Fuse size question

[avanti]

Originally Posted by Nic7320

When radios use chassis ground for signal returns, every RCA cable going to another unit creates another ground pathway somewhere. Hopefully these are routed and bundled together so the loop area created by the ground mesh is not picking up too much RFI.

Good grounding can be done with a mesh, since the inductance and resistance is less. An N x N mesh of any arbitrary size with the same size links has the same resistance across its diagonal as any single link.

Sorry, I just don't buy this. When radios use chassis ground for signal returns, they are mis-designed or broken. No EE who understands the issues would do such a thing.

Chassis ground and circuit ground are two completely different things (just as safety ground and neutral are different things). The chassis should be bonded to the vehicle, and the circuit should get its ground from the -12V wire. They should NOT be connected together (except perhaps through a current-limiting resistor). Any competently-designed modern circuit board will dedicate an entire layer of the PCB to a ground plane, and all devices will get their ground directly from it. In the same way, all devices in your van should get their grounds directly from the chassis. Period.

RCA cables should be grounded to the chassis on ONLY ONE SIDE, else there will be horrible ground loops. When dealing with improperly-designed equipment, this can be accomplished with an isolation transformer, such as this:

81SUoWmjgfL._AC_SL1500_.jpg

Now, I freely admit that a great many mainstream a/v devices ignore these rules, but they are all wrong. If this weren't true, the above device would not be necessary. But, it is a band-aid, not an engineering technique.

Your "mesh" comment is a bit of a red-herring. Obviously, a mesh of evenly-spaced uniform conductors is a good approximation of a solid plane (after all, a solid piece of copper is simply a 'mesh' of atoms). But this is very far from the situation in which you are making interconnects in a van.

Ground loops should always be avoided, and ground/neutral isolation should be observed at all levels of a circuit.

[Nic7320]

This is going out of scope in an RV thread, but since you brought up sone good points, let's look at them. Here, your physics and audio background serve you well.

Star point grounds are great when they work. Often in small, well controlled circuitry.

But they create a whole pile of other issues, such as:

Signal Integrity (signals need a short and predictable return pathway)
Impedance control (RF signals need a constant impedance for transmission lines to maintain a constant impedance and not create reflections)
Shielding (wrapping a single-ended signal with a ground sheath or an entire circuit in shield gives it a nice faraday cage)
Ground bounce (If you have a signal sharing a poorly controlled ground path, you can pick up unwanted noise such as induced RF or L di/dt voltage spikes following that same ground path.)
Limited common-mode signal compliance (even if you use differential signals -- such as LVDS that have their own return path -- they still only tolerate a limited amount of common-mode ground voltage variation before the receiver runs out of differential signal discrimination. To keep this in check, several grounds are sent down the cable)
EMI compliance: (this is where everything is RF, and needs lots of grounds to contain it)
Electrical Safety: (needed to reduce electrocution risk)
Lack of redundancy. (not every ground connection is guaranteed to be a good ground as systems age, due to corrosion, mechanical abuse and configuration changes, like unplugging something)
Mechanical mounting (holding down a circuit board usually requires several metal screws)
Thermal issues (similarly, getting heat out of a circuit board usually requires lots of metal interfaces and fasteners)
Cost (Often isolation measures are too expensive, such as isolation transformers or boron nitride filled thermal strips to isolate grounds. But yes, sometimes they are absolutely necessary, like the thermal strips that saved a program I was on from a complete redesign

Keeping on topic, I discovered my 12V TV in my RV does not shut off when the DC ground is disconnected;
it just finds another ground path on the coax through the Winegard preamp.
I discovered this when I powered it with a little 12V switching box that oddly switched the ground off instead of the +12), so the TV stayed on. TVs can, like you point out, isolate the RF ground with a RF transformer, but that cost money and a some fringe signal performance.

A car radio also has multiple ground connections through the antenna and RCA connectors. That isolation transformer can solve some noise problems, but they can also be solved with ferrite beads or DC filters by making the DC ground pathway a poor high frequency pathway.

So I'm not ready to break all these redundant grounds in the name of single point grounds. There are plenty of reasons for them. Especially on a flight deck where another aircraft's radar system is pointing right at you.

[avanti]

Most of the above is deflection and has nothing to do with the topic. I recommend that anyone with a serious need consult reliable sources and make themselves conversant with best practice.

This topic is closed.