Monitoring section

Hello,

I'm a total newbie in the audio pro DIY world, my first project was a monitoring section as transparent as possible to really hear what I did in my mixes. For this purpose I made a full passive one with the added functionalities I need in my work.
Running some A/B tests, I found my module to not be so transparent...

Here's the setup :

Outputs of the Digidesign 192I/O to custom monitoring section to Dynaudio BM15A
To run tests, I connect directly the outputs of the monitoring sections to the inputs of the 192

The custom monitoring section include a "DACT" stepped attenuator (10K) and some switches to allow mutes, mono, L-R swapped, and phase reverse listening.

Here's a picture of the module (please don't laugh)


The suspects for me are :

1) Wires nature : I'm not sure to use the best appropriate wires, those are Cu oxygen free teflon isolated, but only one conductor...

2) Wires arrangement : Is it better to run the wires together

3) Ground connections : I'm not sure of what I done, I connect all the Pin 1 of XLRs to their own XLR chassis connector, connect the two ground pins of the attenuator to the pin 1 of the Left input, and because this is unbalance, I connect the pin 3 of the output XLR to their Pin 1.

4) Switches quality : I'm wondering about, because when I shortpass the signal path with the nearest attenuator switch (so I just go through 1 switch opposed to 5 for the normal signal path) the sound seems to loose some definition...

5) Impedance adaptation : I'm conscient that my realisation don't meet the impedance adaptation rules, I thought it was better for transparency to avoid buffer stages, to keep full passive, but maybe I'm wrong...

I apologise for my english which is about as good as my electronic knowledge, I hope to be clear enough.

Thanks by advance for answers, I hope my problem could be solve, do not be afraid to tell me the truth about my faults, I post here to learn...

Hi poomka,

Welcome to "The Lab"!

Do you have a schematic of your setup?

That would make it easier to discuss.

Could you possibly resize the picture, so it could be viewed on normal monitors? A size of 640x480 works nice in this forum..

Jakob E.

Sorry about picture size...

I'll make a schematic and post it as soon as possible

Thanks for welcome

how does your box sound "not so transparent"? Does it sound really loaded down? If you dont have your impedance right and your system is being picky about it, no amount of grounding and rewiring is likely to change that. I would start there first- making sure that everything in the system is happy driving everything with your new box inline, and then work back from there to make it as "transparent" as you want it to be. It could be that your wiring is all fine and the box sounds great, its just that the high end is a little constricted because the impedance is too far off. Thats one possibility at least, one of many.

dave

Thanks for your answers,
So here's my very first schematic...



Is it clear enough to be understand ?

When I see my schematics for a so simple path, I fall in admiration for the guys who makes real ones...
(Sorrry for the litteral translation)

Could you recomend me a good soft ? I used (and learned) Vectorworks for this one (I'm sure Vectorworks Company won't give me a job as a demonstrater...)

Thanks


About impedances, I'm not sure of the impedances of my box, The DACT attenuator is 10KOhms, should I consider this as the input/output impedances of the box ?
In this case, the 10X between impedances of the box outputs and the speakersamps inputs is not achieved, could it be the source of my problem ?

Looks to me like you need to see a balanced input, but you ignore one leg or the other, choosing either the hot or the cold (dependant on "phase" -or better expressed 'polarity' selected) and don;t terminate the leg that isn't selected.

This may work well enough for some electronically balanced outputs, but transformer balanced outputs will not work with this arrangement... also, other load-dependant cross-coupled balanced outputs will vary in level when connected to this arangement...

Have you tried it? -does it even work properly with your equipment? I suggest that these considerations are rather more important than what type of wire is best (you will most likely not hear a difference between wires!)

Keith

-By the way, your English is better than my... well...anything else!!! :green: :green: :green: :grin: :thumb:

Keith

Hello,

Many thanks for the answers,

I use this box from about 3 months now, because of work, I haven't yet the time to make reals A/B comps until now.

It's hard for me to explain the sonic differences, So here's 2 samples of a song. the "CR" one is recorded via the box, the other one "XLR" is when I directly connect together the XLRs previously connected to the box, samples are 44.1/24bits. The "XLR" one is 6dB louder because of the unbalanced of the box. I feel a lost in definition in the "CR" one, the voices and the background instruments seems farest.

XLR
CR

The reason why I suspect wires and switches quality is because when I run through the long path (Path selector/Polarity selector/L-R reverse/Mutes/Mono selector/attenuator), the sound is different than when I go by the short path (Path selector/attenuator). I could also post A/B of this if you want... Also when I feed the box with a mono signal, inverse polarity of one side, then sum them via Mono switch, the cancelation is not complete...

Thanks again.

Hi,

And welcome to "The Lab"!

Okay, I've been doing a lot of A/B testing recently on some DIY stuff and one of the most important things in comparing two paths is to make sure that the levels are exactly the same, or you will always perceive an audible difference. By "audible difference" I don't mean the level difference but an actual tonal difference.

So if you want to compare two paths and spot any signal degradation- frequency response problems or signal distortion, then you must have both signals at identical levels. When you mentioned that one path was 6dB down due to the balanced>unbalanced signal change losing one leg of the balanced signal, it made me think that maybe this is the case here.

Insert a 6dB pad in the balanced "through" path, or add 6dB of gain to the monitor section (on a side note, I know that either the pad or the gain will further add to the signal path, but please try it before condemning your Monitor Box!)

Don't forget as well that with the -6dB loss through your monitor box, you will have to turn up your monitor amp by 6dB, which loses some signal-to-noise ratio compared to your straight-through connection. So any background noise will also be boosted by 6dB. Any hiss from the power amp, or hum etc will seem louder.

As Keith mentioned above, any signal deterioration caused by the switches and/or wiring will be minimal.

Hope you sort it out,

:thumb:

Mark

Thanks Mark,

In fact , when I A/B test CR.wav against XLR.wav, I pull down of 6dB the fader of XLR.wav in the PTools, which give me the same peak level, is it a correct procedure ?
I guess that maybe to put the fader down changes the sound a little ?
Have you heard the samples ? I'm curious to have advices about, to know if the difference could be only due to the A/B procedure, or if something more problematic does occur...

Thanks again.

If you have a way to NORMALIZE your audio files in software then that may be the easiest way to make sure both levels you are comparing are equal.

cheers & welcome,
kent

sorry guys be I have a real problem with all this so far.

Kent we can talk about normalising later but lets move directly to the project at hand.
" a monitoring section "

You haven't spec'd this monitoring section. I know this may seem a little pedantic but in so many areas of this thing we call audio are an infinite number of adequate solutions and an infinite number of opinions. I say infinite as I may have a different opinion tomorrow ... and so on ....
Monitoring can mean stereo Genelec 3 way active or it could mean a single self powered crap speaker to monitor ... program sound. Monitor doesn't automatically mean that it is good.

So we need to see exactly the requirements of this unit and then decide if it can be achieved with a passive layout.
It looks like this idea grew from a simple level control with mute facility to a unit with phase and mono.

I can see some issues with the phase and mono selections so can we start again and list the equipment we do have.

Digidesign 192I/O outputs balanced
Dynaudio BM15A inputs balanced ??
Attenuator - simple 2 layer potentiometer ( this spells unbalanced )
Independant Left and Right mute switches ??
A larger change-over switch.

Unbalanced can often mean no way of passively making the reverse phase thingy. You have assumed a fully floating differential balanced output from the Dig192 thaty can be treated as two seperate ouputs. One in phase and one reverse phased. I don't know the chips on the outs of these things and you may well be OK but some questions may be worth asking.

Please specify the requirements of this monitoring unit.

Does anyone know where I may be heading here ?

I think I'm following

In the last week, I've stopped myself from posting about a half dozen times, this was one of them.

This box is a problem that lots of us are trying to conquer, and perhaps there's a meta thread in this if we can tackle the different issues such that people can walk away and roll their own.

I've got a similar box built, and am struggling with some of the same issues. Mine is used mainly as a signal splitter for the headphone and monitor feeds at present, but I'd like to add attenuation, mute and mono functionality.

For me, the specs are:
HD192 balanced outputs
reference amp for monitors
reference amp for heaphone mix
would like independent attenuation for both monitor feed and headphone feed
single mute switch for each feed
mono switch for each feed ( or if too difficult, a single mono switch would suffice)

So passive or active?

If passive, how to deal with attenuating the stereo balanced source? (for me I'm a fan of stepped attenuators, how many positions would be adequate? I'm thinking like 4 pol, 12 pos, but where to find? )

If active, I'd assume we want something that won't "color" the sound. Would a straightforward opamp running differential suffice.

Kev, I know the group diy pages have been building lots of information, is this covered somewhere there, or is this another chapter that can be added, once discussion has ensued?

Is this where your headed?

ju

What are the resistor values in the mono summing circuit?

I'm interested!!! I also have a Lynx Two card. I'm monitoring through a mackie right now. :?

Many thanks for all those incomes,

If you have a way to NORMALIZE your audio files in software then that may be the easiest way to make sure both levels you are comparing are equal.

Click to expand...

kent, normalize operation doesn't changes the sound ? Anyway, I'm sure both levels I'm comparing are equals when I adjust them in PTools, nobody yet seems to have download and listen to the samples, I would be really interested to get some advices about the size of the differences...


Kev, I'm not sure about what you mean by spec'd the unit, but I will try to do it.
- The BM15A inputs are balanced
- I was going for unbalanced path because I thought that it will be too complex to keep the signal balanced for mono reduction and sides reverse

My requirements actually are :
- Stereo inputs and outputs
- Steeped attenuator to recall exactly some listening levels
- Mono reduction
- L-R inverse (because as 99.9% of humans I don't ear exactly the same with my two ears)
- Independants L-R mutes before the mono section so it allows to listen only one side on the two speakers
- Phase/Polarity reverse before the mono section to allow phase cancelation of the mono informations

I was thinking to have this working fine before adding more elements, like a Vu meter, a source selector, an output selector, headphone amplifier, etc...

Dale : I will check soon the resistors values, but I don't remember actually, by the way, are resistors needed for those ?

And yes Viitalahde, I will be very interested by tour work.


Thanks again.

Currently there are no references at Group DIY to cover this.

We have had a number of switcher and monitoring units there BUT generally people constantly want different things so I've removed them as they became redundant to the various discussions.

Kev, I'm not sure about what you mean by spec'd the unit, but I will try to do it.

Click to expand...

well things always seem to get added.
some are easy and some just blow the whole project away with something that may seem simple but just isn't.

The more electronics you add the more UN transparent it can become.
Then you need to UP the quality of the active stuff and it all becomes more than just a stepped attenuator and a couple of GOLD switches.

I was thinking to have this working fine before adding more elements, like a Vu meter, a source selector, an output selector, headphone amplifier, etc...

Click to expand...

see what I mean :roll:


back to the point
- Stereo inputs and outputs
no problem
- Steeped attenuator to recall exactly some listening levels
yep ... see Gold Point Switches for ideas
- Mono reduction
could be a problem ... explain to me what Mono reduction is ??
- L-R inverse (because as 99.9% of humans I don't ear exactly the same with my two ears)
stereo flip .. again Gold Point can help , shouldn't be a problem.
- Independants L-R mutes before the mono section so it allows to listen only one side on the two speakers
Left/Right speaker ON/OFF easy ... we are back to the mono bit .... passive ?
- Phase/Polarity reverse before the mono section to allow phase cancelation of the mono informations
When totally balanced then Polarity reverse can be easy given the output from the source and input to the amp is true balanced and there is no tricks we don't know about.

General comments
This sort of stuff can work BUT you expectations need to be established. Going passive may have some consequences with loadings and levels. Driving amp and headphone amp at the same time and then independently may have subtle level changes as you switch.

Level of the mono could be difficult to set up and will involve the balancing of the sources and destinations ... then you go and add a headphone amp.

Something you haven't though about is that the more complex this gets the larger it may grow. Where do you mount it ? It may be possible to do this with relays in a Rack Box and have Remote Control on your desk top.
Just a thought.

I see no problem with starting with a simple attenuator box and then adding the sinmple switches,
like Polarity flip and Left/Right flip and Left/Right - On/Off

Hi!

Fum, you can build modular 4pole, 12 position rotaries using the NSF index mechanism and wafers found at Farnell.

Their website is www.nsfcontrols.co.uk

I'll be using these for a similar thing in the future.....but it gets very complicated very quickly - mine will probably be an active design though with ForssellJFET992s.

Poomka, what value is your attenuator - 10k Ladder??

Regards Tom

Many thanks for answers

Kev, your inputs are really helpfull,

When you talk about Gold Point Switches, you're talking about rotaring switches, aren't you ?
Could you recomend me some good ones ?
I couldn't find some toggles switches specificated to audio, are they uncommon for this use ?
Did someone knows witch ones Summit uses ?

What I call mono reduction is a summation of the Left and Right Channels, with the passive path I actually use, I have a level lost when I listen in mono, this drop is a bit too much, but I prefer it to the usual boost you get when you press mono...

I'm going on the passive way because I thought it would be easier for a newbie to get a correct result, the power supply scare me a little...
I'm conscient that the Vu-meter and headphone outputs require power supply, but I thought they're less criticals paths (I don't listen often with the headphones)

Maybe my actual box could be improved with inputs and/or outputs active buffers ?

Did the inputs and/or outputs impedances varies when I change the level with the attenuator ?

How could I mesure the inputs/outputs impedances with a resistance meter ?

The relay idea is in my head since a few months but one more time my ignorance on this subject take me away from this solution...


Tom, I'm sorry but I don't understand "ladder" (my Harrap's dictionary tells me it's usefull to get on another floor, but I'm not sure in the context...)

Thanks again

[quote author="TomWaterman"]Hi!

Fum, you can build modular 4pole, 12 position rotaries using the NSF index mechanism and wafers found at Farnell.

Their website is www.nsfcontrols.co.uk

[/quote]

I just tried to order some switch parts, but Farnell doesn't ship to the US? Such a bummer! It is so hard to get one off large rotary switches for Neves, etc.

Ian