Gary, Jeff and any other EE out there

I'm currently putting together some servo motors for work, the drawing calls out for the wiring harness to be shielded.
Also, the shielding is pinned out to the connector which I am guessing is grounded at some point???


Would shielding help or hinder a fuel injection harness???

Isn't the shielding only helpful when the voltages are very small? The FI runs on fairly high voltages I thought.
I'm no EE thou.....just a hack.

Geoff

I'm not an EE by training, but I do have an opinion....

Everything D-jet or L-jet is rather medieval - it's all switched by pretty serious transistors and the like, and most of the signals are analog, so brief spikes may be reacted to, but the reaction goes away when the spike is done.

If you are running a megasquirt (or other processor based systems), there is a microprocessor trying to recieve some signals and interpret them digitally. The one specific example (and problem) I have is the tach signal, for which I use the negative side of the coil. This signal is ridiculously noisy, and the MS does have some trouble ironing it out. The effect of this is that sometimes the megasquirt's idea of engine RPM flips up to about 15,000 once in a while, which causes a brief and rather serious over-inject (this tends to happen at low rpms), which causes the idle to be a little lumpy.

Having said that, the other RFI that is around the engine bay has not caused any problems for the MS, it has been suprisingly bulletproof, even though it needs to be dialled in better than it is.

The other datapoint I have is that the subaru engine I am working with - that doesn't seem to have any shielding whatsoever, and that is fully digital and micropro driven. So I guess they haven't had any problems.

So I guess it's nothing to worry about - you should clearly worry more about trying to bring your paint job up to the standard of my car

Shielding have been used on Porsches for years...mostly on the spark plug wires....and it's grounded...the idea is to keep the induction/high voltage inside and not interfering the signals for DME/CD units.....

You could do shielding on any wiring you want only if you see there's electrical/signal interferance on other part of electrical system....

Basically, yes the shielding are grounded....

j

I don't think shielding would help much like everyone just said. Ok, so EMI is not my field, but I think that because the stock ECU has no digital circuitry or high impedance analog signals (that I know of) shielding would not help you at all. The most it might do is reduce the chance of ignition noise getting into your electrical system and your radio, etc.

I agree that you will only reduce EMI/RFI. You might have problems though if the shielding is not done just right. Especially for the high voltage lines, this presents a very close ground point and could possibly make it more difficult to troubleshoot ignition problems if instead of having a few possible shorting spots you have a whole mess of them. Also if the connections between the grounding wire and the chassis corrode over time the problems could become worse.

-Tony

Mike, I'm not an EE, but did have this experience:

When installing my MIC3 ECU I encountered a problem. The spike caused by the colapsing magnetic field in the injector solenoids had enough inductive strength to trigger the ignition module via it's signal wire. As you know, ignition modules are just powered by a little transistor and it doesn't take much current to set them off. It drove me nuts trying to diagnose the problem. The ignition would fire in time with the injectors regardless of how I programed the ECU. Re-routing and shielding the ignition signal lead was the cure.

I just set this up as an example of situations that might require shielding. Any low-current signal lead ought to be shielded IMHO. Crank possition hal effect sensor, pointless ignition pickup, ignition module trigger, tach signal (if from source other than coil which is going to be noisy no matter what), etc.

-Ben

ps - I got a little milling machine. I'm so happy. Little lathe is next. Then I can make all sorts of fun little stuff.

Mike,

To minimize intererence keep the supply and return close, in this case in the same cable when suppling the load. Ground the shield at one end only. If the unit supplies the shield do not ground the other end. The reason for this is the ground current can induce noise. Let the power ground carry the current.
Also keep the control inputs remote (separate cables and routing path), from outputs.

In automotive engine systems:
In general, minimizing loop area (keeping load supply and returns on same path) and keeping distance in relation to noise generators (ignition primary and secondary, and injector circuits) from the sensing system (Timing, temperature, sensing, TPS, and MAP circuits) eliminates noise problems. This is a good practice even with all the built-in transient protection systems.

KitCarlson

Kit is right as far as my book learnin' taught me. There are
really 2 areas though that you may or may not need to be
concerned with; shielding your product from outside noise,
and minimizing noise you generate that would disturb other
nearby systems.

The type of signals inside the harness would be important
to know before you could figure the necessity and type of
shielding to use.

Bosch LH-jet has a shielded wire from the Hall Effect dizzy to the amplifier continuing to the ECU. I don't remember any on my Maxima engine. Both ends were grounded.

HTH,
Chris

On a stock d-jet system I am pretty sure you could put a Tesla coil in the engine bay and you will not have any ill effects (electrically speaking). If you start talking about modern EFI systems with microprocessors, knock sensors, and other high impedance circuits then you will benefit from shielding (so much that it may not work without). Things like injector lines are going to be low impedance no matter what, so I would not worry about those. It's the inputs to the ECU that matter more.

Been down that road...

When in doubt, always shield the wiring between a servo (or stepper) controller and its motor. Servo, stepper, and variable frequency drives are notorious for producing EMI because they throttle the motor by constantly "chopping" the current on and off, a lot like the original ignition systems in our teeners, but with WAY more current.

If the install permits, locate your controllers as close as practical to the motors themselves to keep the motor leads as short as possible. If you take ground braid and compress it along its axis, it forms a little tunnel and you can push your motor leads into it and make a great shield. Some motor connectors don't always provide a low impedance ground, you may need to find a chassis ground close to the motor connector. Tie the shield itself off to the ground, don't use a skinny jumper wire, it will reduce the effectiveness of the shield. Don't connect the motor side of the shield.

If you do this and your equipment still does strange things, your controller may be "leaking" EMI back into your primary power wiring. Add a line filter (or a BF capacitor at the controller input if you're feeding it DC) or find a better controller. Also, this time of year especially, make sure your moving parts are grounded in such a way to prevent static charge build up.

As far as shielding a teener's injection harness, the chopping rate is much lower than your servo's, the current is most probably much lower, and the entire thing is more or less shielded inside a big metal box (your engine bay) anyway. So, unless you want to add a little rice style detailing, don't waste your time.

So how big are the servo's you are working with?

The factory D-Jet and L-Jet FI harnesses are shielded, but the early Sixer engine harnesses do have a shielded section to them. The Sixer harnesses though have no FI, but they do have everything bu the ignition wires in them including the alternator cable and battery cable.

I have heard reports though that occasionally there can be erratic FI problems if the ignition wires (to the spark plugs) get close the the FI wiring harness. Its rare, but I have heard of it happening. Just separate the FI harness from the plug wires.