things better, but we are still fighting the noise. Much improved from removing the hei8, but still present.

Tach signal is a sheilded pair to distributor. Output to msd box along the points trigger wire. Msd grounded to the battery. Coil leads are a twisted pair. 

Theres two independant harnesses running side by side. Only crossover of the harnesses is at the 12v distribution block and the relay grounds. Other than that, and the basic power/ground distribution system (batter to bussbar, chassis is a ground, and engine has dedicated 2 gauge from head to battery for ground) theres no crossover.

Im searching for each and every source of noise. 

Wes wants me to start with looking for voltage drop. Near as i can tell from google, i hook ine lead if my multimeter to the battery positive terminal and use the ither to probe ground points and compare to battery voltage. Use 12v setting on multimeter. 

Then doe the same for the pisitive sudes of the circuits with 12v

After that, check ohms of resistance between negative ground termal and ground points.

Thats all i got.

Gimme more!

I like the Aluminum foil troubleshooting method.  My Focus was doing similar things.  Tach jumping all over, cruise control doing funny stuff.  Thought it was the Vehicle Speed Sensor.  Replaced the VSS and still had the problem.  Wrapped the VSS wire with Aluminum foil and the problem went away.  But that wasn't the source of the problem so I started wrapping other wires to find the source.  Turned out it was the #4 plug wire was breaking down and leaking electricity. 

Did you have to grpund the aluminium foil?

Dusterbd13 said:

Did you have to grpund the aluminium foil?

I didn't ground it. Just wrapped the wife after moving it away from the spark plug wires.    

Ground one end of the foil, yes.

Noise is so much easier to find if you have a scope where you can actually see and measure the stuff.

Is there a time constant that this repeats? If if seems intermittent, start timing it to see if it has a longer term pattern.

One source of noise that has not been covered is relays. Anything that switches on and off can have a flyback current that impacts the voltage on other lines. Good relays come with a flyback diode to prevent this, but not all of them do. Could this be the fuel pump relay cutting in and out? That would explain a seemingly random noise event.

I am guessing that you could filter the signal out if you wanted to dive down the rabbit hole of filter design.

Pete

wawazat said:

Dusterbd13 said:

Did you have to grpund the aluminium foil?

I didn't ground it. Just wrapped the wife after moving it away from the spark plug wires.    

I tried to wrap my wife in aluminum foil once, but I couldn't get her to stay still long enough.

Bad grounds are typically the most common source for electrical noise.  One thing to point out - if you have any shielded cables, ground only one end of the shield wire and not both ends.

T.J. said:

wawazat said:

Dusterbd13 said:

Did you have to grpund the aluminium foil?

I didn't ground it. Just wrapped the wife after moving it away from the spark plug wires.    

I tried to wrap my wife in aluminum foil once, but I couldn't get her to stay still long enough.

HA!  In my case it nicely complimented her tinfoil hat.