Shielding a passive audio switchbox

This is because despite the box having all of the inputs' shield/gnd lines tied together, yet they themselves are all each discreet elements.

So, there would be a chance that IF one were to tie all of those input lines' shields together at their lead ends, instead of at the switch box end, the loops would subside (not even be present). Of course, this is typically a logistical issue if the input sources are distant from each other. If they are all in the same stack or locality, it is a bit more feasible.

Some say at that point then that the tail end (at the sw box) needs no common ties at all.

The snail mail method is to get RCA terminators for all inputs and hand terminate the open feeds and plug in the wanted feed.

So, ideally, the switch box would be switching IN a terminator on all open feed ports, while un-terminating and connecting the desired input. Then, the shielding issue would surely go back to the shadows of attenuation.

The opamp's configured as an summing inverter with unity gain. Input audio signals are inverted and then /summed/ together at the Output. Normally, only one audio signal is present at any given time. That said, if for some reason you want to simultaneously listen to music on Input 1 and the ham radio on Input 2, there is no attenuation of audio signal. When +1V appears at Input 1 and +1V appears at Input 2 the Output goes to -2V. You can prove it to yourself if you happen to have a copy of LTSpice available. There's an LTSpice file available for download [1] that simulates the circuit. The simulation contains three sinusoidal inputs. The amplitude of each input is 1V. The frequency of each input is different. That makes it easy to see how the input voltages sum together at the output.

Note.

1. formatting link

Thank you,

--
Don Kuenz, KB7RPU 
"To invent, you need a good imagination and a pile of junk" 
  - Thomas Alva Edison

I build my first hi-fi audio amplifier about a half century ago. In those days it was (at least in Europe) standard practice to keep chassis ground and signal grounds separated as long as possible and use a _single_ signal ground connection to the chassis.

You may dislike DIN connectors at much as you like, but the connector shell could be installed directly to the chassis, the signal ground (pin 2) was by default separate from the chassis ground.

Unfortunately brain dead RCA and 6.35 mm jacks had by default a connection between the signal ground and chassis ground.

You may get rid with this in small systems, but in larger (stage) systems with multiple devices with more or less leaky mains transformers, you end up with all kinds of hum problems.

I have solved some guitar amplifier hum problems by simply using isolated 6.35 mm jacks with all signal grounds connected directly to the input tube.

Of course, XLR is even better, with the connector case connected to he device chassis, the shield possibly connected to pin 1 and the actual signal lines separated from all this ground mess :-).

This applies to mains/oudio frequencies, for RF (at least UHF/SHF) you really have to use BNC/TNC/N/SMA connectors with direct chassis connection for proper transmission line operation.

However, if you want to feed and audio signal to such a device (e.g. amateur radio transceiver), you should use some audio isolation transformer.

impedance."

So it runs on current like a bipolar. Big deal, the result is the same.

If you don't believe me get a breadboard and an OP AMP and experiment for yourself. Seriously, I wish you would. If I was loaded with money I would buy you the shit to do it.

available."

Fuck a simulator,it is easy enough to do for real wwith a beadboard That is what I told Pimpom to do.

If the other two inputs are grounded, that is 5 K to ground shunting the si gnal. How can it not be ? As such it shunts part of the signal on the used input. If the unused inputs are an open circuit then 10K is correct, but I would bet they are low impedance which makes me right.

Try it for real, in fact you can do it with one transistor in a common emit ter configuration. I would bet on it proving me right, IF the unused inputs act as a ground because of being low impedance. Disconnect them and it is different, but then why not just use a switch ? The supposed advantage of t he Altioids or whatever is that you can leave everything plugged in at the same time.

Try it sand see. A transistor, six resistors and a few caps.

chassis ground and signal grounds separated as long as possible and use a _single_ signal ground connection to the chassis. "

Yes but all the signal grounds were common TO EACH OTHER before being connected to the chassis.

Yup.

shows a very nice diagram of the layout.

There _is_ a difference between the way a linear supply and a switching supply use their transformers. In a linear supply, the transformer comes "first", before the rectifiers - the rectified DC voltage is already isolated from the line by the transformer.

In most switchers, the rectifiers come first, creating a high-voltage DC which is *not* isolated from the mains. This high voltage is then PWM'ed into the transformer, which feeds it to a second rectifier stage, and you end up with a lower-voltage DC which is isolated from the mains.

The feedback from the switcher's output stage, back to the PWM logic, is also isolated - either by a transformer, or capacitively, or via an optocoupler.

Now, there are a few (very few) "off-line" switching power supplies which produce a DC voltage that is not transformer-isolated from the lines... they may have capacitor isolation (good only for very low currents, I believe) or simply rectify and then buck-convert down to a lower DC voltage. As far as I can tell, these sorts of non-isolated converters are limited to use in "sealed" equipment, where there's no possibility of human contact with either side of the connection, and are not allowed in "wall wart" supplies or laptop-computer power bricks, or in desktop computers/servers either.

It's not a case of believing or not believing you. I'm not an audio specialist but I have quite a lot of experience designing, building, troubleshooting and simulating audio circuits -

*including* the summing amplifier we're talking about.

It's a well-established elementary building block which is fairly common in audio and other circuits. It's solidly based on theory and well proven in practice.

And I could do without that patronizing last para in your post. Many regulars here at s.e.d. wouldn't be as restrained in responding to it.

You keep telling others to try out your theory in practice. You keep saying that you "would bet" that the result would prove you right. You become vulgar when someone politely suggests a means to check out your theory.

Why don't *you* try it yourself?

Many regulars here at s.e.d. wouldn't be as restrained in responding to it. "

Then try it. It is that simple.

switching supply use their transformers. "

I know that, but the brainiac over there thinks it is not a transformer. I would like to know what he calls it. But know-it-all has no time between baseless insults to say anything useful.

The opamp's jelly bean magic changes what, at first blush, may look like a simple resistor network. Let's review some ideal opamp basics. No current flows into the inputs of an ideal opamp. The non-inverting input is grounded in my circuit so the ideal opamp holds its inverting input at virtual ground. Kirkoff's Current Law implies that the sum of all input currents must then go out of the non-inverting node through Rf. Here's the well known equation for a summing opamp as applied to my left channel input:

Vout = -R9(V1/R2 + V2/R3 + V3/R4)

When an input is grounded, the numerator for that input becomes zero and its term drops out of the equation. Looking at it from another perspective, when an input's grounded the opamp's inverting input is held at virtual ground, so no current flows through the input resistor from the input ground to the virtual ground at the non-inverting node. As expected, nothing happens when you disconnect audio lines from the audio input jacks. The sound doesn't become louder.

Thank you,

--
Don Kuenz, KB7RPU 
"To invent, you need a good imagination and a pile of junk" 
  - Thomas Alva Edison

I don't have to "try" it. I have *used* summing amplifiers with low impedance sources in several of my designs - both audio and others - from the vacuum tube days to modern solidstate circuits.

You come up with a half-baked theory disputing something that's explained in every elementary book on electronics and then keep telling others to try it out. You refuse to do a simulation that will so easily prove you wrong.

*You* try it and you'll see.

You're AlwaysWrong.

AlwaysWrong strikes again.

to check out your theory. "

Vulgar ? How ? I didn't mean to come off that way, maybe you are used to th e highest forms of etiquette, but not me. his is an electronics forum, and haven't you read Phil or the brainiac over there ? Now with him he is going to inoke vulgar if he keeps it up.

I just might do that. I lost my camera though so I have to borrow one. I am sure I can dig up an OP AMP, and I know a guy learning who has a breadboar d. I'll get the shit together, you can see the scope waveforms even in the pictures as I ground the unused inputs. I am that sure of this. If the extr a inputs are open circuit, like I said, no, but in real life they will be e ffectively a short.

Alright, you MIGHT be right. I will try it. Put this to rest once and for all.

"AN954 Transformerless Power Supplies: Resistive and Capacitive" [1] is a fairly good resource. So is "Transformerless Power Supply" [2]. Experimenters need an isolation transformer to scope their circuit. And it turns out that the hospital grade "isolation transformer" sold on ebay may not be so isolated after all. Although it's almost an hour long, Toddfun's video [3] shows you the pitfalls. It shows you how to mod the ebay transformer to make it truly isolated. It also shows you how to use a light bulb to verify your isolation. The light bulb is very important. It prevents crowbar mains shorts.

Note.

1. formatting link
2. formatting link
3. formatting link

Thank you,

--
Don Kuenz, KB7RPU 
"To invent, you need a good imagination and a pile of junk" 
  - Thomas Alva Edison

-----------------

** You are thinking of "transformerless" PSUs, not "switching".

** Typically, an AC mains capacitor is used to drop a most of the AC voltage so that a low voltage DC supply can be created. The great advantage being the cap creates no heat as using resistors would. ** More simply, they cannot have an external output connection but are simply incorporated. AC powered smoke detectors/alarms are one common use.

.... Phil