garage19
|
| posted on 31/10/07 at 07:13 PM |
|
|
Twisting sensor wires??
I spoke to a mate who said he has been advised by a semi proffesional ECU loom builder to twist all sensor wires together?
Th loom builder didn't give reason why so he has now asked me, but i'm struggling to come up with answer.
He thinks its to do with stopping interference. I know the EMF from one wire will have a small effect on the other but i'm not sure how twisting
them would effect this?? Or that the effect is big enough to worry about at all??
I have never seen wires twisted together in a stock engine loom from one of the big car manufacturers so can't see why this guy does it?
Surely the end result of twisting your wires is longer wires that will have a greater resistance?
Any body have any thoughts?
|
|
|
|
|
llionellis
|
| posted on 31/10/07 at 07:18 PM |
|
|
Here is one explanation, but I have never seen it done on auto wiring
http://en.wikipedia.org/wiki/Twisted_Pair
|
|
|
britishtrident
|
| posted on 31/10/07 at 07:22 PM |
|
|
Earth and signal should follow as closely as possible the same path to cancel interference, usual electronic practice is a "twisted pair"
as in wired ethernet.
Many sensor wires have an additional shielding sheath over the cable.
[I] “ What use our work, Bennet, if we cannot care for those we love? .”
― From BBC TV/Amazon's Ripper Street.
[/I]
|
|
|
RazMan
|
| posted on 31/10/07 at 07:45 PM |
|
|
I had all sorts of misfire problems until I shielded the crank sensor. I would certainly shield anything that gives a digital signal to avoid
problems.
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
Confused but excited.
|
| posted on 31/10/07 at 08:00 PM |
|
|
This is what used to be used in the old days of wireless equipment( radio to you young whippersnappers), before screened leads.
Tell them about the bent treacle edges!
|
|
|
zilspeed
|
| posted on 31/10/07 at 08:12 PM |
|
|
My only experience of this is the Dunnell supplied loom on our old zetec. Crank sensor was twisted pair.
If it works for them etc.... 
|
|
|
chriscook
|
| posted on 31/10/07 at 08:32 PM |
|
|
The CAN wires on more modern cars are normally twisted pairs.
|
|
|
Danozeman
|
| posted on 31/10/07 at 08:32 PM |
|
|
They stop interference. Most new vehicles has can bus systems. They are twisted wires for this reason.
Dan
Built the purple peril!! Let the modifications begin!!
http://www.eastangliankitcars.co.uk
|
|
|
mangogrooveworkshop
|
| posted on 31/10/07 at 08:44 PM |
|
|
prevents crosstalk.....hate telco copper twisted pairs.......when at work....
|
|
|
Bluemoon
|
| posted on 1/11/07 at 08:41 AM |
|
|
Yep minizes the magnetic coupling to other sources of electromagnetic (EM) waves (i.e ignition leads etc) by miniizeing the loop area between signal
and earth and hence reducing the magnetic pick up from EM waves...
A belt and braces method would see twisted pair for each signal, with a Faraday screen sheath (i.e. shield) insulated and only connected to earth at
one end. The farady screen then intercepts and the electric part of the EM wave, and earths any induced current via the low impedance (resistance)
screen.
Ideally a twisted pair should be used on balanced signals, not unbalanced ones (i.e not an earth wire and signal wire but a symetrical +signal and
-signal, so the "noise"/"interferance" you get is common to both wires and cancels out)...
Any how basicaly it's good practice to twist signal and earth for signal cables. For very sensitive stuff you might want a screen over the
twisted pair, insulated and earthed at one end only.
Cheers
Dan
[Edited on 1/11/07 by Bluemoon]
[Edited on 1/11/07 by Bluemoon]
|
|
|
MikeRJ
|
| posted on 1/11/07 at 09:53 AM |
|
|
As Bluemoon says, it all depends on the type of interface the sensor is connected to:
A balanced input uses a transformer or differential amplifier to literally measure the voltage difference between the two sensor wires. By twisting
the wires you ensure that any interfering signals are coupled onto both wires as equally as possible (known as "common mode" ), and will
therefore be ignored by the balanced input.
A singled ended input measures a sensors output with respect to ground, twisting the wires in this case will achieve very little. In this case
shielding/screening of the sensor wires is appropriate to prevent any interfering signals being coupled into the sensors in the first place. Using
twisted wires on high impedance sensors (e.g. knock sensors) that feed a single ended input is not necessarily a good idea, the extra capacitance
added by twisting the wires can have undesirable effects, and of course there will be no protection from interference.
[Edited on 1/11/07 by MikeRJ]
|
|
|
