PDA

View Full Version : Braided ground stap or Cable? I'm curious if one is better than other?


CardinalFJ60
06-17-2013, 11:47 AM
I'm going to be cleaning up the circuits and electricals in the near future. My VWs typically had ground straps instead of a cable. Any advice on one over the other?

I currently have the David Dearborn massive cables...maybe overkill, but I'm curious about any advantages of strap vs. cable.

DaveInDenver
06-17-2013, 11:56 AM
For what purpose, simple grounding (e.g. DC circuit returns) or overall? With AC and particularly RF you need to consider skin effect. For the simple return path for your battery it's doesn't matter, the largest cross section (or equivalent thereof) is what you want regardless of construction. The reason straps and high count stranded cables are used is practical, they flex and are easy to route. To some extent using them also helps reduce EMI as they work better as conductors than antennas for the pulses created by the alternator, ignition and the digital noise produced by all the computers in cars. But in general it's not important. Now for RF paths, braided cable is a necessity as you've probably seen mentioned often. It's absolutely critical to having any usable HF in a mobile situation.

Rzeppa
06-18-2013, 06:11 PM
I'm going to be cleaning up the circuits and electricals in the near future. My VWs typically had ground straps instead of a cable. Any advice on one over the other?

I currently have the David Dearborn massive cables...maybe overkill, but I'm curious about any advantages of strap vs. cable.

As mentioned, braided cable's main advantage in automotive applications is it's flexibility. And frankly, it's "skin effect" advantage in HF applications is way over-rated when you run the numbers. That's more marketing hype than real physics. If I were you, I would use what you already have, especially given the cost of copper these days.

Rezarf
06-18-2013, 07:21 PM
Anything with high strand count... that's what I have been told and seems to be working.

CardinalFJ60
06-19-2013, 07:07 AM
thanks for the great replies. what I've got is more than sufficient and it's just the battery grounds I was referring to.thanks again.

...here's a follow up question. Is a ground connection a ground connection, that is, is it 'grounded or not'? or maybe a better question is how can you test for the quality of the ground?

DaveInDenver
06-19-2013, 07:50 AM
Resistance.

But just because it shows conductivity does not mean it's a sufficient return path. Remember Ohm's Law, V = I * R. A circuit that flows small current will tolerate a higher resistance, but a circuit carrying significant current requires low resistance to prevent voltage drop and heat. Grounding is not really the correct term, it's return meaning the path for current back to what forced the current to start with. The feed and ground conductors must both be sufficient to carry the current that is not used for work and not impact the circuit much.

Problem is a handheld DMM is not accurate enough to really measure, so you have to do some of this analytically and verify workmanship with a test tool. IOW, if you know your current and target voltage drop, you can calculate the size of wire you need. Say a winch cable, where you want resistance to be in the tens of milliohms. It's not practical to measure your return to verify, so once installed you measure and are satisfied that your meter reads zero. You know it's working, your connections are fine and that it's something less than the minimum resolution is of the meter. Thing is most all wires are going to look like zero ohms to a lot of portable meters, so unless they are exceptionally poor you can only say its go/no-go and have to engage the maths.

CardinalFJ60
06-19-2013, 10:39 AM
this is great. I've been experiencing weird voltge drops for years, and always figured it was an overtaxed alternator. As I re-do much of this wiring I'll do the ohms law trick. Good info on handheld MM, I tested resistance and they are, from what I can see, measuring like a hundredth of an ohm. or maybe not. given the sensitivity of my radio shack mulitmeter.

DaveInDenver
06-19-2013, 02:15 PM
Is it a digital or analog meter? In any case you have to know what your meter is telling you when making a measurement. For most stuff in vehicle wiring you are just looking for either an open or closed state and the actual measurement is not important. I'm betting that if you're using a typical 3-1/2 digit DMM which means you're probably looking at a minimum resolution of probably 100mΩ on the 400 Ω scale and an accuracy of around +/- 250mΩ. Some can be more accurate (I have a Fluke 87 that has a special 4-1/2 digit mode that allows 10mΩ resolution and is good to +/- 0.2% + 2 counts on the minimum range), but it's still best if you calculate needed wire size and validate it using voltage instead of resistance. Most DMM have 10mV resolution on a 40V scale and their accuracy will be decent enough for practical use, say within about 50 mV at ~12V.

TIMZTOY
06-19-2013, 09:50 PM
I didn't read any posts , but I always use the thickest gauge battery cable from the parts store.
There easy to get ahold of, and are cheap..

Rzeppa
06-19-2013, 11:17 PM
Problem is a handheld DMM is not accurate enough to really measure

I have yet to encounter a modern DMM that isn't accurate enough for general automotive applications. Even the little $4 jobs that HFT sells. They work just fine for our automotive applications. We're not measuring cold fusion here.

I spent several decades in metrology, and spent quite a bit of time up at NIST in Boulder and the Danish Institute for Fundamental Metrology in Copenhagen. I certainly know uncertainties and sources of error. The latter will get the uninitiated, but not modern DMMs' accuracy.

FWIW, "Accuracy" is a combination of precision (repeatability), and calibration to some accepted standard, such as NIST, and of course resolution. You can have resolution to however many digits past the decimal place but still have poor precision or be uncalibrated. You can have great precision and calibration but insufficient resolution. When you have all three, you are generally good, but still need to know how to apply your instruments to achieve their potential.

For milliohm measurements we always use a Kelvin connection for our milliohmmeter. Using a milliohmmeter, you can use a regular DMM in parallel and discover techniques that allow a regular handheld DMM to give you really accurate measurements, so that when you learn the techniques you no longer need the fancy milliohmmeter and kelvin connections for your automotive troubleshooting and diagnostics.

DaveInDenver
06-20-2013, 05:43 AM
I have yet to encounter a modern DMM that isn't accurate enough for general automotive applications. Even the little $4 jobs that HFT sells. They work just fine for our automotive applications. We're not measuring cold fusion here.
Yes, exactly. My point, apparently poorly stated, was that most portable DMMs are not sufficient to measure the true resistance of a large size wire. It's just beyond their design criteria, even though a $4 ICL7106-based meter can be quite good, you are just at the limit when you are operating your ADC at the LSB range on the V x T curve with budget-minded op-amps. I also said that it does not matter for this because as you have followed proper engineering principles when you sized the wire, so you are just validating your workmanship with the meter.

The Fluke 87 I carry considers conductivity for purposes of making the little beeper beep anything up to about 100 Ω, so in the same vein no reason to be hypercritically worried about 5 decimal places and resolution. The only reason I even mentioned it was to point out limitations with DMMs at the edges of ranges. With 3-1/2 digits on a 200 or 400 Ω scale there is a lot of error for measurements lower than a few ohms and certainly so for less than 100 mΩ or so, both due to resolution and accuracy.

U4JFeU-o2kc

rover67
06-20-2013, 08:57 AM
Yeah i think I got it, I can't put my DMM across a big fat wire and measure its resistance. It just reads zeros.

Even a small wire for that matter.

or even a junction sometimes just measures zero....

thefatkid
06-20-2013, 08:58 AM
Resistance checks are really worthless in most automotive applications. A correct way of diagnosing a problem is using voltage drop tests. Measure from after the load on any circuit to the negative terminal of the battery. Any voltage over a couple tenths shows a possible issue on the ground side. A DMM does not load a circuit and does not provide any real valuable information.

Inukshuk
06-20-2013, 09:20 AM
Times like this we need that little popcorn emoticon.

Rzeppa
06-22-2013, 02:04 PM
Resistance checks are really worthless in most automotive applications. A correct way of diagnosing a problem is using voltage drop tests. Measure from after the load on any circuit to the negative terminal of the battery. Any voltage over a couple tenths shows a possible issue on the ground side. A DMM does not load a circuit and does not provide any real valuable information.

Excellent observation Brian!

In fact, the best way to validate design and workmanship in a high current application is to measure voltage drop across whatever is suspect, whether it is a connector or a conductor. Especially suspect is the return as you mentioned. I like to use star washers on frame and body connections so they dig into any coatings, whether oxidation or residual paint after it has been sanded.