HELP with designing a headphones distributor please

[sonolink]

Box A nearly finished





There's WAY too many B boxes



The board is looking good and fits fine...



Cheers
Sono

 

[sonolink]

Hi guys,
Following my overcompulsive paranoid instincts I decided to test that everything is working ok before soldering and mounting 20 B-boxes to find out that the volume pots or the switches aren't behaving correctly.
The good news is that everything is actually behaving correctly.
The "not so good" news is that I seem to have spill. Let me explain myself.

First this is how I tested:



-I don't have a mini-jack to 2x jacks adapter so I tested each channel individually with a mini-jack to stereo jack cable. In other words only ONE channel receives signal.
When the switch is on the channel that is receiving signal everything is fine. Sound is good and loud.
BUT, if I switch to the other channel (the one that isn't receiving signal) I can hear spill, leakage, cross talk (or whatever the right name).

So my question is:
-is this due to poor shielding due to the test circumstances?
-does this happen because the "empty" channel, since it's not connected to anything is picking up all kind of crap (the "active" channel being the nearest).

I must point out that there's a HUGE difference in amplitude between signal and spill. I don't think you would hear the leakage with a signal present.
Of course the exact same thing happens when testing the other channel.

Should I worry? Should I test it using both channels connected to see if that is the culprit?
As usual all your ideas and suggestions are very welcome

Cheers
Sono

 

[Hubbub]

-does this happen because the "empty" channel, since it's not connected to anything is picking up all kind of crap (the "active" channel being the nearest).

Good possibility. Put a shorted plug in the second input jack and see what happens. And remember your ears can get really sensitive if the environment is quiet for a while. A better question is how many dB down is the crosstalk? (Test with a steady tone - near clipping at the 3.5mm output jack - from your signal generator.) Also, the budget stereo amplifier implementation might not be as good as we'd like, and it may have some inherent crosstalk.

Shielding is probably not an issue, except for maybe the "A" box, but I'd think that is more for hum rejection than crosstalk rejection.

Good that you posted a pic of the test setup. We can certainly rule out any "long cable" issues.

 

[Voyager10]

Does the crosstalk change depending on what the other output on the B box is doing e.g. its volume setting or which channel it's switched to?

 

[sonolink]

Put a shorted plug in the second input jack and see what happens.

How do you mean shorted? T and R together?

Does the crosstalk change depending on what the other output on the B box is doing e.g. its volume setting or which channel it's switched to?

No. Moving the other channel controls doesn't change anything.


I rechecked and noticed that when listening to the "unconnected" channel (hearing the cross talk), the volume control makes noise when moving it. So I measured voltage at the pots pins and I read 4 volts between pins 2 and 3 when the pot is fully cranked. Both pots read the same.

Could that be the reason for the cross talk?
Should I decouple the amp's output anyway with some cap?
Or maybe I'm just missing some grounding in the A-Box?

Any ideas?

Thanks and Merry Christmas everyone!!
Cheers
Sono

 

[Hubbub]

How do you mean shorted? T and R together?

T and R and S. That connection scheme grounds all inputs so noise pickup is eliminated (assuming ground is noise-free).

I measured voltage at the pots pins and I read 4 volts between pins 2 and 3 when the pot is fully cranked.

If this is 4V DC from the amplifier then there is a problem with the amplifier. A proper bridge amplifier should not have more than a few millivolts of DC between the outputs. Please remind us of what IC is used on the amplifier board. (Also, I'm not sure what pins 2 and 3 are. I would use the terms Input / CW - clockwise, Output / Arm, and Common / CCW - counterclockwise.)

maybe I'm just missing some grounding in the A-Box?

Possibly, but the entire bridge-tied output / RJ-45 scheme needs to be ungrounded except for a shield on the cable.

Here's hoping you have a Merry Christmas as well !

 

[ccaudle]

I measured voltage at the pots pins and I read 4 volts

Where is your final A box and B box schematics? I looked back through the earlier posts, but there was so much back and forth between different ideas I'm not sure where the final design is located in the 168 posts so far.

 

[sonolink]

Please remind us of what IC is used on the amplifier board.

TDA7297

(Also, I'm not sure what pins 2 and 3 are. I would use the terms Input / CW - clockwise, Output / Arm, and Common / CCW - counterclockwise.)

Pin 2 is the middle terminal of the pot that connects to the transformer primary directly (P1)
Pin 3 is the right hand terminal of the pot that connects to switch


I hope this new schem makes things clearer
Thanks!
Cheers
Sono

 

[Hubbub]

From the TDA7297 datasheet the maximum DC offset is 120mV measured at the output with an 8 Ohm load. That offset should not increase much with a lighter load, so 4V DC is a problem. Check the amplifier with out anything connected to the outputs - although your schematic looks basically good.

A note on the input jacks: If you are being fed from a professional audio console the output is on tip and ring, so they should not be shorted together. Use just tip and sleeve for an unbalanced connection, with ring floating.

 

[sonolink]

Check the amplifier with out anything connected to the outputs - although your schematic looks basically good.

I measure at the output of the amp, 4.3vdc between L+ and L-.
0.3vdc between R+ and R-.

If that's a problem coming from the amp I guess I should return it asap. Otherwise would a cap (say 10uF) on L+ and another on L- be a fix?

A note on the input jacks: If you are being fed from a professional audio console the output is on tip and ring, so they should not be shorted together. Use just tip and sleeve for an unbalanced connection, with ring floating.

Your thinking is right since that would be a balanced signal coming out of a proper aux bus, but my friend is going to use 2 headphone outputs on a cheap Zoom podcast console to send both Channels A and B to Box-A.
Those outputs are stereo jacks. So the idea behind shorting T and R is to convert them to mono

Thanks a lot for your time and help
Cheers
Sono

 

[ccaudle]

A note on the input jacks: If you are being fed from a professional audio console the output is on tip and ring, so they should not be shorted together.

And if you are using some kind of headphone output, T and R should really only be shorted together if both channels are playing identical signal, if you are playing a stereo signal it is not very nice to the output driver to short L and R signals together when they are different. It may be OK if there are decent sized build out resistors on the outputs, but if the outputs are low impedance any difference in signal between L and R is going to cause a large current to flow from one output driver into the other.

my friend is going to use 2 headphone outputs on a cheap Zoom podcast console

I thought you previously said 2 balanced aux outputs. That plan is out now?

 

[Voyager10]

Hope everyone had a good Christmas!

4.3v dc between L+ and L- is definitely wrong. Assuming you've connected the power supply correctly the only reasonable explanation is that the amp module is defective.

 

[Hubbub]

If taking a stereo headphone feed to mono I would connect tip to ring via two equal resistors in series, say 10 to 33 Ohms each. Take the sum off the junction of the two resistors, with sleeve as common.

It appears drtechno is thinking you are driving one set of headphones, in actuality you are driving a higher voltage line (bus) that is capable of driving many (impedance shifted) headphones in parallel. Driving that line with a low impedance makes it quite immune to loading, whether the bus sees one headphone load or forty. Yes, you could put an amplifier in each "B" box, but that reduces system reliability, and now requires pumping more power down the smallish wire seen in Cat-5 type cable. Lots of voltage regulation and crosstalk issues would happen. Fortunately the transformers to be used are a lot less costly than $10 each, and a voice type application has no need to keep distortion down to 0.006%!

 

[sonolink]

Hope everyone had a good Christmas!

4.3v dc between L+ and L- is definitely wrong. Assuming you've connected the power supply correctly the only reasonable explanation is that the amp module is defective.

Ok. So I guess I should return it asap.
Otherwise would a cap (say 10uF) on L+ and another on L- be a temporary fix until I get the amp replaced?

Cheers
Sono

 

[sonolink]

Take the sum off the junction of the two resistors, with sleeve as common.

Sorry, I don't understand this last bit...

 

[DarkSky]

Sorry, I don't understand this last bit...

I believe"sum off the junction of the two resistors, with sleeve as common" is meant to describe something along the lines of the following:

 

[Brian Roth]

At first, the POE standard puzzled me until AHA! Both blue pairs in parallel so you don't blow up Grandma's ancient cordless phone if that ancient RJ-11 plug was accidentally inserted into the RJ-45.

Bri

 

[Brian Roth]

I never had to fiddle with POE after I retired from doing any network stuff besides making my Home Office "play nice" with pre-made cables.

100W and 52VDC.....2 Amps down a pair of "doubled up" 24 AWG. Shrugs....seems plausible!

What's next.....running my electric dryer down 24 AWG pairs....JUST KIDDING!

Bri

 

[Brian Roth]

Apologies....but I have NO interest in allowing any remote controls to diddle anything in my hacienda! The wall mounted wall switches that control the Romex cables seem to do what I need. <g>

I recall a few years ago when working a studio project and they were fiddling with "new" internet enabled "light bulbs". A friend knew "whazzup" and began changing the ;lght level and color via the IP connection.

Screw that...I can make my own adjustments locally!

Bri

 

[Brian Roth]

IP enabled light bulbs are just spooky...

Bri