DIY stereo to mono summing unit

[paully]

Unless I'm totally misreading what you're asking, wouldn't it be easier to simply reassign everything to a stereo sub, then pan the L/R outputs, which are normally hard left and right, to center? That should give you a quick mono mix. Pardon the dumb question, but does the board not have a mono switch of some kind? If you really need a hardware switch, any double pole on/off toggle switch could be used to short pin 2 L to pin 2 R on one pole, and pin 3 L to pin 3 L on the other pole. Could be mounted on the board or in a seperate metal box with appropriate connectors.

Paul

[A27Hull]

Well you are right.

my 002r/PT does have the pan functions on the busses, and I guess (uneducatedly) that that would be the same. The 002r does have a Mono function on the hardware that monos the HP mix/output, but..

I want to make something , with which I could hook up to analog outputs other than those feeding my main stereo mix, and assign the Master bus to several outputs..

analog 1/2 > stereo analog

S/PDIF 1/2 > dig. monitor inputs

analog 3/4 > stereo to mono sum>fed to another monitor amp/speaker combo

essentially then I could keep the internal mix setting the same, but choose what to monitor externally.

I guess then there would be the problem of having an external monitor selector for the analog outputs.

I wonder if making a ST>M box with bypass feature would be too complex?

[paully]

[QUOTE=A27Hull;1432383]
analog 1/2 > stereo analog

analog 3/4 > stereo to mono sum>fed to another monitor amp/speaker combo

I wonder if making a ST>M box with bypass feature would be too complex?c/QUOTE]
>>>>>>>>>>>>>>>>
What you're suggesting is really a bad idea for this reason. Listening to a mono or stereo mix thru a seperate amp/speaker combo is not going to yield the same results as monitoring thru the main amp/speaker combo. Different components = different sound. Better to maintain a constant listening environment. A simple shorting switch like I suggested, placed after the analog 1/2 bus, should work just fine. Switch on = mono, switch off = stereo. It's really not a hard project to whip together if your into DIY.

Best, Paul

[A27Hull]

Thanks Paul!

You make a good point.

So, the double pole switch that you suggested above would act as the summing device/switcher?

Would that be a single throw, double throw, or changeover switch?

[A27Hull]

If you really need a hardware switch, any double pole on/off toggle switch could be used to short pin 2 L to pin 2 R on one pole, and pin 3 L to pin 3 L on the other pole.

Like this?

[paully]

I just caught your thumbnail. Pin 1 is always ground, and doesn't need to connect to the switch. The diagram shows pin 2 L connecting(shorting) to pin 2 R (red line) for mono 'on', but pin 3 L should also short to pin 3 R (black lines). A must for a balanced output.

The simplest switch would be a 2 pole, single throw, on/off. A toggle switch is commonly available. No offense, but are you really sure you are ready to tackle something like this?

Paul

[A27Hull]

yeah, I'm basically just starting with electronics, I went to a 4 yr college for audio school, and after graduating, have a desire to get technical with it, not just knob turning and such.

I've only built some simple LED gadgets, a lot of audio cables, a WC terminator, and a Velleman Super Stereo ear kit.

I'm comfortable with soldering, and I'm starting my DC/AC courses this fall.

I haven't a clue of the theories behind electronics yet, other than my light reading thus far into the basics, and a little math.

I truly appreciate your kindness and patience. I obviously have a lot to learn

[A27Hull]

also,

if I were to use an aluminum enclosure, would it benefit me to connect the signal grounds to the chassis at one point (star config.), or should I not bother with the chassis ground?

-thanks,

Andrew

[paully]

Originally Posted by A27Hull

if I were to use an aluminum enclosure...

would it benefit me to connect the signal grounds to the chassis at one point (star config.), or should I not bother with the chassis ground?...

>>>>>>>>>>>>>>>>
Andrew,

The best metal case you can use is steel. Aluminum not recommended. A small box that can sit on the desktop is sufficient. It must be directly grounded to pin 1 of the L and R ouputs. That ground must continue from the box to the power amp. That's the shield, and eliminates interference/hum. The shield is a part of the wire that you use to connect the box, both from the board and to the amp. Called coaxial wire. You can find all the parts that you need at Mouser Electronics - Electronic Component Distributor .

For a switch, I'd use a latching (when depressed, it stays on until you push it a second time, then it turns off) DPDT pushbutton switch. If you go to Mouser, type in the number 612-PBH2UEEK at the search port. That's a decent switch. For a box, do the same thing with the number 546-1441-6.

If you need help with the wiring diagram, you need to post up some info.

1) The type of jack on the L/R board outputs: XLR(male or female) or 1/4" TRS.

2) The same info for the amp's input jacks.

Once we know that, it shouldn't be hard to get you going. Parts may run you a few bucks, so be prepared.

FWIW, go to Radio Shack and pick up a cheap, hand-held multmeter like this that can read continuity in ohms RadioShack.com - Cables, Parts & Connectors: Test & measuring equipment: Multimeters: Pocket-Sized 8-Range Analog Multimeter . Money well spent.

Paul

[A27Hull]

Thank you very much for the help Paul,

As far as the signal chain, the DA outputs are going to run through an ATTY to the...ehem "MonoBlok," then to some active JBLs.

the ins ans out could either be TRS or XLR, but I can convert either way, I guess it just matters on the size of the box, (which I did have a look at on Mouser btw,)

Panel mount XLRs might take up a little less room than TRS jacks internally, but more externally. I'd like to keep it XLR if possible.

I'm gonna need a 'nibbler'



I think I'm devising something silly though... The ATTY is due this week, and as soon as I can get it, I'm gonna open it up to see inside. The box size is perfect, and since it already be on my DAW stand anyway, why not add a mono switch to it? That is of course the insides look about like I expect, a volume pot and a mute switch, and some wire.

Its a passive device, so perhaps theres room in the unit to drill/punch another hole, and hook up a DPDT switch at the back end, without disturbing the front end volume control and mute button.

I'm willing to be discouraged of this idea, but the warranty I would end up voiding isn't going to deter me for such a low cost item.

I set up a Mouser accnt just in case.

I'll let you know what I think when it gets here.

As far as a DMM, I recently purchased one I thought was rather accurate..a fluke 177, but as I found out through my experience on another geekslutz thread, and additionally through other web sources, its not as accurate as needed for working with audio analysis. It only has something like 5000 counts or so per range.

I should'a gotten the 187..

Edit: Should I be worrying about impedance differences in the new 'Monoblok,' especially should I worry if I were to modify the ATTY?

[A27Hull]

So, as far as hook up,

I'm confused, if the output of the "monoblok" looses a pin on each side via the switch diagram above, that would make it an unbalanced line with an extra conductor in the wire,

I assume then the pict listed below is actually whats happening in the switch.

In the "on" mode, Pin 2L connects to Pin 2L and Pin 2R, Pin 3L connects to Pin 3L and Pin 3 R simulatneously...

Tell me if this is right.

thanks,

Andrew

[paully]

The second diagram is right. I've attached a thumbnail of the way I'd put it together as a stand-alone box. The box would not need jacks, just the switch. That's why I asked about the jack type earlier. Assuming XLR on the ATTN outs and on the amp inputs, a simple premade, 2 conductor shielded mic wire, cut in half, could be used on each leg. Great way to save money and real estate. With XLR, outputs(like a board's stereo outs) are usually male, and inputs (like a power amp's) are usually female. With a mic wire, you would merely cut it in half, run it into the box, and solder it up. It's ready to go. Same deal if the ins and outs are TRS. Just buy a couple of premade TRS 2 conductor shielded wires. You'll save a lot of time and expense. In the thumbnail, I included a resistor on each leg to knock down the mono signal if it's too loud. BTW, nothing wrong with that Fluke 177. They make great stuff.

[A27Hull]

Thanks very much Paul!

Very nice diagram.

I like the pre-made wire idea, but I got some spare 2552 on a spool, and I plan to hunt the parts you suggested tomorrow. Theres a nice parts store locally, but, as you suggested, I'll probably have to order the switch and the box.

[barbaroja]

I found this on the net:

"When two outputs are directly wired in parallel, you are effectively fighting with each other. When they are sending out different audio signal, then most of the time one of them is trying to push the output to different voltage than other output. This effectivley causes stress and overload to outputs. In best case you get uneven mix of the signals, lowered output voltage and increased distortion in the sound. In some cases you even risk on damaging the equipments because you are constatly overloading the outputs. The reason for this to happen is that line-level outputs are low-impedance voltage sources and they just aren't going to be very happy if you connect them to each other. Anytime the output differs between the two (as it will almost all the time in stereo) each will be trying to ram current through the other -- more current than they're designed to source, and way more current than they're designed to accept (they're designed to accept practicly nothing)."
Simple mixer

¿Are we risking our sound card outputs if doing the above diagram?


¿What happens if I am in mono and the pan goes alll to one side? ¿Does the other sound card output gets fried up?


¿Then, using no resistor can damage the sound quality as the above pharagraph says?

[A27Hull]

I'm not sure what the paragraph you quoted is referring too.

What we are essentially doing is using a switch to sum L and R to Mono(one side). Think of it as a "Y" connection, only the switch allows us to change from "l l" to "Y"

When signals are summed the total signal level may increase, and that is why Paul suggested the resistors. Theoretically, each resistor would need to be matched.

To answer your questions,

1. No, I don't think so.

2. For the above situation, my speakers will only produce sound out of one side when the switch is engaged, but in stereo when the switch is not.

3. If using no resistor and the audio signal (mono) is a large enough level to distort the amplifier, then yes, audio degradation will occur. But that is not what the paragraph is saying.

[Radioman]

Originally Posted by barbaroja

"When two outputs are directly wired in parallel, you are effectively fighting with each other. In some cases you even risk on damaging the equipments because you are constatly overloading the outputs. Anytime the output differs between the two each will be trying to ram more current than they're designed to source, and way more current than they're designed to accept"
Simple mixer

Then, using no resistor can damage the sound quality as the above pharagraph says?

Absolutely true, the outputs must be followed by a couple of resistors before wiring together hot (or cold) L and R wires as shown in the "simple mixer" link.
The 10 kohm and the wiring diagram are optimized for a consumer unbalanced gear, for a balanced enviroment the circuit must be doubled for the hot and cold wires, and the resistors value could be smaller.
If the output impedance is low enough (about 100-200 ohm) and the speakers input impedance is high enough (10 kohm or greater) the R could be a 1,5 kohm-2,2 kohm and the signal loss could be very small (about 1-2 dB).

[barbaroja]

Originally Posted by Radioman

Absolutely true, the outputs must be followed by a couple of resistors before wiring together hot (or cold) L and R wires as shown in the "simple mixer" link.
The 10 kohm and the wiring diagram are optimized for a consumer unbalanced gear, for a balanced enviroment the circuit must be doubled for the hot and cold wires, and the resistors value could be smaller.
If the output impedance is low enough (about 100-200 ohm) and the speakers input impedance is high enough (10 kohm or greater) the R could be a 1,5 kohm-2,2 kohm and the signal loss could be very small (about 1-2 dB).

Most of the gear I researched has this values: For monitors input impedance is always 10Kohm. Exceptions to the rule were the barefoot micromain 27 wich are 50Kohm, I suppose the swbwoofers raise the impedance.
For audio interfaces they were ranging between 51(presonus firepod) and 300ohm (M-Audio Audiophile 192 when balanced). Mine (TC electronic konnekt 8) has a impedance of 100ohm balanced or unbalanced.

So,
¿is a DIY project in this topic, made exactly for the impedances of the gear I actually have so, If I change soundcard or monitors, will I have to recheck the circuit?

¿With my values (10K input monitors, 100ohm output soundcard), which would be the perfect ones to match the circuit?

¿May I trust this circuit sonically? Somewhere I read about HF loss.

Thanks!

[A27Hull]

Originally Posted by Radioman

Absolutely true, the outputs must be followed by a couple of resistors before wiring together hot (or cold) L and R wires as shown in the "simple mixer" link.
The 10 kohm and the wiring diagram are optimized for a consumer unbalanced gear, for a balanced enviroment the circuit must be doubled for the hot and cold wires, and the resistors value could be smaller.
If the output impedance is low enough (about 100-200 ohm) and the speakers input impedance is high enough (10 kohm or greater) the R could be a 1,5 kohm-2,2 kohm and the signal loss could be very small (about 1-2 dB).

Exactly why i shouldn't be the one telling him what is what.

I would think that since both impedances are low, and in parallel, assuming that they are the same value, the sum of them divides as such: assuming 50Ω output Z of each channel, 1/.02+.02 or 1/.04 = approx 25 Zeq

This would seem pretty low if the outputs were shorted directly. The question is whether or not the output circuit used in my 002r is capable of driving that low of an Z and how the input of the amp responds to increased current and voltag levels. Overloading, more clearly, pulling more current from the output than under normal load is definitely possible.

Also, the capability of a sound card to safely drive a lower Z will depend on the circuitry used, and it is most likely different from card to card. More current pull equals higher heat dissipation, which can lead to thermal runaway, and subsequent distruction of the output transistors/ICs.

But,

At +4dBu, or 1.228 Vrms, the current and wattage would be this:

1.228/25 roughly 50mA, and 1.228(.05) roughly 62mW.

Surely a output IC could handle 62mW heat dissipation right? A higher output Z, say 50 Zeq, would dissipate 32mW. I realize that much lower than 25 Zeq, one would dissipate even more W, but isn't it still a possibility that the output circuits are designed to protect if not handle lower Zeq and possibly output shorts?

Disregarding impedances, if two dissimilar audio channels are summed, the product is will have a differential effect, masking, resonating, and canceling along the frequency spectrum. The overall effect of the resonances and cancelations along the frequency spectrum equals an overall rms level increase in the amplitude domain. True?

If so, it would seem that the two added resistors are not affecting the parallel impedance, but attenuating the series rms. If the amp input section has enough headroom to safely deal with a 6dB increase in rms voltage level or 2.5Vrms, the lack of the resistors shouldn't matter should it?

Concerning amp input current, 2.5Vrms divided by say 10K input Z is roughly 250uA, 1.228 is 123uA. If the difference between out and in is 49.8mA, where does the extra current go?

If I am wrong I need to know.

[Radioman]

Originally Posted by barbaroja

¿is a DIY project in this topic, made exactly for the impedances of the gear I actually have so, If I change soundcard or monitors, will I have to recheck the circuit?

With Pro balanced gears at 99% no, but they are only 4 resistors...
Anyway if the soundcard output impedance remains the same (or became smaller) and/or the input monitors impedance remains the same (or become greater) definitevely not at all. With the value you shown above M-Audio would be the worst case.

Originally Posted by barbaroja

¿With my values (10K input monitors, 100ohm output soundcard), which would be the perfect ones to match the circuit?

A good rule is to have a input impedance 10 times greater in respect to the output impedance of the gear before so 1 kohm should be perfect, 10 times greater than 100 ohm and 10 times smaller than 10kohm. If you use 2.2k ohm it could be a good compromise for M-Audio too

Originally Posted by barbaroja

¿May I trust this circuit sonically? Somewhere I read about HF loss

with 1 kohm impedance output you must have 8 nF of parasite cable capacitance between pole and sleeve to obtain a HF loss of -3dB at 20 khz (3.6 nF for 2,2 kohm). A typical factor for a cable is 100pF (0.1 nF) for 1 metres, so you must have about 80 metres of balanced cables to have a -3dB HF loss at 20 khz.
Do you have the PC in your room and the speakers in your neighbour's kitchen ? If yes, maybe a -3dB HF loss at 20 khz is not your main monitoring problem ;-)

[barbaroja]

Originally Posted by Radioman

With Pro balanced gears at 99% no, but they are only 4 resistors...
Anyway if the soundcard output impedance remains the same (or became smaller) and/or the input monitors impedance remains the same (or become greater) definitevely not at all. With the value you shown above M-Audio would be the worst case.


A good rule is to have a input impedance 10 times greater in respect to the output impedance of the gear before so 1 kohm should be perfect, 10 times greater than 100 ohm and 10 times smaller than 10kohm. If you use 2.2k ohm it could be a good compromise for M-Audio too


with 1 kohm impedance output you must have 8 nF of parasite cable capacitance between pole and sleeve to obtain a HF loss of -3dB at 20 khz (3.6 nF for 2,2 kohm). A typical factor for a cable is 100pF (0.1 nF) for 1 metres, so you must have about 80 metres of balanced cables to have a -3dB HF loss at 20 khz.
Do you have the PC in your room and the speakers in your neighbour's kitchen ? If yes, maybe a -3dB HF loss at 20 khz is not your main monitoring problem ;-)

Thanks for the straight answers.

Sorry about my lack of understanding electronics, many of the questions may seem like I am asking without reason, but I am not.

So, what would happen with the M-Audio card if using 1K resistors?

When I asked about degrading the audio quality is not just in HF loss, wich now I know is irrelevant. Is the quality going down in any other measure? say THD or destroying the audio spectrum or anything like that?

Are commercial "mono maker" circuits like the ones found in monitor managers as Mackie ig knob or presonus central station based on this principle or would those circuits be better designed/more complex for mono making?

I still feel some fear to pan the signal hard right or left before reaching the device. Is the silent output side going to suffer?

Thanks a lot.

[Radioman]

Originally Posted by barbaroja

So, what would happen with the M-Audio card if using 1K resistors?

At 99% nothing relevant, 2.2 k it's for minimize the outputs load, but 1 k works well too. 1.5 kohm could be an optimum value for all the 3 soundcars.

Originally Posted by barbaroja

When I asked about degrading the audio quality is not just in HF loss, wich now I know is irrelevant. Is the quality going down in any other measure? say THD or destroying the audio spectrum or anything like that?

At 99% no at all, but you should have a THD measure set to be secure at 100%. If you want you can try to measure THD with your soundcard and a PC based signal generator and analyzer software (you can use a spectrum analizer SW with 90-100 dB of range and a SW sine generator)

Originally Posted by barbaroja

Are commercial "mono maker" circuits like the ones found in monitor managers as Mackie ig knob or presonus central station based on this principle or would those circuits be better designed/more complex for mono making?

At 99% they acts exactly like this one

Originally Posted by barbaroja

I still feel some fear to pan the signal hard right or left before reaching the device. Is the silent output side going to suffer?

No at all if you respect the 1:10 (but 1:5 is a good value too) "source to input" impedance ratio.

[barbaroja]

I cannot thank you enough. I think all my doubts are clear now. I will make the circuit and let you know my impressions.

One last question. If I make the box, and put a switch in the input so I can select or combine one or more sources into a selected output (lets say a digital piano plus the soundcard output plus the onboard sound), what would be the circuit for that?

Thanks

[Radioman]

Originally Posted by barbaroja

One last question. If I make the box, and put a switch in the input so I can select or combine one or more sources into a selected output (lets say a digital piano plus the soundcard output plus the onboard sound), what would be the circuit for that?

Something like this (balanced version)
One input for each source (8 for this diagram), resistors about 1-1.5 kohm instead 5 kohm, mono switch joining the same L and R output XLR pins.
Not selected inputs must be closed (input pins 2 and 3 must be wired together).
More the paralleled inputs, more the attenuation (1/2 for 2 inputs, 1/3 for 3 inputs and so on...) like explained for the passive summing network.

[barbaroja]

Originally Posted by Radioman

Something like this (balanced version)
One input for each source (8 for this diagram), resistors about 1-1.5 kohm instead 5 kohm

Why does the schematics suggest the 5ohm ones?

Originally Posted by Radioman

Not selected inputs must be closed (input pins 2 and 3 must be wired together).

Whats the reason for doing that? Is not that bad for the card? I mean I am shorting the hot and cold pins.

Originally Posted by Radioman

More the paralleled inputs, more the attenuation (1/2 for 2 inputs, 1/3 for 3 inputs and so on...) like explained for the passive summing network.

I see. Is there a big difference in gain if using 1,5kohm resistors and just three channels?

[Radioman]

Originally Posted by barbaroja

Why does the schematics suggest the 5ohm ones?

Whats the reason for doing that? Is not that bad for the card? I mean I am shorting the hot and cold pins.

I see. Is there a big difference in gain if using 1,5kohm resistors and just three channels?

1) 5000/8= 625 ohm
2) Shorting 2 and 3 detached from the DAW, only on passive mixer side
3) No, always 1/3 (about -10 dB)
For further informations look at the pdf

[stellar]

OK,

I built the circuit in post #13. my chain is

DAC, or tape machine (aka, input) --> mono/stereo switch --> stepped attentuator --> mute switch --> powered speaker .

the stepped attentuator is 10k per deck, 4 deck, so 20k balanced stereo, which matches my monitors.

for the mono stereo switch like I said its like in post # 13 and I am using a pair of 2.2k for the resistors.

Now, if there is one signal patched to the inputs, and I flip to mono, the signal DOES come out of both speakers. But if BOTH signals are patched into the inputs, and I flip to mono, the signal stays separate. what I want is to switch to mono, and L+R will combine to either L or R or both speakers if I choose, which is why the mute switches are per speaker and last in the chain.

First, I tried a "Y" configuration using 20k in order to "bridge" after the attentuator. This provided a dropped signal with losses in highs and lows.

Then, I moved the switch to before the attentuator and changed to 2.2k, also in a "Y" configuration. This gave a signal that was even in volume and frequency, but when in "stereo" mode the signals were still combining somewhat.

So I tried the configuration in post #13 using 2.2k before the attentuator and it also sounds good but doesn't make it mono.

Any ideas??

[A27Hull]

Originally Posted by stellar

DAC, or tape machine (aka, input) --> mono/stereo switch --> stepped attentuator --> mute switch --> powered speaker .

Any ideas??

If you want the mono-ed signal to reach both speakers perhaps a different circuit is needed, as the one above turns a stereo output to one mono output.

Using the exact picture above, only one speaker is used.

Now, if there is one signal patched to the inputs, and I flip to mono, the signal DOES come out of both speakers. But if BOTH signals are patched into the inputs, and I flip to mono, the signal stays separate. what I want is to switch to mono, and L+R will combine to either L or R or both speakers if I choose, which is why the mute switches are per speaker and last in the chain.

If you patch a stereo out to the inputs of the mono switch w/resistance, when the switch is engaged you should only have one output feeding into one side of the attenuator and mute switch.

You could modify the circuit to feed the mono-ed ouput of the switch back to both L and R inputs of the attenuator, etc...

Then, I moved the switch to before the attentuator and changed to 2.2k, also in a "Y" configuration. This gave a signal that was even in volume and frequency, but when in "stereo" mode the signals were still combining somewhat.

Perhaps the switch is the culprit, in that it is not completely isolating each channel from one another. The ideal situation would be complete isolation.

[Radioman]

Originally Posted by stellar

OK,
I built the circuit in post #13.
Now, if there is one signal patched to the inputs, and I flip to mono, the signal DOES come out of both speakers. But if BOTH signals are patched into the inputs, and I flip to mono, the signal stays separate.
Any ideas??

The post #13 schematic can't work, R1 and R2 allows only very partial blending in depending the source output impedance.
A correct schematic is the "simple mixer" linked in post #15 as explained in the following posts.

[stellar]

ok so I guess that means that in order to properly execute this for balanced stereo operation I will need an 8 pole switch? is that correct? This is the only way I can figure to get the resistors totally out of the circuit when not in mono. thanks.