A/B USB switch box noise

[whiterabbit]

On Saturday I built an A/B switch box for my home studio so I can easily direct communication between my audio interface and my two computers.

I built a working switch box, but I want to make a better one. The one I built over the weekend adds a fair amount of fuzz/static in the high frequency range. Mostly noticeable only when the monitors are sitting idle.

I used 3 female USB jacks, a 6PDT slide switch (on-on), and added separate power indicator LEDs in parallel for each of the INs before the switch and a third indicator AFTER the switch- each with a 220O hm resistor.

What can I say? I like to make stuff that glows...

Anyways- I want to make a second model that doesn't affect the audio passing through it. (As much as possible, anyways- I know that any additional hardware will have an effect to some degree)

I'm wondering if the LEDs are causing the distortion, or maybe the quality of the materials is too low? Both?

Any and all help is appreciated, as always!

 

[(*steve*)]

Are you switching input signals (from your signal source) or output signals (to the speaker(s))?

How are you powering the LEDs?

Where does the noise come from, the active speakers (computer), or the inactive one?

Oh, and where is USB involved?

I may have other questions depending on your answers to these.

 

[whiterabbit]

First need to correct a small mistake- I actually used a 4PDT switch, not a 6PDT. Doesn't actually matter in terms of the circuit but wanted to fix it anyways.

To answer your questions:

My current setup includes two desktops, both connected by USB to my audio interface, which then goes on to connect to my studio monitors (by TRS, if you're curious). The audio interface only has one USB connection so I can only plug one computer (source) in at a time. I can quite easily just switch the input cables, but instead chose to view this as an opportunity to build something useful on my own.

Note that the schematic sketch doesn't accurately reflect where each of the four USB wires is physically connected to the switch on the finished product, but was just a quick reference for me to know that each of them connects to a different pole. The principle of it is all correct though... hopefully that made sense.

So there is a blue LED on each of the two USB inputs to indicate that the cable is connected correctly, and a green LED right before the single USB output that goes into the audio interface. (as a side note, my original intention was to have only one blue LED on at a time to indicate which of the two inputs was being used, but I realized that I needed a switch with more poles and just built the rest of it anyways)

The noise comes from the two monitors when the switch box is plugged in between the source (computers) and the output (audio interface / monitors), and is the same for both inputs (computer 1 and computer 2). It is present when there is nothing being output from the source, and also when sound IS being output.

Also, the output cable looks loose in the image where the box is actually plugged in and on, and I just wanted to be clear that I DID reseat the connections multiple times to make sure they weren't just loose. I also swapped the two input cables around a few times to see if that made any difference; it did not.

 

[(*steve*)]

OK, now I understand.

It's a very curious problem because there is no simple way that this could introduce noise into your signal. (At the point the signal goes through your box it is digital and highly immune from noise. Noise would cause errors, and these are not likely to be heard as noise)

The only thing I can think of is that the ground connection is poor and that the ground connection is being made some other way, but even that isn't certain.

Do toy get the noise if you only have one computer plugged in?

the test would be to connect a source to a single computer. Check the noise. Now replace the cable with your switch box and see if it makes a difference (try both ports).

If neither give any noise, try plugging in the other computer. Are you getting noise now?

 

[whiterabbit]

I hang my head in shame and offer you the cord of my soldering iron [<<< awkward Game of Thrones reference...].

You nailed it- I didn't properly connect the casings of the USB cables. This morning I turned everything on and laid a piece of scrap wire across the metal case of the input over to the metal case on the output and it completely eliminated the hum.

The good news is that because of the open-side design I used, I can easily solder a permanent connection between the jacks.

I guess I'm a little confused on why the casings (is that an accurate word, by the way) need to be connected as well, or how it causes the noise. Can you please explain that, OR point me to an article that explains why? After posting this I will also start poking around Google; Sometimes I have a hard time finding the answers I'm looking for because I don't know how to formulate my queries properly.

Thank you so much for your help!

 

[Arouse1973]

Yes Steve I think your right sounds like it might be a ground loop problem to me.
Thanks
Adam

 

[(*steve*)]

Can you please explain that

At this point I should offer you a gem of information showing my true depth of understanding and knowledge.

However, I can't.

The one thing you learn is that you need to respect ground wiring, especially in combined analog/digital circuits.

I have a few half-formed ideas as to why this might have happened, but it's really hard to know for sure why it happened.

 

[whiterabbit]

I'll keep poking around and see if I can find any possible answers to that question. Knowing WHY something works or what it does helps me remember it. Which really screwed with me in a lot of math classes because the teachers could never give me a realistic situation where I would need to use an equation in my life. But that's besides the point.

I asked a friend who is minoring in electronic engineering right now and he said he thought that connecting the case to ground helped to shield the wires from any external fields that may interfere. He mentioned electromagnetic fields, and then stated that he wasn't exactly sure... but the point was that it keeps the in IN, and the out OUT.

So when I go back to fix this, should I connect all three jack casings to each other with one piece of wire OR should I connect the case of each jack to it's own ground? (ie connect case of input jack 1 to the ground of USB input 1, case of input jack 2 to the ground of USB input 2, case of output jack to the ground that is being output)

This morning when I did my quick test, the wire I used to connect the casings was going AROUND the switch instead of connecting THROUGH the switch. I guess my concern would be that if I then also connected the 2nd input in the same way AROUND the switch it might be potentially dangerous to the equipment somehow. Maybe I'm just overthinking it...

 

[whiterabbit]

Seems legit... http://www.newark.com/pdfs/techarticles/alphaWire/USC.pdf




[edit]
After reading through that, I think it probably makes more sense to have the shield connection connect to the respective grounds that then pass THROUGH the switch. No need to shield the unused input.

 

[Arouse1973]

Try connecting everything you are using to one mains extension lead. Even if it means moving the equipment into one room. Tell me that helps.
Adam

 

[whiterabbit]

*RESOLVED*

Today I connected each of the three jack casings to the ground lead of the LED closest (respectively).

Now the switch is working exactly as intended, no noticeable noise!

Thanks again for the help!!

 

[KrisBlueNZ]

After reading through that, I think it probably makes more sense to have the shield connection connect to the respective grounds that then pass THROUGH the switch. No need to shield the unused input.

Right. It's probably best to keep the grounds from the two computers isolated from each other. Run everything through the switch, so the unused computer is completely disconnected from the path from the active computer to the audio box.