Sensor-, Data- and ignition trigger-wiring twisted shielding useful or useless?

Hi all,

Very first post for me over here. I've followed both Fundamentals and Practical Wiring Club level and currently standalone EFI Tuning..

I'm modifying Motronic M4.4 ECU for Volvo 5 bangers. Some smart people reversed engineered the ECU and they came up with ability of WBO2 controlled fueling, Coil on Plug conversion, being able to use unused sensor inputs and log. Volvo 850 now has ability of OBD2 instead of OBD1 cause ECU upgrade (from M4.3) is fully pin compatible. It's for learning and applying in practice for future standalone ECU (tuning) with custom wiring harnas.

Now I've designed a PCB with connectors which fits within ecu casing. For every extra sensor signal wire, OBD2 data (K-line) wire, WB02 signal wire and COP trigger wires I also thought of adding a grounding wires in and create concentric twisted wires just for shielding (of courses grounded at 1 side on location of grounding ECU). My reasoning "while I'm at it it's not lot of extra work" to mitigate EMI on these input/outputs.

Over here I only read usage of completely shielded wiring for trigger sensors and knock sensors. Is my effort of using twisted shielding wire useless or can it suite a purpose? Cause when I create something which solve a problem which doesn't actually exist it's a waste of the little effort right.

Any thoughts?

It doesn't sound like you're doing something that requires shielded wire, unless the wires will run along something you expect to emit lots of interference.

@Mike so you don't think twisted grounding is needed? What rule of thumb can I apply in which cases twisted grounding can help? So I'm not referring to shielded wiring.

Volvo don't seem to use shielded wiring for knock, crank or cam sensors. The configuration with distributor creates some audible hissing signal on audio and changes with RPM. Tried to mitigate by wiring in common ground points to combat earth loops etc. Doesn't solve the issue. So I might got paranoid of creating my own harnas to later find out all kinds of EMI issues will popup. Maybe I'm overthinking it. I'm no electrical engineer though.

It's possible I'm not understanding what you're wiring, but I'm reading your question as if you are considering twisting and shielding analog sensor wires. That generally isn't required unless you're running them along high output ignition wires, alternator, something that generates a large amount of interference.

Hissing in your audio system does sound like a ground loop has been created.

@Mike I'm wiring extra extra capabilities of reversed engineered ECU (WB02 analog input (from ALM controller), K-line output (log+flash+OBD2 reader)). I'll add pictures. Probably provides you more insights.

Now I mainly focus on wiring in the OBD2+WB02 capabilities which originally car doesn't have. Wiring is going through firewall cause WB-controller is more isolated from environment engine compartment And connecting to laptop for calibration WB analog output and diagnosing is easier from within car.

I already screw up "OBD2+WB02 harnass" cause I overheated heat shrink so "OBD wires" overheated and melted a bit. No short circuits though. So I need to do a second attempt. Yep learning the hard way. GX20 connectors on Aliexpress looked like a cheap great alternative. But soldering 1mm2 power wires to pins are headache. Bought new ones which I'll likely fill with 2K epoxy so no problem for vibrating loose.

As I previously mentioned it's all to learn and apply HPA courses. For now I'm using cheap PVC wire. How long does it have to last... When i'm capable of DIY wiring up to standards and I'm able to tune my car a bit on Factory ECU I would like to jump to standalone ECU. A safeguard for cutting engine for eg. low oil pressure too high EGTs, logging intake air temp, compensation on Ethanol contents, map-switching and tuning for running on LPG etc.

The bit offtopic thing and less priority for now brought up by myself; hissing RPM audible stereo. I've wired both headunit, GROM CD emulator and aftermarket amplifier grounds to same grounding point near rear seat. Problem still exists. Let's ignore this one for now cause it skews topic on my more urgent ECU wiring question.

I've used Micro-molex connectors cause they fit tight space and can be easily disconnected to be able pull metal lever to unmount ECU from holder. This way I can bench-flash ECU. In wiring course it didn't pick up rule of thumb for fully shielded cabling and effective measures, hence my question on twisted shielding ground wires. How about power and signal stuff in same harnas? Rule of thumb for that stuff? Somebody recently told me EMI will become an issue on high frequency circuitry not so much on DC-stuff And devices inducing huge oscilating magnetic fields (eg. sparkplug wiring) can cause problems.

OK, I stand by my original statement that shielding, concentric twisting aren't required for what you're describing. It doesn't mean you can't do it, or it would be bad to do it, but as you said you don't want to add complexity where it isn't needed.

Your current implementation doesn't appear to be weather sealed, so I would be concerned about potential corrosion over time, possibly water intrusion depending on how well the top of that box and your passage for the wires seals up.

Mike your eyes are not deceiving you. It is not weather sealed. ECU is mounted in a plastic box with a lid on top and on as entry of duct for pushing air in from behind the radiator. Over time corrosion will have an effect. Those scrapper ECUs have some corrosion on pins in connector foot as well. The metal lever, to have ECU fixed and secured, has some restrictions for location connectors. This is the 'smartest' thing I could come up with those Micro Molex connectors also with height restrictions (the lid) in mind.

Following your reasoning. It would make sense to have only sensors have a twisted ground. Path of wiring will be 1 to 1.5meter parallel with ignition trigger signals. While we're at it.... Would switched 12V (+ GND) make sense to pull from ECU to feed WB-controller? Now these wires will run parallel in harnas. Or splice in from other switched power location better for signals (K-line/OBD2 + ANOUT WB-controller)? Almost entire factory engine wiring runs in a duct at the front of engine about 20cm from alternator and some distance from starter.

General question could be; what can I do to mitigate potential EMI problems? Better address potential issues instead of tracing weird input/output behaviour and chasing your tale for many hours. That's my logic.

I don't know the electrical system of that particular vehicle, but generally I use a different switched 12 V power source for ECU and wideband controller. The wideband heater can draw significant power, so when adding a wideband to a vehicle with stock wiring, I typically don't want that additional load on the ECU circuit in case it may overload it.

When creating vehicle wiring from scratch, and when all loads are known, you can plan to power the ECU and wideband on the same circuit. When I've done this I haven't had an noise issues. With that in mind, it might be fine on your vehicle, I just don't know the current max load on the circuit, how much capacity it has to spare, etc.

Have you completed a test install and experienced EMI issues? If not it seems it would be quicker to lay a unfinished harness in place without wire sheathing, start and idle then rev engine in neutral while monitoring to see if signals exhibit large amounts of noise or not.