DIY Monitor Controller Project

[Matt Syson]

Hi
Your basic plan will work but there are several practicalities that are hindering performance.
Firstly the 10 K resistors would be better as say 2K or 3K, enough so that the monitor outs aren't ever overloaded when monoed but low enough so that the following monitor amps and wiring see a reasonably low impedance to maintain HF response. If the monitor output can handle 600 Ohms comfortably you could take the resistors to 1K happily.
Secondly the 'outgoing' switches hould 'terminate' (short) the INPUT of the amplifier as it could pick up hum and or crosstalk. This is simply a matter of using 2 pole changeover switches rather than 'on off' that you have shown.
You will get varying levels switching mono and the various permutations of speakers but unless you go for a rather more complicated setup of switches so that the various loads stay constant whichever combination you have, it is inevitable.
Hint, having more than one pair of resistors may be a better way to go if you can figure out the switching necessary. If you are using 'toggles' you can get 4 pole changeover if needed and of course the 'push' switches are available up to 6 or 8 poles (more difficult to mount).
Matt S

[killanotrace]

Thanks for your input Matt, it's much appreciated. Here's a new schematic based on your recommendations:

Originally Posted by Matt Syson

Firstly the 10 K resistors would be better as say 2K or 3K, enough so that the monitor outs aren't ever overloaded when monoed but low enough so that the following monitor amps and wiring see a reasonably low impedance to maintain HF response. If the monitor output can handle 600 Ohms comfortably you could take the resistors to 1K happily.

Here's some information about the gear that will be connected to this circuit, which should help to home in on an ideal resistor value:

Control room monitor outputs - 120 ohms impedance, unbalanced

Main monitor inputs - 10 kohms, balanced

The monitor inputs are balanced XLR, but I will be running them unbalanced by using cables with pins 1 and 3 shorted. The product specs don't provide a different impedance for unbalanced operation and they recommend unbalancing in the way I just described... So I'm assuming the input impedance is still 10 kohms. Is that correct?

Originally Posted by Matt Syson

Secondly the 'outgoing' switches hould 'terminate' (short) the INPUT of the amplifier as it could pick up hum and or crosstalk. This is simply a matter of using 2 pole changeover switches rather than 'on off' that you have shown.

In the new drawing I've got SPDT switches that short the inputs directly to ground. As far as I know there is no problem with doing this but some people recommend a small value resistor instead of a direct short. Is there any merit in that?

Also, the main mute switch is just a DPST because to my understanding you should never short low impedance line outputs and they don't pick up interference anyway. Is that good or is there a better way?

Originally Posted by Matt Syson

You will get varying levels switching mono and the various permutations of speakers but unless you go for a rather more complicated setup of switches so that the various loads stay constant whichever combination you have, it is inevitable.

I can definitely live with that and I'm trying to keep this one simple. Later on I will attempt something more advanced but for now this will do.

Thanks again Matt, I checked your webpage and it's obvious you know your stuff... that's some really classy gear. It's way out of my price range at the moment (which you probably figured by the fact I'm running unbalanced). Nonetheless I want to get the most out of what I do have and do it right to the best of my ability.

I welcome input from anyone who has something to offer on this and hope I will be able to give back as my knowledge grows.

[ulysses]

A couple of refinements I can suggest:
1. You should use the double-throw switch method, with the OFF position grounded, on the master mute just as you've done for the output selector switches. You won't be shorting the console output to ground, just as you're not with the selector switches. You'll be shorting the monitor inputs to ground, which is fine. I would put the master mute on the other side of the series resistors. And if you want to ground through a resistor, I would use 120 ohms - same as the source impedance of the console output.

2. I would suggest using a 4P3T switch (perhaps a Grayhill rotary switch) to do the monitor selecting. Since there's probably no reason you'd want to use one speaker from each pair, this seems like a good idea to me. I don't see the value of your "dual mono vs. one-speaker-mono" anyway. At least, I see far more benefit to being able to switch quickly between Monitors A and Monitors B. With a 3-position switch, you can have A, B, and None, which effectively combines your A/B and your Master Mute into a single switch.

3. You can make this whole thing into a balanced system fairly easily. At the console's monitor output, connect the "cold" of your balanced cabling to ground through a 120-ohm resistor (which matches the "hot" output's source impedance). This is easy to do inside the body of a cable-end XLR or TRS plug. If you have a schematic, you can see if there's a series capacitor as well, and match that too. If not, don't worry about it. This will create an "impedance balanced" output which offers the same noise-canceling ability of any other balanced connection, but without the 6dB headroom gain of a symmetrically-driven output, and without the galvanic isolation of a transformer. It will double the number of poles you need in your switches, so you can decide if that's worth the cost and trouble.

That's all I've got for now.

[killanotrace]

Great stuff Ulysses, thank you. I need some time to digest all that and I'll post again shortly. Just checked your link, you have some serious gear on offer too. Thankful to get advice from people like this, seriously.

[Matt Syson]

Hi
I agree with Ulysses suggestions but would add that instead of having one pair of 2K resistors that you have 2 pairs and move them to the 'other side' of the point where the signal goes to the two switches.
This then means that the mono switch will have to have 2 poles (as shown unbalanced) or 4 poles to do it balanced.
I have not explained that well but my head hurts this morning.
This would have an advantage that the levels would not change as much if you were to feed both together although having said that you could argue that it should change to maintain a constant sound pressure in the room. Having 2 pairs of resistors would only be necessary if you don't use a rotary switch as Ulysses suggested.
Get yourself a handful of switches and put it together. As long as you have enough 'poles' you can always jig it around if it is not quite right.
There are many ways to do this, and apart from some 'circuit issues' as Ulysses and I have directed there is no 'right' or 'wrong' way to do this. The parts are cheap and you won't blow anything up so you can keep rewiring it until it does what you want.
Matt S

[killanotrace]

I appreciate the input, you've both been very helpful. Once I get around to building it I'll post again.

[killanotrace]

One last question though, just wanted to make sure 2k is a good value for the resistors considering the 120 ohm output impedance of the mixer?

[ulysses]

You want it as low as possible without being too low. How low is too low depends on the amp driving it. 2K is probably fine.

[Matt Syson]

Hi
I suggest 2K as it is a tradeoff between mainly, distortion if too low and HF roll off and possible noise pickup issues if too high.
If the mixer stated that it will perform to rated specifications driving say 600 Ohms then it will of course be fine at 2k and you could possibly go lower. I suspect that it won't actually tell you so the next best thing is to 'assume' it is happy into 1K or more so 2K would give some 'safety' margin as you are possibly going to get a signal at max positive peak on one 'side' and maybe max negative peak on the other.
Most 'chips' say in their data sheets what they can manage so if you knew what the stage uses you could find out that way. Even well known designers can't be bothered to read ALL of the data sheets and properly interpret the specifications so you are probably overdoing it to get so worried.
You could also work out the HF roll off if you wished. Cable is often about 120pF/foot nd working with say 2K or whichever value you can calculate what the roll off would be. Remember the usual equation gives a -3dB figure and you should be aiming for about 0.1 or 0.2 dB value at say 30KHz.
Matt S

[OMorgan]

I'm looking to do the same thing as the OP and this thread has been a great insight into how to do it in a slightly neater way than my proposed, easy as you can get two SPDT switches.

I have a question though, ulysses mentiones using a 4P3T switch which seems the a good way forward to me but I just want to clarify, that would be for a balanced system right? Thats what I would like to build,

So the 4 poles would attach to input left cold and hot and input right cold and hot. the three throw's would send those to monitors a, b or ground?

Also, in that setup wouldn't it be unavoidable to ground the OUTPUTS of your soundcard monitor outs, in the '3' position? Is it ok to do that? Or would you just leave the 3rd throw disconnected?

I've suddenly discovered electronics after years of audio engineering and its so exciting, though I can remember all but the most basic facts from my high school electronics of years ago, not that we ever got that deep into it even then!

EDIT - if wanting to add a 'mono' switch to a balanced system would you just have a DPDT switch which routes the hot-left to hot-right and cold-left to cold-right?

[Matt Syson]

Hi
While the scheme proposed by Ulysses will work, it is bad engineering practice to simply 'open circuit' any unused output (input to the monitor amplifier). This makes the switch requirement more complicated as you need to either short or terminate (say 75 Ohms) the amplifier input when NOT used.
Thus you would need a 4 pole 3 way to do this unbalanced and 6 pole 3 way (I think) to do this balanced.
A cheap compromise would be to fit say 1K5 resistors across all OUTPUT circuits so that there is some loading but not so muchj that it will upset the incoming feed.
Matt S

[citternman]

Hello fellow slutz,
I am interested in building something similar, but with a volume control. Would this be better placed in the circuit before or after the mono switch. And what value pot should be used.
Cheers,
-T

[guitarguy12387]

You may wanna look at a stepped attenuator for this type of application... Should sound 'better'/more balanced

I've seen the attenuator stages after the mono circuit... but it really shouldn't matter.

[Colombian]

I was wondering if the OP finally built this box. I am trying to build one very similar, just a passive switch for 3 audio sources to two sets of monitors ( one active and one passive with amp)

audio sources are RME scard(1/4") , Project Mix I/O (1/4") and a Dac(rca)

any help with schematics will be very appreciated!!

[royaldust]

Hello everyone,

instead of starting another thread about DIY monitor selectors, I thought I'd just write here and see if I can manage to build one too.

This is the diagram i did based on the one made by killanotrace.

I just added a stepped attenuator for volume control.

I am pretty sure I forgot to add some smaller resistors before the OUTPUTS, but maybe some of you can help me get this clear.

Also, I take any suggestion as far as brand and models for the switches / attenuator !



edit: sorry... I just realized the image is really small ! Hope it's big enough to understand !

edit 2: I don't know if that matters, but my inputs vary between apogee ensamble and symphony and my outputs are NS10 trough a crown xls 202 and genelec 8040.

[Bill@WelcomeHome]

I used a TKD from Michael Percy Audio, but he may not stock them any longer. You can buy resistors and a switch and wire your own. DACT, Goldpoint, Michael Percy.... I'm sure that there are other vendors. The cheaper you go with the resistors and switches (or stepped attenuator), the less likely the stereo image will stay centered. Wouldn't you need a trimming resistor on each channel to balance the outputs in case of any difference? I'm not a tech, I just know that there are a pair on my unit.

[royaldust]

Originally Posted by Bill@WelcomeHome

I used a TKD from Michael Percy Audio, but he may not stock them any longer. You can buy resistors and a switch and wire your own. DACT, Goldpoint, Michael Percy.... I'm sure that there are other vendors. The cheaper you go with the resistors and switches (or stepped attenuator), the less likely the stereo image will stay centered. Wouldn't you need a trimming resistor on each channel to balance the outputs in case of any difference? I'm not a tech, I just know that there are a pair on my unit.

Yeah, I think there should be a resistor on each output, I'm not a tech either but I'm slowly slowly starting to understand a little bit of this.
I'm hoping someone like Bri or Matt can help me out

[ulysses]

If you want to save money on the stepped attenuator, do it by having fewer positions rather than by using a crappier switch. If you think about it, you should really only have about four positions on your monitor volume control anyway:
99% of the time you should be listening at a calibrated reference level (0VU = -18dBFS = 85dBSPL or some such standard). Then you need a mute (-?), a "quiet" setting (-40dB or so, for checking instrument balance and for answering the phone) and a "rock out" setting (perhaps a momentary switch, maybe 10-15dB louder than your reference level). A continuously-variable monitor volume control encourages constant twiddling with the monitor volume and that makes it impossible to maintain any mental reference at all. A calibrated reference level means you know how loud things are.

[Bill@WelcomeHome]

The problem is though, that at closer to the reference level, you often want finer resolution. This usually translates to more steps.

[Matt Syson]

Hi
I am attracted by Ulysses proposition. To expand it to 11 (12) positions for which switches are readily available would be an interesting way forward, with the 'basic' positions as suggested then using up to 8 finer steps around the 'nominal' reference level.
How about 2dB steps (negotiable at the point of assembly) above and below the nominal?
Ref to previous, you should not be adding extra 'trimmer' resistors to compensate for incorrectly calibrated gear external to the attenuator, which is STILL a bad idea.
Actually a better idea has just struck me to 'solve' many of these issues. A relay switched (mono) attenuator which is placed with MINIMAL cabling (say 150mm or less) which would live behind EACH monitor (amplifier). This can then be a higher impedance reducing loading and the short link would minimise HF loss, and there would be NO crosstalk.
As a 'concept' it could handle 'surround' systems too.
Matt S
PS, if anyone gets around to building this before I get to it, can I have a pair of boards please?

[ulysses]

I don't think I got my point across. Right around the reference level position is exactly where you DON'T want a bunch more steps. It's the reference level for a reason. The whole point is to train your ears to recognize and feel that reference level. If you're monitoring at 0VU=-18dBFS=85dBSPL and you think it's too loud or too quiet, it's because your mix is too loud or too quiet. Adjust the mix, not the monitors. Our ears adapt very readily. If you turn the monitors up or down a couple of dB, it throws off your "internal calibration".

[Matt Syson]

Hi
Yep true, I was not thinking of it in terms of your 'reference' just 'convenience. However the principle is valid in that a few more 'usefull' levels could be added. Off, very chilled, fairly chilled, I like this level, REFERENCE, darn this is good' and 'lets broadcast it to the street' levels. Lastly the infamous, number 11 position (or is it now inflated to 12 these days?).
If anyone has access to a reasonably accurate controller at present, would they care to suggest dB levels with a 'starting' point of 0dB for the 'reference' level, so there would in my scenario be a couple of 'notches' above, then a few below this.
Matt S

[ulysses]

Yes, there you go.
But since he seems to be after a passive attenuator for a stereo pair of balanced lines, he's going to need a 4-pole switch. maybe this Grayhill: 71BDF30-02QAJN (Digikey has it in stock)

[undertone]

I created a 5.1 balanced passive monitor box that I've been using successfully for several years now. It uses 3 fixed calibrated levels (K-system) and a variable one for "casual" non-calibrated use. BTW the following only works if you're feeding balanced outs to balanced inputs and also relies on the accuracy of both.

The way I did the the attenuation was to use a simple 1/2 H-attenuator, with matched 1K (to .01%!) on the balance lines and trimmers across them. The converter impedance is around 20 ohms and the amp in is 10k. My cable run is pretty short (from converter out to attenutor, to amps is max 15 feet). I haven't found any measurable hi-Fq loss.

The variable one uses a 6-channel audio taper I found on eBay. It has to be audio taper.

Over the years I've posted this idea here, on DIY Audio and Prodigy Pro and no one has ever dismissed it as a bad idea. It was the simplest most economical way of building this and I used relays to add muting to all circuits. I grouped them as L/R, Center, Sub and Rears, with a master that mutes them all. I also have a secondary circuit that routes the mains to another amp chain to test stereo mixes. It was a PITA to wire, but it's stood up well now over many years. I posted the concept schematic here. A stereo version would be much simpler; actually 2/3rds simpler.

I second the idea of fixed monitoring levels. You'll find that's the norm in the film industry and it makes so much sense as you're always referencing to the same exact levels. You just have to be careful when you play back commercial music as most of it will naturally be much too loud; what the pot is for...

[timotheee]

Hi all -

I stumbled across this useful thread while planning my own DIY monitor selector and figured I'd share the schematics of the version I will build. Notice I said "will" build - haven't gotten there yet. It'd be great if someone more knowledgeable would chime in and give a "thumbs up" if this schematic looks good to go.

This is intended to take the 90 ohm unbalanced outputs from an Apogee Duet 1 and select between two pairs of monitors which both have input impedances of 10k ohms. Like the OP, I've chosen to implement a mute and mono switch. No need for volume control as the duet's knob works fine for that purpose. Hopefully this is helpful to someone, and thanks in advance for any verification / suggestions of problems & fixes should someone catch this in time before I start tinkering!

Cheers!
Tim