summing speaker/power signals, help !

[isophase]

Hi all,
i did search on the forum, but no success.
if i want to drive a headphone set from a stereo power amp of lets say 100 Watt per channel, and insert a mono switch in beween ? is it better to add a summing netwotk (with high wattage resistors) or just combine the signals hardwire?

Ok, let me explain why.
we are fed up at the studio of having to constantly install/move 3 different heaphone amps around the (rather small) studio space, and would like to try out a passive setup using power amps instead (we got spare/unused amplifiers and 8 speaker tie lines at the studio that could do the job).
thinking of using a guitar box type eclosure fixed on a round base mic stand, the box would have speakon (2+2) chassis connector underneath , a stereo attenuator, a switch for stereo to Mono and a jack plug. simple and clear.

it means no more sh*tty headphone power supplies with their tiny sh*tty cables, and no more heaphone amps laying on the floor eating up space, just one cable.

so to add a mono switch, would it be best to combine the signals directly at the switch or use resistors, i'm sure both would work, just wondering which is the better solution.
thanks!
jon

 

[PRR]

.

 

[audiomixer]

two notes: you could probably skip the two 1R resistors, they are paralleled at the amp return terminal anyway and will not do anything (except 'protecting' the amp if it is a bridged amp).

do not use a bridged amplifier, make sure you have a common ground for LS return.

- Michael

 

[ruffrecords]

Replace the two 33 ohm resistors with a two pole 6 position switch pot to give the artists their own level control.

Cheers

Ian

 

[isophase]

thank you!
PRR, i'm trying to understand the different resistor values.
are these values chosen to arrive at a certain load (about 60ohm?) per channel in order to match the headphone impedance?
wouldn't it be better to choose a load that is about half the impedance of the headphone in order to not have too much change in level when using diffenret type/impedance headphones (not really important since there would be an attenuator at the end).
thinking of using an L-Pad/constant impedance speaker attenuattor for level adjustment. is this a good idea? or is should i just use a 3watt (or 5watt but hard to find) wirewound stereo pot?
i build a similar type box several years ago (without a mono switch) using linear wirewound pot for the same purpose, it was working fine exept the sensytivity of the pot was not optimum (change in level was happening only in the first quarter of the turn) i never found wirewound Log pots...
i know this is very basic stuff, but i prefer to ask and do it the right way as i'm not so good with the theory as you can see.
thanks a lot!

 

[isophase]

Forgot to mention,
the mono switch would be an option to permit making two discrete Mono headphone monitor circuit if needed, and still hear the music in the center (spread the one channel to both earphones)
(EDIT:with two separate boxes and a stereo amp)
thanks Ian for suggesting the rotary switch for attenuation, a very good and less expensive solution for sure, but i bet at some point one musician is going to say that the sound is eather too high or too low, i would take a bet on that  

 

[ruffrecords]

I think PRR's resistor values are a good compromise. His network has an effective source impedance of about 25 ohms. If you use 32 ohm phones then the level will drop but that's fine as these need less voltage. If you plug in a higher impedance phone you get less signal drop but higher impedance phones generally need more voltage anyway so the sound level won't alter that much.

Looked at another way, 100W into 8 ohms is about 28V rms. With 32 ohm headphones, PRR's network attenuates this by just over 7 times to about 4V rms. That is more than enough to drive any 32 ohm phone made today to painful levels.

Regarding resistor wattage, , even if you short the output, the worst case is 28V rms across 100 ohms which is 8 watts. Most of the time you are going to be nowhere near this, especially as its music we are talking about not sine waves. 5 watt resistors should be fine.

I built a box like this about 40 years ago. I used a Sinclair Z30 amp (30 watts??) to feed a box with 4 headphone sockets in it each fed via 120 ohm resistors. I only used quarter watt resistor types and they never blew up.

Cheers

Ian

 

[landis]

This was a pretty common set up back in the days before hear back type systems came along. If I'm not mistaken. I've only seen it implemented once and I can't remember the name on the boxes. But they had xlr in and out on the bottom stereo/mono switch and separate level pots for left and right. They were all attached to music stands and I've seen orchestral sessions run all the head phones off of one line. Never got to se e the inside of the boxes or the amps used though.

 

[PRR]

> not have too much change in level when using diffenret type/impedance headphones

As a trend: the higher-Z phones have better power sensitivity and lower wattage ratings.

Somewhere else (now vanished) I surveyed many phones and noticed that 7V behind 29 ohms (not critical) would drive "any" phone (excluding several exotics) to HIGH but non-burning level. The small coil in the typical hi-Z phone simply took less power.

The dots are impedance/power numbers from Dennis "Rane" Bohn's 1983 survey of headphones. While the data is old, my sense is that model-numbers change, gut technology doesn't.

For 25 Watts change 100+33 to 68+68.

Yes, the 1 ohms are to smoke if somebody tries this with a bridge-mode amplifier. They exist, and often look the same as a common return amp.

Individual level controls *probably* get into problems which bring you back to individual active (powered) amplifiers. With ~~100 ohms in you can plug a dozen of these to one amp. Speaker attenuators tend to be 8 ohm, so you can only use 1 or 2 per amp. A 25 or 50 ohm couple-Watt dual-pot would work in place of the 33 ohms.... but where do you find those? If you do, it's much more precious than a simple dual-100K and a chip.

> 40 years ago ... 30 watts?? ... via 120 ohm

Me too.... and I used to find dead headphones when people switched to speakers and didn't notice the headphones squawking. The plot above shows 32W 100r is a decent fit to many 32 ohm phones, but potentially over-powering to 60r and up. (We had some expensive 150 ohm cans.)

 

[Brian Roth]

When I began working in the (pro) studio biz back in the early 1970's, the scheme for cue feeds was using something like a Crown D-60 (and often a D-150) power amp fed from the cue outs of the desk.  Since the Crowns had both "black" output posts commoned together to ground, shielded 1 pair wire was used to move the signal out to the tracking space(s).  Out there, the cue feed terminated onto male XLR-3 jacks on the wall plates, to prevent mic/cue mis-connections.


The cue boxes had two A-B Type J, 2 Watt carbon pots...one for each ear.  IIRC, the pots were 500 Ohms wired as a potentiometer (not as a rheostat), and definitely linear taper.  The headphone load impedance "slugged" the wiper of the pot, providing a quasi audio taper.

The next "advancement" added a 3 position switch on the cue boxes.  It would select stereo from the amp, to the pots and earpieces.  Another position would drive the pots only from the left channel of the amp, while the third switch position would feed the amp's right channel to both pots/earpieces.  Thus, the latter two switch positions would allow two independent mono cue mixes vs. a single stereo cue feed.

The independent left/right level pots were used because of easy availability, AND had the bonus of allowing a performer  to reduce the level in one ear, vs. having to slide an earpiece off his ear so he could hear himself or something else in the tracking space.

20+ years ago, I took the concept a step further at the request of the owner/engineer of a new studio that I was designing/constructing.  We used two stereo amps, and (IIRC) 4-conductor cable with overall shield, and XLR-5 connectors.  On the cue box, I added a six position switch:

1.  Stereo from amp #1.
2.  Stereo from amp #2.
3.  Left from amp #1 to both pots/ears.
4.  Right from amp #1 to both ears.
5.  Left from amp #2 to both ears.
6.  Right from amp #2 to both ears.

Thus, various combinations ranging from two stereo mixes through 4 mono mixes were available.

Passive cue systems are still being used, with off-the-shelf boxes such as:

http://www.adorama.com/FUHR2.html

http://www.rolls.com/product.php?pid=PM52

They are variations of the same 'classic' scheme.  They don't have the switching features I described above, and use stereo pots.  I still think the 2 Watt type J pots are more robust.

Typically, an amp rated around 100 W into 8 ohms works well with any of the passive systems.

EDIT.  My friend Rick Chinn wrote this articles years ago, which discusses powering headphones:

http://www.uneeda-audio.com/phones.htm

Bri

 

[joaquins]

It always have been hard compromise this kind of things, in the market you have a sony V700 with 24Ω and 3W max power and some AKG (121, 240, etc) with 55Ω and 200mW...

For match those max power for both you would need to feed them with 2,2V and -18Ω resistor (yes! negative) With this (not imposible but nuts) you will get about 1W for 30Ω but forget about high Z ones or less sensitive like my HD380 (500mW 54Ω)

I like PRR compromise, but will be putting a high 400mW on those 200mW AKG and only 400mW on my 500mW Sennheiser. (both almost the same Z and very different power rating)

In the other hand... how does damping factor and output Z of the feeding device affect the sound of headphones? I mean, for example, they have inductance and feeding them with a resistive source will make a HPF... I've never read something abot this nor experiment but I think this every time I see some of this amps...

Returnig to the topic... I think the schem PRR sent on saturday doesn't do what you are asking for... you may want a selector that switch between one or the other input to both outputs in parallel or let them stereo...

JS

 

[ruffrecords]

> 40 years ago ... 30 watts?? ... via 120 ohm

Me too.... and I used to find dead headphones when people switched to speakers and didn't notice the headphones squawking. The plot above shows 32W 100r is a decent fit to many 32 ohm phones, but potentially over-powering to 60r and up. (We had some expensive 150 ohm cans.)

My overall system was a tad more sophisticated. Recordings were mono so I had a playback speaker in the recording area. On top of it was a relay operated cue light box. The relay cut the speaker and connected it to the aforesaid phones box. never ones blew up any phones.

Cheers

Ian

 

[PRR]

> only 400mW on my 500mW Sennheiser.

It is not necessary to hit the power spec; more important, is it loud enough?

Convert the claimed sensitivity to 110dB SPL, 120dB SPL, whatever your clients think is loud enough.

The "3W" phone may be loud enough with 0.4W. (Or it may not be--- don't use it, there are other choices.)

I know there are several phones which need huge power to get loud. Fortunately most are too expensive to kick around a studio.

There are the in-ear phones derived from hearing-aid technology with VERY high sensitivity. These should not be mixed with common ear-things.

Damping: AFAICT, the WalkMan/iPod phones hardly care. 30 ohm source causes less error than the tolerance from one to another of the same model, much less different models and brands. The hi-Z AKG does have response peaks which are better with low-Z source, but on my 150 ohm version 30 ohms was low enough. Speaking audio-phile, it is not essential that the 30 ohms also be the source impedance; it may be incorporated inside the NFB loop or in the power feed. Speaking about tracking monitors, get the beat and the pitch to the players, don't worry about few-dB variations in response, the engineer will EQ on the good speakers.

> doesn't do what you are asking for... you may want a selector

Yes, there are infinite variations in what musicians may need in the cans. Hence Brian's 4-ch 6-way switch, and boxes which take 8 buses and allow *each* musician to blend a *custom* mix. But Iso's studio is small, this isn't CBS nor Western Recorders. To some degree the musicians must take what they get or book elsewhere. Stereo is useful. Stereo-to-Mono is useful when a player prefers one ear out in the open, and just a SPST toggle or added jack. Brian's plan goes from 4 to 5 wires and a 6-way switch-- you get more choices but more to rig and buy.

Pump the whole console to the Cloud and let the players connect via iPhone. Should be mildly easy to do a server-side app which proportions 64 channels for each of a dozen users and WiFi/4Gs it back.

 

[joaquins]

> only 400mW on my 500mW Sennheiser.

It is not necessary to hit the power spec; more important, is it loud enough?

Convert the claimed sensitivity to 110dB SPL, 120dB SPL, whatever your clients think is loud enough.
...

I know is not necessary to get the whole power, just an example... What I wanted to show is that any amp you build is going to overdrive some headphones and under drive some others, I didn't make the math for SPL but I will and post it.
I've used V700 a couple of times but can't tell how was the amp, I don't remember which I used.

About the damping, just a question i have, when I get near a dummy head maybe I'll test it.

JS

 

[audiomixer]

Forgot to mention,
the mono switch would be an option to permit making two discrete Mono headphone monitor circuit if needed, and still hear the music in the center (spread the one channel to both earphones)
(EDIT:with two separate boxes and a stereo amp)
thanks Ian for suggesting the rotary switch for attenuation, a very good and less expensive solution for sure, but i bet at some point one musician is going to say that the sound is eather too high or too low, i would take a bet on that

your mono switch needs three or maybe even four positions then: L only, R only, L-R, R-L....

I have had an headphone  application were there are 200+ phones, six different lines and volume control - all run on a 70V system with a transformer for each output. old school tech but works reasonably well, sure better then 200 headphone amps around a room!

- michael

 

[sodderboy]

The mono switch can be done with a DPDT ON-ON-ON style switch.  That is how the Simon Systems CB4, the high-bar of commercial distributed cue boxes does it.  The ON-ON-ON switch is both poles connected to COM on either side, and one pole from each side connected to it's COM in the middle position.
The box itself is a tank and I have never had one blow, even in use at recording schools, where gain structure is in one ear and out the other. . .
The simple parts solution is to emulate a ProCo cue box which uses separate switches for mono/stereo for each side, and single pots for L and R volume.
I can dig and describe the simple schematics if needed.
Mike

 

[Brian Roth]

Sodderboy, I've run across some of the Simon Systems boxes at a few studios, but never took the time to study them closely.  Externally, they look like very sturdy boxes.

The passive cue boxes I recall building 35+ years ago (EGADS...LOL) were a hybrid between the Simon and ProCo examples that you linked.  A single full-size bat handle, 3-position ON-ON-ON 'DPDT' toggle switch plus individual pots for each ear.

I had to spend a few mins. "doodling" the schemo from memory just now, but indeed it's a simple schemo.

We used A-B Type J pots, which are REALLY sturdy.  Currently, the PEI pots from Canada are the closest available.  Digikey's website kinda annoys me, but it appears that two mono pots are more readily available AND less expensive that a single dual gang pot.

"Back Then", the cue boxes sat of the floor, so STURDY construction was a must.  The Simon and ProCo have "lips" on the sides and/or top-bottom to help reduce damages when the box itself is stepped on by accident.

I did the same thing when I designed/built up the dual power amp units discussed earlier in this thread.

BUT, the "one pot per ear" design confuses inexperienced kids who are used to one pot for overall level control.  Shrug.

Later systems are now a distro of line level L/R/whatever into multiple little (wall wart) powered boxes, and that is a NIGHTMARE with all of the little cue amps/warts on the studio floor.

Which brings me to what should be in another thread...today's cue systems where every player has his own mixer panel.  Furman, Aviom, HearBack, etc.  In the hands of experienced session musicians, they are GREAT.  Otherwise, a headache for the session engineer.

Bri

 

[isophase]

thanks to all for your help and contribution!
I spoke to the guys at the studio and they are not sure yet if they want to take this route or invest in a dedicated system, we'll see... Personnaly, i think its a very good idea since the power amps are already available, and it would make a much easier/cleaner setup in the studio.
i added the words headphone cue box to the tittle of this thread to make more sens.
Attached is the last plan with updates highly inspired with ideas from the above posts
the box now has a 3x4 rotary switch in order to be able to select eather:
1- mono mix/Cue1 (in both ears)
2- stereo (with dedicated mono switch)
3- mono mix/Cue2 (in both ears)
A stereo 8ohm speaker attenuator, and added a second headphone jack terminated with resistors in order to keep a fictive headphone load impedance in case only one headphone is plugged in the box (and i just realise the correct value is more 66ohms rather than 33ohms since the jacks are in parallele.. i'm sure there are other mistakes! you tell me)
oh, and the speakon patchbay with the power amp split output i know would be a bit "dangerous" if used inproperly, it lets you the option of making 4 independent mono headphone mixes (which i think is insane! but my friends at the studio say they want the possibity of doing it) or 2 stereo mix, or 2 mono + 1 stereo...
in the end i think its getting a bit too complicated/compromising if such a system has too much options (like this one) but i'm not the only person working in this studio and i certainly don't want to impose anything seen as a limitation. we'll see...

 

[isophase]

it would be nicer to have only one rotary selector and skip the extra mono switch to make the box more simple and less confusing for the musicians, but how to do it? what switch to use?

 

[sodderboy]

The big simplification is to lose the two - lines.  I never jumped the two -'s together at the amp and simply connected one side only.  You only need one and run the distributed cue with 12-3 or 10-3, and component count reduces by 25%.  Plus, you do not need to switch the amp - .  You only need to switch the +'s, so you can do it with a ON-ON-ON DPDT toggle.  Draw the switch and it should make sense to you: one center COM goes to the L and one to the R volume pots.  The LHS of the switch poles both connect to the box R input, and the RHS poles both connect to L.  When the switch is thrown LEFT the box L input is fed to both pots, when thrown RIGHT the R input feeds the pots, and when in the center one pole from each side, L and R, feeds the pots.  Get stubby toggles because they do not flip or break easily.

Personally I think distributed CUE is robust and sounds better than the CAT5 mixers but very inflexible when recording bands playing together.  Even when going MONO on all sides you get only 4 cue mixes, and the engineer has to do the work getting everyone what they want/need.  With CAT5 cue mixers everyone gets what they want themselves.  It usually sounds like cacophony but it works for them.  Look at your competition studios and see what they have- from a sales perspective a cue mixer system will always sound better to a potential client than distributed, no matter how geeky and custom the cue boxes are.  If you are doing post or recording large ensembles or orchestras, then distributed is the best way to go.  You have to look at your target clients.

Mike