HELP with designing a headphones distributor please

[Hubbub]

While 100V line stuff exists, in the USA 70V is more common. Regardless of line voltage for stereo each "B" box would need four transformers ($5 each in USA), plus a pair of stereo L-pads ($13 each in USA), not to mention box, switches, and connectors. Interesting idea, maybe not the most cost-effective, certainly doable. Much cheaper if stereo is not needed!

Regarding Voyager10's other suggestions, I would say I agree completely, although a star topology may be impractical. From an efficiency viewpoint it would be better to use a 48V feed to minimize supply current, as well as Cat-6 cable, which has lower resistive loss than Cat-5. Each "B" box would then use an isolated step-down supply that allows for using a common high-efficiency class-d (stereo?) headphone amp for each output. In that case the use of higher impedance headphones is not required, so any cheap earbud would work.

 

[MidnightArrakis]

It would help if you drop back to basic electronics. Unless I missed something in the last 60 years impedance matching is critical in only two situations: 1) power transfer must be maximized, 2) cable length is such that reflections can occur. For audio case 1 is almost never a concern, although it might be in specialized situations. Headphone powering is not one of those situations. For case 2 reflections start to be a concern if the cable approaches 1/10 wavelength of the highest frequency. At 20KHz the wavelength in a vacuum (I know... but this is for simplification) is 15000 meters. One tenth of that is 1500 meters, or almost 5000 feet. Are you using the furthest "B" box at about 5000 feet from box "A"? If "NO" then case 2 doesn't apply and forget about impedance matching.

That said, impedance is important. You want a low enough source impedance that you can provide a good signal down the bus for the maximum number of connected loads. You want a high enough load impedance that connecting or disconnecting any one (or number of, you pick) load(s) has negligible effect on the signal level of the bus. For example, let's say the source output impedance was 10 Ohms, the load input impedance was 100K Ohms, and you had a 1V RMS signal from the source with no load. Connect anywhere from 1 to 100 loads and the input never drops below 0.990V, a drop of 0.086dB. That is so little that no one will notice switching on (or off) 100 loads!

As to whether you need a buffer, it all depends on the impedances you are dealing with, and how many loads you expect to have. Design the system right and the correct answer is NO, or maybe YES. It's all in the tradeoffs. You need to understand the basics well before you can intelligently manage the tradeoffs.

[Joined: Apr 18, 2023] -- I "thought" that I had felt something special taking place on that day this year!!! That day (April 18th) was my birthday and with you joining this forum on that day, I just - KNEW - that something had taken place!!! GLAD YOU'RE HERE!!!

[impedance is important] -- As far as what you had commented on within the rest of your posting, that is all for Sonolink to figure out. I'm just here attempting to assist him with determining - HOW - his system will be physically designed and implemented (as shown within my various images). Should you wish to come up with a schematic diagram detailing exactly how - YOU - would design this system.....GO HAVE AT IT!!! As I have tried to make clear.....I am -- NOT -- a "Circuit Designer" type of person. I just take completed schematics and convert them into a "physical reality". A totally different kind of skill set.

/

 

[ccaudle]

"Good Ol' NEUTRIK" has a set of "Heavy-Duty" RJ-45 connectors that have been designed specifically for the "Entertainment Industry"

I like those Neutrik Ethercon connectors, but they are just strain relief shells that go around a standard Ethernet connector+cable.
The chassis side connector can take a standard cable without the cable side shell, so if you can stomach the cost I would always recommend putting the chassis side Neutrik connector in, and then if you want to go cheap just plug in standard cables, if the cables break too frequently then pay the cost of the XLR style shells to add on as strain relief to your cables.

Can someone please tell me if a buffer is really needed

To determine that you need to know the input impedance of the headphone amplifier circuit and the minimum and maximum number of B boxes which might be attached.
Once you know that then determine what the impedance will be when both outputs of all devices are switched to the same input signal. Can the A box drive that minimum impedance? Are you OK with the associated level change and distortion change when going from 1 output of 1 B box on an input channel to all outputs connected to the same channel?

I may have missed some information just skimming over the weekend, but I don't think you have provided information for either of the two necessary data points to make a decision.
If you don't know and want to keep your options open, then the conservative choice is to put high impedance buffers on each input before the switches then it doesn't really matter.

 

[Voyager10]

So the "100V line" suggestion wasn't entirely serious, but it got me thinking.

The humble Eagle LT700 output transformer (e.g. LT700 Audio transformer) is a couple of dollars or so, and steps down from 1.2K primary to 3.2Ohm secondary. So it shouldn't have any trouble with 16 or 32 ohm headphones, and ought to be good for 100mW or so.

If you don't need super audio quality, you could drive a hundred of these from a medium power audio amplifier, with no other power needed and really easy construction. (Two per box, placed _after_ the A/B switch).

 

[sonolink]

Ok guys, thanks a bunch for all the comments. I apologize for not giving a description of the project and its application. TBH I am not (as you are already probably aware) an EE. I'm an old self taught electronics hobbyist. I've been building guitar tube amps and pedals for over a decade and I've tried to learn as much as I could (I still am eager to learn). My real trade is sound engineering.

So I will try to describe as clearly as possible the goal for this project: A good friend of mine (also a sound engineer) is married to a girl that works as a freelance simultaneous translator. He asked me if I could build a setup for her to use in some conferences they do.

So, what he does so far is to mic a guy giving a speech in one language in a lecture hall, and at the same time mic his wife in a cabin translating the guy in another language. Both mics get sent to 2 different outputs of a mixer.

What he needs is a box in which 2 people can plug a pair of headphones (their own set) on minijacks, set the volume they desire and switch between languages at will. Boxes would be max 3 meters away from eachother. My friend suggested daisy chaining the boxes using RJ-45 since it's cheap and easy to deploy and pickup after the session. He said he's need like 20 boxes (40 people/headphone pairs).

Trying to keep costs down I thought it could be a good idea to try to find cheap ready made circuits (also space wise since they use SMD components that save space) and mount them inside a Hammond enclosure with a couple of pots and minijacks. Hence my search of those Headphone amps in Aliexpress.

To determine that you need to know the input impedance of the headphone amplifier circuit and the minimum and maximum number of B boxes which might be attached.
Once you know that then determine what the impedance will be when both outputs of all devices are switched to the same input signal. Can the A box drive that minimum impedance? Are you OK with the associated level change and distortion change when going from 1 output of 1 B box on an input channel to all outputs connected to the same channel?

I have no idea what the input impedance of the headphone amp circuit would be. If I buy ready made circuits maybe I could ask that information to the seller...
Minimum number of B boxes would probably be 10 and max 20.

 

[MidnightArrakis]

[they are just strain relief shells that go around a standard Ethernet connector+cable] -- I am well aware of that. But, somewhat as you have also mentioned, I have suggested using them in order to hopefully maintain some structural integrity of the RJ-45 cables during their use and lifetime. Should Sonolink decide - NOT - to use them.....that's his choice.

[1) need to know the input impedance of the headphone amplifier circuit and the minimum and maximum number of B boxes which might be attached / 2) the conservative choice is to put high impedance buffers on each input before the switches] -- Since the number of "B" boxes will be an unknown variable, that is what my earlier diagram was attempting to show - without - actually "saying" it!!! Using a set of buffers as I have indicated just makes sense to me.

[2 people can plug (in) a pair of headphones (their own set)] -- THAT is a -- HUGE -- system variable RIGHT THERE!!!

[on minijacks] -- You're talking about 3.5mm jacks and - NOT - ( standard ) 1/4" headphone jacks? I am guessing with today's use of earbuds and whatnot that nobody uses 1/4" headphone plugs anymore, huh???

[Boxes would be max 3 meters away from each other] -- While the boxes may only be 3-meters (i.e., ~10-feet) away from one another, the actual cable length will probably be closer to 5-meters (i.e., ~15-feet), assuming that the boxes are placed onto a desk or table. This is because the cables will be going into/out-of the boxes, then down onto the floor, over to the next station and back-up to the next box. Therefore, you also need to include the distance from the desk height to the floor.....times 2!!! Makes sense???

[I have no idea what the input impedance of the headphone amp circuit would be] -- Which is yet another reason of why it "makes more sense" (at least to.....me!!!) to just use a known "headphone amplifier circuit" and incorporate it into a PCB.

[If I buy ready made circuits] -- While you may believe and/or "think" that buying "ready-made" circuits is the cheap and quick way to go for this project, what you are opening yourself up to is an -- ENTIRE ASSEMBLY NIGHTMARE -- completely fraught with inherent possible errors and untold problems!!! What my earlier images have shown using a single PCB is something that even - YOU - could easily put together >> OR!!! <<.....if the PCB-fabrication shop also has a "PCB-assembly" department, these PCB's could be very easily assembled either by hand or by using automated equipment. I could easily get this done here in the U.S., but I don't know what the general capabilities are over there in the EU.

Having a quantity of 25 PCB's fabricated and assembled would qualify as a "low-volume" production or a "short-run" assembly. It is just enough to possibly even qualify for some kind of a discount. Who knows???

[maybe I could ask that information to the seller] -- yeah.....you can ask them for that technical information, but.....I would guess -- FAT CHANCE -- of you receiving an answer to your question. Those types of sellers either just don't know that type of information or they don't - GIVE A RATS ASS - of providing you with an answer!!! That's just how it is. In my humble opinion, you would be a whole lot better off by just having a complete circuit designed (Ccaudle???) of what it is that you need and then having a PCB designed and assembled to fit an enclosure. That way.....you are relatively guaranteed that this whole entire system will work "right out of the box"!!!

Another idea that came to me as I was futzing around with some stuff relating to your project is...............maybe your buddy's wife might know of and/or could inquire about within her translating world.....she could find out if such a system as you are designing here could be useful to some other organizations for the same and/or similar purpose? If so, now that you would have done all of the R&D work to get such a system completely built and functioning, you could have one or more additional systems built and you could SELL THEM to these other organizations!!! And.....why would you want to do that? SO YOU AND YOUR BUDDY CAN RECOUP YOUR ORIGINAL COSTS AND MAYBE EVEN EARN A SMALL PROFIT FOR ALL OF YOUR EFFORT!!! How does THAT sound to you???

[B boxes would probably be 10 and max 20] -- And.....I would still go for a quantity of 25 boxes because:

1) You never know who else just might show up, you know???

2) Spares.....just in case someone gets pissed-off and stomps on their box or hurls it across the room!!!

3) Possible quantity discount. Price breaks when purchasing items usually starts at a quantity of 25 and then multiples of 25.

Just sayin'.....

/

 

[sonolink]

[2 people can plug (in) a pair of headphones (their own set)] -- THAT is a -- HUGE -- system variable RIGHT THERE!!!

I know.....it's just the way things go nowadays I guess...

[on minijacks] -- You're talking about 3.5mm jacks and - NOT - ( standard ) 1/4" headphone jacks? I am guessing with today's use of earbuds and whatnot that nobody uses 1/4" headphone plugs anymore, huh???

Correct

[Boxes would be max 3 meters away from each other] -- While the boxes may only be 3-meters (i.e., ~10-feet) away from one another, the actual cable length will probably be closer to 5-meters (i.e., ~15-feet), assuming that the boxes are placed onto a desk or table. This is because the cables will be going into/out-of the boxes, then down onto the floor, over to the next station and back-up to the next box. Therefore, you also need to include the distance from the desk height to the floor.....times 2!!! Makes sense???

No, no, 3m includes the extra cable lenght needed

Another idea that came to me as I was futzing around with some stuff relating to your project is...............maybe your buddy's wife might know of and/or could inquire about within her translating world.....she could find out if such a system as you are designing here could be useful to some other organizations for the same and/or similar purpose? If so, now that you would have done all of the R&D work to get such a system completely built and functioning, you could have one or more additional systems built and you could SELL THEM to these other organizations!!! And.....why would you want to do that? SO YOU AND YOUR BUDDY CAN RECOUP YOUR ORIGINAL COSTS AND MAYBE EVEN EARN A SMALL PROFIT FOR ALL OF YOUR EFFORT!!! How does THAT sound to you???

That's not a bad idea at all. I'll try to get some info on that. Thanks for the tip

[B boxes would probably be 10 and max 20] -- And.....I would still go for a quantity of 25 boxes because:

1) You never know who else just might show up, you know???

2) Spares.....just in case someone gets pissed-off and stomps on their box or hurls it across the room!!!

3) Possible quantity discount. Price breaks when purchasing items usually starts at a quantity of 25 and then multiples of 25.

You're right, I'll mention that to my friend. It's wiser to have some spares. Thanks again for that tip too

 

[MidnightArrakis]

[3m includes the extra cable lenght needed] -- GREAT!!! It's good that you had already thought of that. NOW!!!.....Start making a BOM/Parts List and start the list off with:

QTY: 25 -- 3-meter Heavy-Duty RJ-45 CAT5 (or, CAT6) cables.

(WATCH THIS SPACE FOR FURTHER DEVELOPMENTS)!!!

/

 

[sonolink]

The PAM8043 board has 3W output on each side. I realise that 3W is waaay more power than you need for headphones. So it would be best to use this amp below its full power rating.

I found the following to achieve that. Can somebody please confirm that would work?

"PAM8403 datasheet doesn't say what's actually inside it, but it's a reasonably safe bet that it's a full bridge for each channel. As Out_L+ is driven positive, OUT_L- is driven negative, and so on, giving you twice the voltage and 4x the power that a half-bridge will do. (With no input signal, both outputs should sit at half the supply rail). So I would simply ignore OUT_L- and use OUT_L+ as a half-bridge (a normal push-pull amplifier). You will need to AC couple the output via 10uF (maybe 47uF would extend the bass a little) but then you can simply return the common GND from your headphones to GND."



Impressions?
Thanks!
Cheers
Sono

 

[Hubbub]

Getting back to the use of transformers in each "B" box could result in a serious simplification and allow for full A and B sources in stereo on the RJ-45 connector. Assume each headphone jack is fed directly from two each 1300:8 transformer. Running at headphone level they can be small "EE14" and still sound good. These transformers cost about $1.10 each from China, https://www.aliexpress.us/item/2251832856173900.html

For a volume control consider using a 2K Ohm dual pot ($0.89 https://www.aliexpress.us/item/3256804359484078.html) with the A-B switch feeding the input and common, the transformer primary connected to the arm and common. The inputs to the switch are the bus signals. Add box, knobs, connectors and PCB and you are done. FWIW I like a small extruded box with flat plate (metal or PCB material) front and rear panels, that way one main circuit board holds everything - RJ-45s to rear, all else to front - and the box is nearly indestructible, with the PCB having almost no contents to shake loose.

How does this design work? Assume 20 "B" boxes all set to the same feed, 40 loads in total. Headphones are more like 32 ohms, the transformer makes these look like 5200 Ohms, so the 2K pot works fine for the volume control. With 40 of those in parallel at full volume the impedance for one channel is still 20-36+ Ohms, with those 40 loads wanting maybe 0.1W each, or 4 W total. So feed each of the 4 channels on the bus with an amplifier good for maybe 10W-20W into 8 Ohms and you are probably good to go. Any loads switching on or off will not affect the output of a decent amplifier, so no perceptible interaction between boxes/users. I would avoid any bridge-tied load type amplifier, as they truly require a floating load and that would mean using 4PDT switches for the A-B selection. Add a power supply for the 4 amplifiers and the "A" box is done.

And I'd use Cat-6 cable - bigger wire for less loss and keeping the load interaction to an imperceptible level. Might even make sense for the "A" box to have dual outputs, so it can sit in the middle of the system, with 10 "B" boxes on each output.

Something to think about...

 

[sonolink]

Getting back to the use of transformers in each "B" box could result in a serious simplification and allow for full A and B sources in stereo on the RJ-45 connector. Assume each headphone jack is fed directly from two each 1300:8 transformer. Running at headphone level they can be small "EE14" and still sound good. These transformers cost about $1.10 each from China, https://www.aliexpress.us/item/2251832856173900.html

For a volume control consider using a 2K Ohm dual pot ($0.89 https://www.aliexpress.us/item/3256804359484078.html) with the A-B switch feeding the input and common, the transformer primary connected to the arm and common. The inputs to the switch are the bus signals. Add box, knobs, connectors and PCB and you are done. FWIW I like a small extruded box with flat plate (metal or PCB material) front and rear panels, that way one main circuit board holds everything - RJ-45s to rear, all else to front - and the box is nearly indestructible, with the PCB having almost no contents to shake loose.

How does this design work? Assume 20 "B" boxes all set to the same feed, 40 loads in total. Headphones are more like 32 ohms, the transformer makes these look like 5200 Ohms, so the 2K pot works fine for the volume control. With 40 of those in parallel at full volume the impedance for one channel is still 20-36+ Ohms, with those 40 loads wanting maybe 0.1W each, or 4 W total. So feed each of the 4 channels on the bus with an amplifier good for maybe 10W-20W into 8 Ohms and you are probably good to go. Any loads switching on or off will not affect the output of a decent amplifier, so no perceptible interaction between boxes/users. I would avoid any bridge-tied load type amplifier, as they truly require a floating load and that would mean using 4PDT switches for the A-B selection. Add a power supply for the 4 amplifiers and the "A" box is done.

And I'd use Cat-6 cable - bigger wire for less loss and keeping the load interaction to an imperceptible level. Might even make sense for the "A" box to have dual outputs, so it can sit in the middle of the system, with 10 "B" boxes on each output.

Something to think about...

Thanks for the tip Hubbub,
I'll try to draw a diagram/schem in the following days and report back

Cheers
Sono

 

[sonolink]

So feed each of the 4 channels on the bus with an amplifier good for maybe 10W-20W into 8 Ohms and you are probably good to go.

Sorry, but if I understand you correctly, what you're saying is to use a 10W-20W amp into 8 Ohms to feed the entire bus and from there use transformers to "convert" the signal to "headphones level"?

So the signal chain would be:
mics -> mixer -> mixer aux sends -> 10w-20w amp into 8Ohms -> Boxes Bus -> transformer -> volume pot -> headphones

Is that right?

 

[MidnightArrakis]

Sorry, but if I understand you correctly, what you're saying is to use a 10W-20W amp into 8 Ohms to feed the entire bus and from there use transformers to "convert" the signal to "headphones level"?

So the signal chain would be:
mics -> mixer -> mixer aux sends -> 10w-20w amp into 8Ohms -> Boxes Bus -> transformer -> volume pot -> headphones

Is that right?

These transformers are the exact same thing, but are less expensive from another supplier:

https://www.aliexpress.us/item/2251832839593934.html

Buying -- 100 -- of these transformers (25 boxes X 4-transformers per box) will set you back a staggering USD $31.00 plus shipping.

Using these transformers also completely changes -- EVERYTHING -- I have just completed!!! Oh, well.....

/

 

[sonolink]

These transformers are the exact same thing, but are less expensive from another supplier:

https://www.aliexpress.us/item/2251832839593934.html

Buying -- 100 -- of these transformers (25 boxes X 4-transformers per box) will set you back a staggering USD $31.00 plus shipping.

Using these transformers also completely changes -- EVERYTHING -- I have just completed!!! Oh, well.....

/

Thanks for the link Jerry, but they ask me more than 100$ shipping for that one!!

Using these transformers also completely changes -- EVERYTHING -- I have just completed!!! Oh, well.....

Absolutely, but I find it's a brilliant and elegant idea (if I have understood Hubbub's idea correctly, of course)

Cheers
Sono

 

[MidnightArrakis]

Thanks for the link Jerry, but they ask me more than 100$ shipping for that one!!



Absolutely, but I find it's a brilliant and elegant idea (if I have understood Hubbub's idea correctly, of course)

Cheers
Sono

[they ask me more than 100$ shipping for that one!!] -- Without my taking the time to check a map or a world globe.....but, I cannot imagine that you are any further -- WEST -- of China than I am being -- EAST -- of China. On the AliExpress website, it states that for buying a lot of 5-transformers, the shipping charge is:



So.....you might want to contact the seller directly and message them. Tell them you are interested in purchasing 100 - E14-222A transformers and request a quote for the shipping to Madrid, Spain and/or London, England.

Also.....ask them for the complete and total shipping weight for all 100 transformers. Whenever it comes time for you to purchase these transformers.....you can let me know this.....and, if you can let me know what the total shipping weight is, I can then contact one of the "logistics" companies that I am aware of who will bid and quote you a more "sane" shipping price. I have purchased huge power toroid transformers from China and these "logistics" companies are all too eager to provide a competitive shipping price.

/

 

[Voyager10]

At the risk of over-simplifying, do you even need stereo, if it's just a speech feed?

If mono will do, it's now totally simple:



You just need a DPDT switch and a single-gang pot, and drive both headphone channels from one transformer.

(A 4k7 pot will probably do just fine. With a pot driven from a low-impedance source, the maximum output impedance is one-quarter the total impedance, which occurs at 50% track resistance).

 

[sonolink]

At the risk of over-simplifying, do you even need stereo, if it's just a speech feed?

If mono will do, it's now totally simple:

View attachment 117570

You just need a DPDT switch and a single-gang pot, and drive both headphone channels from one transformer.

(A 4k7 pot will probably do just fine. With a pot driven from a low-impedance source, the maximum output impedance is one-quarter the total impedance, which occurs at 50% track resistance).

Actually you also have a good point there!
Thanks for the brilliant suggestion

 

[Hubbub]

Sorry, but if I understand you correctly, what you're saying is to use a 10W-20W amp into 8 Ohms to feed the entire bus and from there use transformers to "convert" the signal to "headphones level"?

So the signal chain would be:
mics -> mixer -> mixer aux sends -> 10w-20w amp into 8Ohms -> Boxes Bus -> transformer -> volume pot -> headphones

Is that right?

Basically correct, except the volume pot goes before the transformer as Voyager10 shows.

 

[Hubbub]

At the risk of over-simplifying, do you even need stereo, if it's just a speech feed?

If mono will do, it's now totally simple:

View attachment 117570

You just need a DPDT switch and a single-gang pot, and drive both headphone channels from one transformer.

(A 4k7 pot will probably do just fine. With a pot driven from a low-impedance source, the maximum output impedance is one-quarter the total impedance, which occurs at 50% track resistance).

Your solution is 99% correct. My only quibble is 4k7 is a fixed resistor value, which is not common for potentiometers unless you pay for custom tooling and buy maybe 10,000 or more. Fortunately a value of 5k is common and readily available.

Even a linear taper pot would work decently as loading it reasonably heavy bows the transfer curve down approximating a "linear in dB" type log-taper pot. FWIW I've done this, plotted out via spreadsheet for proof, just recently for an oddball DI box I'm building.

 

[MidnightArrakis]

[10w-20w amp into 8 Ohms -> 1 Box Bus -> volume pot -> transformer -> headphones] -- Equals a 5K load.

[10w-20w amp into 8 Ohms -> 25 Box Bus -> volume pot -> transformer -> headphones] -- Equals a 200 Ohm load.

>> Does this load difference make any difference to any kind of apparent signal-level shift?

[Even a linear taper pot would work decently as loading it reasonably heavy bows the transfer curve down approximating a "linear in dB" type log-taper pot] -- I interpret this as meaning the -- taper -- of all of the pots will vary according to the number of boxes plugged into the system. Does this matter?

[If mono will do, it's now totally simple] -- Should a mono signal and system be used, would there be any benefit to paralleling the RJ-45 conductors to become "2-sets of 4-conductors"? Wouldn't doing so lower the total wiring loss? Just wondering.....

/