peejay

Grounding at the battery is probably part of the problem.

Scrappy_Brap

Mind elaborating on that some? And if not the battery where should I ground the coils?

1BADSP

iTrader: (4)

On a fresh rebuild you should not need to pour oil into the engine.even with used Parts it should start.unless you have big clearances.what did you set your clearances at ?.

Scrappy_Brap

I'm not sure I didn't build the engine, a local rotary guy built it for me.

peejay

The engine. Otherwise the spark current is going to have to go through all the ground circuits.

You really should be grounding the engine computer at the engine as well. MS3Pro is nice in that it has its own sensor ground return and a separate logic ground. I'd put ALL grounds to the same point and make that point on the actual engine. Cuts noise and eliminates issues.

I've fixed a lot of standalone installs where issues were installer-induced by grounding here, there, and everywhere BUT at a single location on the engine. "The OEMs only do that to save a couple cents, ground is ground everywhere." No, not really. Ground is only 0v on a local circuit level, there can be voltage differences at different points. If the hardware is assuming that 0v here is 0v there, 12v here is 12v there, then the only way to ensure that in a highly electrically noisy environment like a car is to source power and ground at the same point.

Ever measure voltage ringback from a relay turning off?

muythaibxr

Where does the MS itself ground?

muythaibxr

WE (the MS developers) are changing our recommendation to ground everything to the battery not the engine. You still want it all grounded the same place, but ground it all to the battery.

Grounding to the engine can cause odd issues... it's better than grounding all over the place if you ground to the engine in 1 place for everything, but it's not as good as going straight to the battery with everything if possible.

Ken

Scrappy_Brap

I will look to see where I grounded the ECU when I get off work, if I'm correct its grounded at a terminal block with jumper wires underneath the dash then ran one wire back to the battery to connect all the grounds that I needed for the wideband, fuel toggle, ect.

peejay

Good to know.

I never saw any recommendation anywhere in MS literature, I'd just gone by observation - OEMs don't have noise issues, OEMs ground everything to one (maybe two) points directly on the engine even if it means running many feet more wire (cost/weight), just do it that way instead of reinventing the wheel.

What issues have you been seeing? I still will prefer to ground at the engine (it hasn't failed me yet) as grounding at the battery is frequently problematic due to placement and logistics issues (hard to find non-OE battery cables with terminals on them) but now I want to keep a look out for potential issues.

Scrappy_Brap

The last time I started it the laptop kept losing sync with the ecu while datalogging (only would do it while dataloging, but gauges would work until I datalogged then it would lose connection) while running it randomly shutoff, didn't sputter it's way down like a miss or anything, just literally stopped

muythaibxr

That works well for OEMs because they spend a good bit of engineering resources on common mode noise rejection. We didn't.

peejay

So noise is the issues you've been running into? At what level? Noticeable on analog inputs or switching/frequency inputs or both?

My previous post may have been vague. "Just do it the way the OEMs do it" wasn't a command, it was my attitude towards avoiding self-generated problems. I like avoiding self generated problems, makes me tear less hair out. Following prior known-good examples is a great way to keep hair...

Scrappy_Brap

I'm not quite sure but I'm going I wrap the CAS wires with foil just to be sure the coils aren't intetfering (I keep getting lost sync code 11) but the crank angle sensor is functioning (or else it wouldn't run at all) Lost sync code 11 is a tooth error that means "Too few teeth before second trigger." Which leads me to believe their could be interference (but nothing has been confirmed)

Any input?

Aaron Cake

iTrader: (1)

If the engine is trying to fire but not quite, you may get sync losses.

But to confirm:

-CAS shielded cable grounded at ECU end only?
-pots turned fully counterclockwise?

Check the sensor gaps in the CAS. Should be the thickness of a business card. Adjust if necessary.

I've almost always grounded Megasquirt installs directly to the battery and really had no problems. I've also grounded to the engine and had no issues, as long as the engine is properly grounded.

muythaibxr

Usually you can see it in all analog signals... sometimes you just see it as a voltage offset in what the MS measures from actual... Most commonly you see bad CAS signals on rx7s causing them to not start when starting with the starter but to start fine any other time. Others have seen similar issues on other cars even with hall-based crank angle sensors.

Ken

muythaibxr

In my experience, if you have used shielded twisted pair wires for the 2 CAS signals and you only grounded the shield on one side, the issue is caused by dirty ground or 12v power.

Power should come through its own relay to the MS stright from the battery with the relay switched by turning the key to "on." Ground should be directly connected to the battery as well.

Factories spend a lot of money on making sure they filter all possible noise sources. Most aftermarket ECUs don't.

peejay

Don't discount relaying the MS power! I measured 1-1.5v drop on one car that used all of the OE wiring. (Not an MS install) 30 year old switches and wiring may have a little bit of voltage drop

So say one volt of drop, and battery voltage drops from 12.4v to 10v when cranking... that is 9v at the computer/injectors... uh oh! I don't know exactly at what voltage MS shuts down/flakes out but a lot of automotive electronics shut off at 9v, so I always use that as a rule of thumb.

Plus think about voltage correction. If your injectors and computer get voltage from different sources, the dead time calculation is going to be incorrect.

My MS2 system will go into relay-clicking mode well before the voltage is low enough to crank too slowly to fire. Definitely something to keep in mind.

Scrappy_Brap

Update: went back to aaroncakes base map, got the car started but once again fell flat on its face, lost sync reasons 11 and 17, started putting foil on the coils and the CAS to test my theory, now I am just left with code 11, noticeable when cranking and I lose sync with rpms and it and drops to 0 for a second then goes back up to 250-260 rpm, I didn't install a capacitor in line with the coils so I'm wondering if that'll be enough to help with the noise suppression. When running I was able to rev it (planted to the floor) to 4500 rpm and the moment I let off the gas it dies,

Any input?

Aaron Cake

iTrader: (1)

Pots fully counterclockwise?

Scrappy_Brap

Here's a vid of it running

Scrappy_Brap

Rev limiter set to 7k, laptop wouldn't even show rpms while running,

Aaron Cake

iTrader: (1)

That sounds like it's running.

If the laptop doesn't show RPM, then there is no communication between the laptop and the 'Squirt. Do you see any other sensors?

Scrappy_Brap

It didn't lose connection with the ECU, all sensors had feedback and we're reading accordingly on the laptop except for the rpm.

Eagle7

I don't remember the details, but you might need to pick a different signal name for the RPM gauge.

Aaron Cake

iTrader: (1)

RPM is one of those gauges that should just work.