ProSoundWeb Community

Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => LAB: The Classic Live Audio Board => Topic started by: Tom Bourke on January 04, 2016, 01:04:48 am

Title: Ying outputs, was I really that wrong?
Post by: Tom Bourke on January 04, 2016, 01:04:48 am
Had a co-worker go off on me over a setup for the new years festivities.  The venue was a large Las Vegas strip hotel with multiple ball rooms.  One of the smaller ballrooms was a "pre-function" area with bars and themed photo areas.  Simple audio needs, premixed "DJ" tracks off a laptop, main ball room feed and maybe an announce mic.  Outputs were L,R, Fill, Sub, Alternate room.  PA was large A level powered speakers with internal DSP that could take full range LR+Sub.

For a mixer I had a Mackie 1202 or something much larger.  I chose the 1202 because it had plenty of inputs, simple to use for the intended opp, and I could role it in on a cart with the cables.  All told I needed a stereo and 3 mono post fade outs with volume knobs.  I chose LR for the mains, aux 1,2 for fill and alternate room, and 'Y'ed the control room outs for sub. The system sounded fine and was signed off on by the powers that be.

A couple days later a co-worker starts ripping me a new one over "never Y two outputs!"  This turned into a huge fiasco.  For the record my position is a couple notches higher than his and I tend to think more like our managers "get it done safe and efficient."

When I chose to Y the outputs I was thinking:

I need another isolated mono out.

The control room outs are buffered from the main LR and the knob is next to the LR knob.  Should make a nice sub-woofer feed.

Because Mackie outs are impedance balanced and the signal is mostly mono I can Y them with out any problem or risk of damage.

I don't care about any added distortion because the sub-woofer processing will filter out most of it anyway.

I have several days to change it if it does not work.

Listening tests show it sounds good, done, onto the other hundred things on my todo list for the day.

So, was I wrong?  Did I commit a "Cardinal sin" by Ying the control room outs?

In hind sight, yes he caught me on a technicality.  However I stand by that the risk of damage was zero and the out come served the needs of the gig well.




Title: Re: Ying outputs, was I really that wrong?
Post by: Steve M Smith on January 04, 2016, 02:47:05 am
I would have thought that most ouputs would have some sort of series resistor to prevent short circuit damage anyway so they should sum together o.k.

If not, a few low value resistors in the Y lead should sort it out.


Steve.
Title: Re: Ying outputs, was I really that wrong?
Post by: Keith Broughton on January 04, 2016, 06:46:11 am


The control room outs are buffered from the main LR and the knob is next to the LR knob.  Should make a nice sub-woofer feed.

Because Mackie outs are impedance balanced and the signal is mostly mono I can Y them with out any problem or risk of damage.

I don't care about any added distortion because the sub-woofer processing will filter out most of it anyway.


Control room out...(good point).

No risk of damage...how did you arrive at that conclusion? (not a good point)

added distortion....I'm looking for the processor that filters out added distortion! (very bad point)
Title: Re: Ying outputs, was I really that wrong?
Post by: Jim McKeveny on January 04, 2016, 07:25:26 am
"Y" to split, or "Y" to sum?
Title: Re: Ying outputs, was I really that wrong?
Post by: Jordan Wolf on January 04, 2016, 07:55:49 am
If you used the wye to split the signal, I see no issue; physical isolation (i.e. transformer) may be a nice extra step since most likely powering from a different AC source.

I would have not used the control room out since soloing affects the output signal. Go mono (Main Left only) for main and use Main Right for sub feed; that still gives you independent level control and you can control how much goes to the sub by using the pan knob.


- Jordan Wolf
Title: Re: Ying outputs, was I really that wrong?
Post by: Ivan Beaver on January 04, 2016, 08:13:43 am
The biggest issue for not Ying outputs is that the output impedance of each channel will load down the other channel and vice versa.

Yes you will have some harmonic distortion-due to the extra load.

Yes you can have other types of distortion-freq response etc because the coupling caps are designed for a particular load and when that load is well below expected, the low freq can roll off.

However the biggest issue could be actual DAMAGE to the output transistors or ICs.

They are not designed to drive a load rated at their output impedance, so they could fail.

SURE-it might "kinda work" for a little while, but it is NOT something you want to do.

You might get lucky and get away with it, but it is a VERY BAD idea.
Title: Re: Ying outputs, was I really that wrong?
Post by: Jim McKeveny on January 04, 2016, 09:14:36 am
The biggest issue for not Ying outputs is that the output impedance of each channel will load down the other channel and vice versa.

Yes you will have some harmonic distortion-due to the extra load.

Yes you can have other types of distortion-freq response etc because the coupling caps are designed for a particular load and when that load is well below expected, the low freq can roll off.

However the biggest issue could be actual DAMAGE to the output transistors or ICs.

They are not designed to drive a load rated at their output impedance, so they could fail.

SURE-it might "kinda work" for a little while, but it is NOT something you want to do.

You might get lucky and get away with it, but it is a VERY BAD idea.

We still do not know if this was a split or sum, so we are still in speculation mode.

However, if it was a split......and a single split not daisy chain, our pal JR has let us know that knowledgeable mfrs. anticipate 'streme apps by the end-user, because building in mild overcapacity is always cheaper than enduring client-caused failure with commensurate warranty costs & mfr. rep damage.

My advice to the OP is to always expect way more I/O than spec'd.

Title: Re: Ying outputs, was I really that wrong?
Post by: g'bye, Dick Rees on January 04, 2016, 09:17:08 am
We still do not know if this was a split or sum, so we are still in speculation mode.

Jim...

Read the thread title again...
Title: Re: Ying outputs, was I really that wrong?
Post by: Jim McKeveny on January 04, 2016, 09:22:54 am
Jim...

Read the thread title again...

It's the stereo/2 chan. thing that leads me here. With differential diagnosis assume nothing.

Was a TRS Stereo Control Room output sent to mono subs? That would be backass by our standards, but could be called a "Y". We dunno.

Like a lawyer, we have to parse every
syllable. It's an affliction, I admit
Title: Re: Ying outputs, was I really that wrong?
Post by: Scott Hofmann on January 04, 2016, 09:36:18 am
It's the stereo/2 chan. thing that leads me here. With differential diagnosis assume nothing.

Was a TRS Stereo Control Room output sent to mono subs? That would be backass by our standards, but could be called a "Y". We dunno.

Like a lawyer, we have to parse every

Parse it again.... "sub" is singular!
syllable. It's an affliction, I admit
Title: Re: Ying outputs, was I really that wrong?
Post by: John Roberts {JR} on January 04, 2016, 09:41:42 am
This comes down to "can you do it" vs. "should you do it". As you've already determined you "can" do it for this particular situation. No smoke was released and you didn't hear audible degradation.

"Should" you do it, NO. Just because some manufacturers anticipate and protect against such questionable connections, don't ASSume you can always do that. Someday you might encounter an unprotected design and it won't work, might even harm the poor thing. 

I recall putting additional dedicated mono outputs on some Peavey mixers.

JR

PS: A custom Y adapter with added series resistors (say 1-2Kohm per leg) could protect weak outputs and still deliver reasonably low source impedance. Ironically products low on the value spectrum often use larger impedance build out resistors, mainly because they can't drive 600 ohms with the wimpy GP op amps, not some grand plan. 
Title: Re: Ying outputs, was I really that wrong?
Post by: Geoff Doane on January 04, 2016, 09:51:34 am

Because Mackie outs are impedance balanced and the signal is mostly mono I can Y them with out any problem or risk of damage.


Impedance balanced means that the "low" side has an impedance to ground, but is not driven.  You're still connecting two driven signals together through a rather low output impedance, and that's what could cause distortion (I have heard it happen), although any damage is less likely.

However if it's really "mostly mono", then why not just use one output? 

Your co-worker was correct that it is not "best practice".  It's somewhat questionable whether it was worth making it an issue however.  For the record, I routinely Y a pair of e604s used for congas, and nothing bad has happened yet.  It's a much better solution than having to drag out a 32 channel console instead of a 16  8)

GTD
Title: Re: Ying outputs, was I really that wrong?
Post by: Scott Helmke on January 04, 2016, 10:04:39 am
I just had the 1202 VLZ schematics handy, and there have been some different versions over the years.

But the Control Room outputs do have 120 ohm resistors. So no real problem, aside from "showing bad form, old boy".
Title: Re: Ying outputs, was I really that wrong?
Post by: Riley Casey on January 04, 2016, 10:05:15 am
So to sum up  :o , its not a good idea to parallel actively driven outputs because they will likely cause distortion when driven to any significant level and in the case of poorly designed equipment might damage the output stages.

Real question in this instance I would propose is why bother?  What is the value of summing L & R when trying to drive a subwoofer feed?  Unless you have some VERY unusual program material the low frequency component of the program is unlikely to be hard panned.  Left OR right will have all of the under 100 hz information needed to ensure that the sub feed is delivering the goods.  You may  well need some additional gain after the mixer output but summing probably wasn't worth the effort.
Title: Re: Ying outputs, was I really that wrong?
Post by: Jim McKeveny on January 04, 2016, 10:05:30 am


A TRS Control Room output to a mono XLR? Without pinouts we have no clarity.
Title: Re: Ying outputs, was I really that wrong?
Post by: Steve M Smith on January 04, 2016, 10:09:15 am
I just had the 1202 VLZ schematics handy, and there have been some different versions over the years.

But the Control Room outputs do have 120 ohm resistors. So no real problem, aside from "showing bad form, old boy".
I just found the same thing on the 1604 circuit so yes, I would have no problem parallel connecting the outputs.
 
Note, every output has these resistors.  Aside from summing together, they are good to protect an output from short circuits to such as putting a two pole 1/4" jack into a TRS output which will ground the ring connection (although this is usually just a 120 ohm resistor to ground).
 
http://bmamps.com/Schematics/Mackie/Mackie_1604-VLZ_16_channel_mixer.pdf (http://bmamps.com/Schematics/Mackie/Mackie_1604-VLZ_16_channel_mixer.pdf)

A TRS Control Room output to a mono XLR? Without pinouts we have no clarity.
In the case of the Mackie, the R connection is just a ground compensation resistor.


Steve.
Title: Re: Ying outputs, was I really that wrong?
Post by: John Roberts {JR} on January 04, 2016, 10:14:30 am
We already know it didn't blow up.

Not best practice... Lets not encourage people to do this.

JR
Title: Re: Ying outputs, was I really that wrong?
Post by: Tom Bourke on January 04, 2016, 11:22:05 am
We already know it didn't blow up.

Not best practice... Lets not encourage people to do this.

JR
I admit not best practice and not something I normally do.  Certainly not for normal gigs or with other gear.  However in THIS case with an older version Mackie, that I know to have robust outputs, and to just add some thump to a pre-function room for 4 hours. I felt a mono sum with a Y cable was acceptable.

When I set it, I did add to the board tape a note not to solo anything.  In this case no need to solo anyway. 

This was more of a political problem.  I was promoted very fast because I am the go to guy for technical problems and troubleshooting.  My real failure was not anticipating the shit-storm from getting "caught" on a technical compromise.
Title: Re: Ying outputs, was I really that wrong?
Post by: Tom Bourke on January 04, 2016, 11:36:35 am
PS: A custom Y adapter with added series resistors (say 1-2Kohm per leg) could protect weak outputs and still deliver reasonably low source impedance. Ironically products low on the value spectrum often use larger impedance build out resistors, mainly because they can't drive 600 ohms with the wimpy GP op amps, not some grand plan.
I do have such a Y cable in my personal pile of gak. It dates back to my first couple years in the biz.  I could not afford much gear so I made due.  The above cable would give me a mono sum and then on the parallel out of the sub amp I put a capacitor shorting pin 2 and 3 to knock the highs out of the subs.  Ahh those were the days, heavy iron amps and Radio Shack.

These days I don't bring much more to work than a personal laptop, Fluke meter, and a nice network cable tester.
Title: Re: Ying outputs, was I really that wrong?
Post by: John Roberts {JR} on January 04, 2016, 12:08:07 pm
I can't opine about the politics, but in general when a subordinate backs me up by bringing something like that to my attention, I thank him for having my back. Of course how he presented that information may make a difference.

Maybe make him your technical guy and continue your focus on management.

Good luck, politics can be a PIA, while should be less of a hassle the smaller the organization.

JR

PS: Maybe show you custom Y cable to you critic (or give it to him) to let him know that you know too.
Title: Re: Ying outputs, was I really that wrong?
Post by: Erik Jerde on January 04, 2016, 12:11:53 pm
Read this:

Why not wye?
http://www.rane.com/note109.html
Title: Re: Ying outputs, was I really that wrong?
Post by: Matthew Knischewsky on January 04, 2016, 01:51:27 pm
Read this:

Why not wye?
http://www.rane.com/note109.html

I carry several adapters built from this note in my work box for the times when summing just has to be done and the "correct" solution is not available. The first one I made was for a Mackie mixer that did everything I needed except mono out.
Matt
Title: Re: Ying outputs, was I really that wrong?
Post by: Luke Geis on January 21, 2016, 11:33:14 pm
Perhaps this is fuel for the fire?

So if you have two balanced outs and you want to make a mono connection that sums them together, could you not use only pin 1 and 2 from the left send and then use only pin 3 from what would be the right side? The connections in the outputs would then never technically be in direct connection with each other. The balanced input at the speaker or amp would sum the signals together as normal right? The impedance would not harm the outputs either as it would just be like dropping a hot or cold pin. Or is this crazy talk? 

I only suggest this as a solution, not as a practice. I have never done this myself as it was my learning as well that coupling outputs into one source is bad juju.

My understanding of most music is that the sub content is almost always up the middle anyway. I suppose if your listening to an old Jimi Hendrix recording you could have drums and bass on one side of the stereo field, but I would solve that problem by just panning things up the middle anyway. In a distributed system where it is simply playback and background, mono and or summed music sources is usually ideal for simplicity in signal delineation and reduction of confusion in recordings where the sound could come from different speakers in a multitude of locations.
Title: Re: Ying outputs, was I really that wrong?
Post by: John Roberts {JR} on January 22, 2016, 12:00:16 am
Perhaps this is fuel for the fire?
why;;;?
Quote
So if you have two balanced outs and you want to make a mono connection that sums them together, could you not use only pin 1 and 2 from the left send and then use only pin 3 from what would be the right side? The connections in the outputs would then never technically be in direct connection with each other.
Ummm Maybe... I think you mean 2 from the left to 2 (+) mono input, and 3 from right to 3 -(-) mono input. (1 is for shield).

The bad news is balanced/differential outputs deliver signal as 2 wrt 3, so if one or the other output line is floating and unterminated the output between them may be inaccurate.   
Quote
The balanced input at the speaker or amp would sum the signals together as normal right? The impedance would not harm the outputs either as it would just be like dropping a hot or cold pin. Or is this crazy talk? 
crazy like a.... (insert your own joke here....)
Quote
I only suggest this as a solution, not as a practice. I have never done this myself as it was my learning as well that coupling outputs into one source is bad juju.
It will work for some outputs and not for others, so not a reliable approach.
Quote
My understanding of most music is that the sub content is almost always up the middle anyway.
an old affectation from the days when vinyl was king... Pure left or pure right LF peaks could push the needle out of the track.
Quote
I suppose if your listening to an old Jimi Hendrix recording you could have drums and bass on one side of the stereo field, but I would solve that problem by just panning things up the middle anyway. In a distributed system where it is simply playback and background, mono and or summed music sources is usually ideal for simplicity in signal delineation and reduction of confusion in recordings where the sound could come from different speakers in a multitude of locations.
SR is mostly mono...

JR
Title: Re: Ying outputs, was I really that wrong?
Post by: Steve M Smith on January 22, 2016, 02:54:14 am
So if you have two balanced outs and you want to make a mono connection that sums them together, could you not use only pin 1 and 2 from the left send and then use only pin 3 from what would be the right side? The connections in the outputs would then never technically be in direct connection with each other. The balanced input at the speaker or amp would sum the signals together as normal right?

Definitely not.  Someone did this to me once.  They gave me an XLR lead to plug into the mixer which had a minidisc player on the other end.  They were using it for backing music to a laser show.

A quick test showed it to be working but when it came to show time, about half way through, it went very quiet with no bottom end.  Next track, it was o.k. again.

It turned out that what I thought was a summed input was actually left on pin 2 and right on pin 3 so the balanced input on the mixer was subtracting one from the other rather than adding.  This means that all you can hear is the difference between the two signals.

The surprising thing was how much difference there was between left and right on most tracks.

On closer investigation, I found that they had two adaptor leads.  Two RCA phonos into a male XLR and two RCA phonos into a female XLR.  They usuallr ran one each end of a mic lead so they could make up a long stereo RCA phono lead.  As they knew I needed an XLR input they just plugged their adaptor in at their end and presented me with their XLR end.


Steve.
Title: Re: Ying outputs, was I really that wrong?
Post by: Luke Geis on January 22, 2016, 03:55:39 am
I was being tongue in cheek so keep that in mind.

If the audio was not panned and the output could pass the signal without fail, then in theory you could make a two into one with the above idea, but there would be no point to do so. It was a rhetorical idea.

In the case of the OP's query, having the desired tracks cued would have been ideal actually. It would have given him two mono outputs instead. Most cue's sum the signal to mono and a control room output would follow that. The relative mix would be different if in PFL, but many mixers also have AFL cue options as well, so that may have worked better in this case.
Title: Re: Ying outputs, was I really that wrong?
Post by: Ivan Beaver on January 22, 2016, 11:57:55 am

Definitely not.
It turned out that what I thought was a summed input was actually left on pin 2 and right on pin 3 so the balanced input on the mixer was subtracting one from the other rather than adding.  This means that all you can hear is the difference between the two signals.



EXACTLY

This also know as "common mode rejection ratio" or CMRR on a spec sheet.

It is EXACTLY what help keep noise out of a balanced line.
 
Whatever is COMMON to both pins 2 and 3 gets rejected by whatever the CMRR is.

DO NOT DO THIS!!!!!!!!!
Title: Re: Ying outputs, was I really that wrong?
Post by: John Roberts {JR} on January 22, 2016, 12:26:26 pm
EXACTLY

This also know as "common mode rejection ratio" or CMRR on a spec sheet.

It is EXACTLY what help keep noise out of a balanced line.
 
Whatever is COMMON to both pins 2 and 3 gets rejected by whatever the CMRR is.

DO NOT DO THIS!!!!!!!!!
For a stereo feed it gives you Left minus Right... For panned stereo, signals panned dead center will completely drop out, for true stereo pair mic'd recordings coherent direct sounds will be diminished, and non-direct incoherent (ambience) sounds will not be diminished.

This L-R is a very old school technique to extract ambience sound from stereo recordings to feed rear speakers, since dominant mono signals cancel, what's left works to enhance ambience effects (generally some delay is added to rear speakers to keep sources localized in front speakers).

And this is while I grabbed pin 2 from one side feeding 2 and pin 3 from the other side to feed 3, to deliver L+R or summed mono... but as I already mentioned, unused outputs may need to be terminated to get a good result, and single legged outputs won't have a negative signal to grab yo get polarity right.

Several reasons why this isn't reliable enough to use. 

JR
Title: Re: Ying outputs, was I really that wrong?
Post by: Steve M Smith on January 22, 2016, 01:07:13 pm
DO NOT DO THIS!!!!!!!!!

I did know not to do this.
However, at the time, I didn't realise I was doing this!
I worked it out quite quickly once it happened though.

This L-R is a very old school technique to extract ambience sound from stereo recordings to feed rear speakers

Many years ago I had a centre rear speaker connected to the left and right outputs of my amplifier to do this.  Known as the Hafler hook up if my memory is to be trusted.


Steve.
Title: Re: Ying outputs, was I really that wrong?
Post by: Ivan Beaver on January 22, 2016, 01:28:58 pm


Many years ago I had a centre rear speaker connected to the left and right outputs of my amplifier to do this.  Known as the Hafler hook up if my memory is to be trusted.
 
Phase Linear suggest hooking up rear speakers between the + terminals (speaker + to L_+ and speaker - to right +) for an effect.

I remember another example was to use 2 rear speaker hooked in series-but of opposite polarity on 1 speaker (so you had the 2 neg hooked together) and the + of each speaker hooked to the + of the amp channels, R and L.

Yes you would get "something".

And that was pretty good back in the 70s-------------


Steve.
Title: Re: Ying outputs, was I really that wrong?
Post by: Geoff Doane on January 22, 2016, 02:34:18 pm
So if you have two balanced outs and you want to make a mono connection that sums them together, could you not use only pin 1 and 2 from the left send and then use only pin 3 from what would be the right side? The connections in the outputs would then never technically be in direct connection with each other. The balanced input at the speaker or amp would sum the signals together as normal right? The impedance would not harm the outputs either as it would just be like dropping a hot or cold pin. Or is this crazy talk? 

JR and I have had this discussion before.  :)  I don't think either one of us convinced the other we were right.

What you suggest will work (I've done it extensively in an installation), but there are too many "gotchas" to assume you can get away with it all the time.  First of all, you need to know that ALL the inputs and outputs will be "balanced", and the outputs have to be simple transformerless outputs with no circuitry that tries to mimic a transformer.  Impedance balanced outputs also don't work.

In my particular case, I was dealing with racks of distribution amplifiers that drove the + and - outputs with individual op-amps.  I could take + from the left, and - from the right into a differential input, and get a mono sum (with no gain or loss, if the signals were identical) of the two signals.  The beauty of this scheme is that it can be done with just simple wire.  No resistors, pad boards or other gack to lose signal or degrade the output impedance.  If the DA outputs are 30Ω per leg, you can still drive 1000 feet of cable.  Probably not a good idea to try that after you've added 1KΩ of series resistance with a passive combiner.

Unfortunately (depending on your point of view), simple active outputs with no cross connection are increasingly rare these days, because they usually don't tolerate having one side connected to ground gracefully.  Connecting a "balanced" source to an unbalanced destination is likely far more common than trying to combine two signals.  I actually stole the idea from the schematic for a piece of broadcast processing gear.  It was a stereo processor, but had left, right and mono outputs.  In that era, it would be very unlikely to encounter any equipment that didn't have a differential input.

GTD
Title: Re: Ying outputs, was I really that wrong?
Post by: Luke Geis on January 22, 2016, 03:27:56 pm
I know ^^^^^^^^^ I was being rhetorical and facetious. Fuel for the fire........

I brought it up only because the theory supports it under specific conditions. I don't have such a cable and don't intend to ever make one. I also don't condone others to make one.
Title: Re: Ying outputs, was I really that wrong?
Post by: Mike Caldwell on January 22, 2016, 05:34:49 pm
And this is "why" I have one these.
Title: Re: Ying outputs, was I really that wrong?
Post by: Luke Geis on January 22, 2016, 08:11:32 pm
Now one of those^^^^^^^^^^^ I do have
Title: Re: Ying outputs, was I really that wrong?
Post by: David Sturzenbecher on January 23, 2016, 01:43:56 am

Now one of those^^^^^^^^^^^ I do have

Best looking resistor box I have ever seen!


Sent from my iPhone using Tapatalk
Title: Re: Ying outputs, was I really that wrong?
Post by: Mike Caldwell on January 23, 2016, 07:56:40 am
Best looking resistor box I have ever seen!


Sent from my iPhone using Tapatalk

The Whirlwind box is a transformer isolated combiner.
Title: Re: Ying outputs, was I really that wrong?
Post by: Luke Geis on January 23, 2016, 07:40:31 pm
mine indeed is.