Cable Shielding and Grounding

That's probably just for heat protection....

"The biggest misconception about "noise" in EFI systems is that it comes from magnetic or radiated sources.

I've never, ever once run into a problem where shielded wire solved it.

The Bosch Automotive Handbook has 4 or 5 pages that are GOLD. They explain that the real problems with EFI system noise comes from "coupling" which means, it comes in on power or ground inputs and is caused by other devices sharing the same power and grounds.
They point out how modern, OEM EFI installations are set up - the only shielded wire you'll see is on a Variable Reluctance (magnetic!) pickup sensor either on the crank, cam, or wheel speed (ABS).
The OEMs run a completely separate power and ground run to the battery terminals, separate from the high-current stuff that causes noise (fuel pump, electric water pump, ignition coil, fuel injectors, electric fan especially).
Instead what you should focus on, at least based on my experience, is keeping those power and ground runs separated from each other. Don't share the same 10' run of 10-gauge wire between the ECU and a fuel pump. Most of what I do is sorting these kinds of problems out and that's where I find most of my fixes for "noise" issues."

http://www.bangshift.com/forum/forum/bangshift/tech-section/36794-efi-shielding

Not having a drain path on a shield will most certainly effect how a shield acts but this may be a desirable effect in your case. The drain-less shield will mean evenly distribute any electrostatic interference onto all wires. Common Mode Rejection of the complimentary transmit and receiving pair of signals can then make the effective noise immunity to be greater than trying to prevent any noise from reaching the wires by the use of a drain path for the shield.

The circuit could also just be so noise immune that it doesn't matter if a shield helps or not.

It all depends on the rest of the circuit.

If feelings are strong, then ground it one end.

Next!

I work with shielded cables regularly, as a designer of control systems for process skids, where there is always at least one VFD to generate a wide array of noise to mess up my 4-20ma circuits. Both Brave Sir Robin and Solar Eagle have valid points. I've definitely seen the grounded connector (IFM Effector does this), and you really can't see this without destroying the cable, and have used shielded where I wanted extra protection for the cable. It is far harder to cut the conductors on a shielded cable with a slashing motion. There is even the possibility that the particular cable was used because the manufacturer got a whopping good deal on a cable with shielding that had the correct number of conductors. I've done that as well. If you have no interference problems, as you didn't mention any, don't worry about it. Any of the above could be the answer.

I neglected to mention that there are EMI problems with this system. The manufacturer acknowledges them and the users crab about it. The problems manifest as a frequent reset of the handheld unit. The cable was about 8' long and I shortened mine to about 3', which greatly reduced the problem. I can see the terminations at each end; the shield and jacket stop short and are shrink-tubed, and only the 4 wires proceed on to the connector. The resets are now infrequent and don't hurt anything. I am experimenting with capacitors from the +12 to ground, and ferrite cores around the power leads. And I will try a drain from the shield to ground and see what happens.

Sounds like a plan to me. I drag raced factory EFI in the 1990's. I often wondered how the after market units worked. I gave it up in 2000 due to bad shoulders limiting my ability to do the still required mechanical work.

It used to be how to change jets and advance distributor curves - now we need to know how to program and tweak electronics. Ever wonder what someone like, say, Don Garlits thinks about this? I guess Jack Rousch (Gapp & Rousch) has some idea of the electronic world of racing.

Not to hijack my own thread, but I'm amazed that most aftermarket distributors still use springs and bushings to set the advance. A few have dip switches and one has a laptop option to design a custom curve. But some SEMA technology is still in the 20th century.

That's why I mentioned Jack Rousch - he was in NHRA, but now in NASCAR, which is barely into the 1980's.You want hi-tech - check out F1. Those cars almost drive themselves now.

F1...those guys are superhuman. Imagine being slammed with high G's in all 3 axes for 90 minutes. I get sick thinking of it. But the cars are engineering marvels.

Hello SSCpal haven't talked in a while. Is there any unused wires in your cable like they were planning for other upgrades in the future not used now. You might try combining them all together to ground as you can also get cross talk with those extra wires. It has caused me a lot of headaches on some of my access control systems when they send cable with more conductors then needed or not shielded. Which system are you using for electronics on your home build system.

Hey Duke! No, just 4 wires, I'm guessing power, ground, RX and TX. It's a FAST EZ-EFI system and the long cable runs from the ECU to the handheld. The whole system is touchy; it needs a clean square wave for the tach input, and won't work with an HEI distributor 'cause the signal is so crappy. FAST says to run the power leads straight to the battery to minimize noise from other car sources. I've already put a little capacitor across the power leads at the ECU and ferrite cores on them. I'm gonna ground the shield and see if anything changes.

Is the cable just for tuning or is it hooked up all of the time. If it is just for tuning then power it from another power source, like a small battery pack when tuning (like a Lipo pack that is rechargeable) or another small lead acid like a 7ah alarm battery to isolate the power source from the vehicle that way no power glitches possible since it can't possibly register any signal variations from the vehicle power source. That way every thing is completely isolated from any false signals.

It's for initial tuning and then can be used as a display:

..the display resets are just annoying, they don't affect the EFI operation. And I think I can eliminate them with some simple fixes that the maker should have done in the first place. SEMA folks...don't get me started.

Yea I know, sometimes you would think they would use the good sense god gave a goose to get in out of the rain, but they just won't do it will they.

I added tin foil around the wires at the end of the cable shield all the way to the ECU. And that eliminates all the resets. Interestingly, it doesn't seem to matter if the shield is grounded or left floating. I went for long drives both ways with no resets. Now my EFI wiring is very close to wiring for tach, fans, water pump. fuel pump and other stuff. I cycled the fans without any reset. But the tach or pump wiring could be the culprit. I think by carefully routing the EFI wiring away from other wiring most folks will be OK.

You will still get capacitive coupling even if you don't make a galvanic connection as long as the foil is intimately close to a grounded metal object. Is the chassis of the ECU grounded?

Yes, the ECU and all electronics and wiring are mounted on an aluminum panel and the panel has a capacitor to power ground. Good catch!

Lotsa wires.

If you can, you might want to route/bundle the wires that carry power away from the wires that carry data/signals.