My 5.1 passive balanced attenuator/muter

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

undertone

Member
Joined
Jan 29, 2007
Messages
10
Hi all. I'm posting this for a number of reasons. First off, I'm not an engineer, just a handy musician with enough knowledge to design and assemble simpler analog circuitry. I built this particular unit because I couldn't find anything that would do this at a price I was ready to pay. It works for me but I'd like a professional's assessment on it's inherent design flaws, as I may have overlooked some issues. I've also noted people on different forums posting requests for similar circuits, so for the benefit of all, here it is. Perhaps some will come up with more clever ways of optimizing the design... or shoot it down completely! BTW: I want to thank this community for providing a valuable resource of shared information. There's a lot of very generous knowledgeable people here. Many many thanks!

The circuit is a 5.1 passive balanced attenuator and muter whose design spec was:

-To be used between a sound card whose balanced outs are stable even if shorted, and the balanced input stage of active monitors. In my case, the card is an RME Multiface and 5 Dynaudio BM6A active monitors + sub.
-Must provide muting to each group (L/R, C, S, rL/rR) as well a master mute for all.
-Must provide 3 static switch selectable attenuation levels: -20 dB-FS = 83 dB-SPL / -14 dB-FS = 83 dB-SPL / -12 dB-FS = 83 dB-SPL. These are K-meter levels I'm used to working with. See this page for details.
-Each calibrated level must provide adjustment for each of the channels through trimmers (6*3 = 18 trimmers)
-Must also provide an additional switchable setting with pot-controlled variable gain for each group, for uncalibrated casual setting.
-The attenuation network must preserve CMRR as much as possible and not attenuate Fq within the audio band.
-Must offer a switchable setting that will mute the L/R circuit and redirect it to an alternate amp/speaker setup for comparative listening through alternate speakers.

So after much deliberation and experimenting I came up with this.

I found surplus 8PDT pushbutton switches that made the design of it simpler and allowed me to use the additional poles for adding LED indicators to the switch functions. All the mutes use surplus 12VDC audio relays and "mickey-mouse" diode logic for the muting and master muting (not shown). I used point-to-point board wiring, which I will never do again for something like this (!).

I based the attenuator design on this circuit because it wast the only one that would give me both variable gain and consistent CMRR. I matched the series resistors to .1%, and chose 20-turn trimmers appropriate for each section's level reduction. The mutes just connect each side of the balanced lines. I used log pots for the variable section (linear will not work here). The 2 switches used to bypass all the in-series 1k resistors are there to take the entire box out of circuit if desired. I used Tascam DB-25 wiring to create a harness connecting the box to my sound card, eliminating useless connectors. A simple differential amp converts the balanced line to an unbalanced out for my alternate speaker rig (not shown).

The attenuator network is what I'd really like some feedback on. The Multiface has an output impedance of 50 ohms while the impedance of the BM6A was given to me as "10k DC and 20k AC", so it seems to fit the requirements of the attenuator network. I settled on the attenuator network values and trimmers based on tests I ran using appropriate test signals and my trusty RS SPL meter. My max cable run is about 5 meters end-to-end and the wiring is standard 22 AWG twisted-pair foil-shielded studio interconnect, but I don't have tech specs for it. My hearing isn't the best [tapers off after 16k) and I don't detect any perceptible high-frequency loss with the box in or out. I've looked at the calc required to determine such loss and this circuit seems to pass, but I'd appreciate some feedback on this especially on worse case behavior.

Some additional notes:

-There is no mechanical linkage between the 4 level switches, so I have to insure only one is in for accurate levels (that's what I put the indicator LEDs in for).
-I use the calibrated levels for mixing, but find myself using the variable when I'm composing or working on an arrangement.
-The variable levels aren't linked. I would have preferred a single 6-way log pot, but I couldn't find a good cheap one.
-The mutes work really well: I never hear a pop when using them.

Comments welcome!
 
Back
Top