A couple of questions...

What you are proposing will work but there are some other things to consider.

What is the output impedence of the device you are coming from?

What is the imput impedence of the device you are going to?

Normally you would want an output impedence of 600 ohms or less (most solid state equipment is in the 5 ohms or less catagory without a transformer) and the pot would be 10 K ohms or less and the imput impedence would be 10 K or better for this all to work. You will be unbalancing the balanced out (if there is a balanced out on the equipment and depending on the equipment it may be a transformer balanced out or a pushpull output if no transformer and it may not like to be unbalanced) and you will be going into the balanced input of the piece of equipment that you are going to. Theoretically this should work but their are lots of problems envolved and lots of things that you will have to consider including the length of the cable that goes from the output of the fader to the input of the device you are feeding (it should be as short as possible). There is also hum problems ( make sure the box that has the faders in it is shielded with steel and the grounds are tied to the box). If you want to keep the whole thing balanced there are special pots that you can buy that are balanced but they are expensive. If you just want to reduce the level from one piece of equipment to the next you could build a simple resistive pad that would do what you want it to do.

This seemingly simple "project" can be a lot more complicated than it first appears.

Hope this helps....

I think what you're looking to build is a passive mixer; no gain stages, just attenuation, right? It would let you combine several sources (each independent of each other) to feed a common source, all at the same relative level. And once you've set the levels, you can pretty much leave it alone and simply use the gear as needed?

Hope I'm not putting words into your mouth, but I think that's what you've got there. A few resistors and (I think) a diode per circuit should do the trick, but I don't build circuits anymore...(that was back in the oldddddd days. ) I'm sure you can google it and find one that works quite nicely for you.

The trick to getting what you want out of a custom piece of gear is often a combination of the build-quality, the components, and the circuit design itself. Many off-the-shelf pieces of gear fall short of this for those very reasons. (Mass-quantity production will do that at times, unfortunately, if the quality control is low.) Oh, and making it LOOK sexy is just as big a part of the job as anything else. :twisted:

You're in good company if you attempt something like this. Andrew Lipinski is one person who comes to mind: A world-renowned engineer (a true Tonmeister, as he's called) who was never satisfied with off-the-shelf gear, so he began building his own. He's got an entire company going now, creating some peerless equipment. (http://www.lipinskisound.com) I believe he's working on something similar, but it might not be passive.... I'd like to try out their preamps someday, as well.

Hey, even good, calibrated volume controls are tough to come by; even the most exotic of consumer gear is compromised. There will always be some kind of level or component mis-match, and who knows what it is you're NOT hearing, when you're down to that critical last db or two.

All in all, a very good argument to do everything you possibly can WITHIN your DAW, and bring it out to the analog world in one place ONLY (monitor board output, or CD player out, when it's DONE.)

Joe, you're absolutely right. I am looking for a passive attenuator.

The goal:
I like to remote mount my preamps - place them within 20 feet of the mics themselves. I then run the line-level signal to my mixer, but this is where I feel the biggest compromise takes place. Mixers are designed to do many things decently, but no one thing perfectly.

I feel that, a passive attenuator with line in and line out would be the absolute no-compromise solution to what I'm looking for. In other words - set the gain to a realistic level on the preamp and walk away from it. Then do any necessary gain changes on the Passive Attenuator and then go straight into the A/D.

I'm intrigued by Lipinski's work and will likely contact him regarding this and other items. I tried the web-link though and it doesn't come up. I'll google him.

Of course, Speck makes the LiLo, but it's 10,000 and it's active - I'd like to stay small, passive and a little more affordable.

J...

Hmmm....maybe you need the http thing in front; this works, I just tested it....

http://www.lipinskisound.com

Quite dissapointing! I'm on a government system and this web-address must be blocked at the firewall. Hmmm...what did Mr. Lipinski do to get blocked by big brother? :shock:

He..He.. I'll try it from my home system. Thanks!

Cucco,
I have always liked the old ad slogan "just do it". And it applies here as well. I´ve done a few similar circuits myself, but so far only on unbalanced line level signals. Especially for connecting active speakers, they save lots of money compared to for example "Mackie Big Knob". Plus i always want to hear mono in only one speaker, which is not sort of how most boxes work.

Anyway, your suggestion, if you only want attenuation, it should work in most cases. It might interfere with the sound quality, but that is rather remote. Mind you, I have no hands on experience on transformer outputs, so no guarantees there.

What I would do is exactly like you say:

hot input (input to the box that is) goes to "top end" of the fader. That is the end "nearest" to maximum output on the fader.

hot output is fetched from the fader

cold goes right through

neutral goes both to output and to "bottom end" of fader

The box where the fader is should be carefully shielded. In worst case, normal household aluminium foil, glued to the inside goes a long way towards that. A steel enclosure is of course better.

If for some reasone you want to mix several outputs, things are not quite as simple, but still not unsurmountable.

On the other hand, using 24bit AD, known mics and pres, and known music, I pretty much find that I can set the gain as a hit-and-miss thing. Not really needed this kind of thing. And on yet another note, unless the cables are really long (say 300 feet or above) a typical condensor mic should have no problems. It is of course different with other types of mics. Anyway, part of the fun with this hobby is to try things out.

Gunnar.

You might want to look into DACT controls

http://www.diycable.com/main/product_info.php?cPath=31_71_77&products_id=274

or there home site

http://www.dact.com/

Also http://www.dact.com/html/ct101_datasheet.html

and http://www.shallco.com/frameset.html

Also http://www.suertenich.com/html/audio/stepped_attenuator.html

and http://www.siteswithstyle.com/VoltSecond/12_posistion_shunt/12_Position_Pure_Shunt.html

If you need for info the http://www.dact.com site is the a good one for information.

Hope this helps.

-TOM-

ps DACT's line amps are GREAT. They are transparent and reasonably priced. It may do exactly what you are trying to do.

Best of luck

and

Hello,

I have designed and built several small format specialty mixers for remote recording. I have found that the passive approach can be successful only under certain conditions. First, your mic preamp's circuitry must be designed to happily run a long UNbalanced line. Next, your a/d must be content with receiving an unbalanced signal. Note that you will be loosing 6dB of micpre gain, and another 6dB going into the a/d. Thirdly, the number of signals combined, per channel, should be only three or less.

Circuit losses and interactions start to out-weigh the advantages of a passive circuit when combining greater numbers of signals. In this situation a minimalist, single stage summing amp circuit makes sense and performs quite nicely.

As stated earlier, high quality parts should be used if one is to maintain transparency. To give you an idea: for a four-in, two-out passive device, you would want to spend a minimum of $150 in parts. However, a really esoteric device could easily run you $800.

You probably will have to experiment with grounding on pin 1 and 3 to unbalance your cables properly, depending on how your pre's and a/d react.

Hey Zilla! Thanks for the info. I'm not really wanting to sum any of the inputs - essentially, I want one attenuator per channel thus allowing me to remote mount my preamps and have a local gain structure with which to work. So, combining channels won't be a big issue.

I don't know that I follow the logic behind losing a total of 12 dB, could you explain??

Thanks!

J.

I remember designing and building a 24 position stepped passive attenuator using a dual-ganged Shallco switch. That thing worked flawlessly for over a decade of hard use. Heck, the preamp I designed around it was finally pulled out of service but that attenuator is going nowhere - I could clean the contacts and it would still be awesome.

BTW, if I had to do it again, I'd go with a 32 position switch

If you get into the DIY thing, you need to be on your toes regarding impedances, cables, crosstalk, and the like. There's a reason good stuff costs so much!

Zilla-

Nice to see you here... He's built some pretty awesome sounding gear-including some great summing mixers (and is a very fine engineer in his own right...).

--Ben

-blush-

Cucco, sorry, I misunderstood your implementation. So you propose to place the mic pres within 20' of the mics, then go the distance at line level to your "control room" where you will ride the level as it goes into each of your a/d inputs. No analog summing, is that right? If so, then passive attenuators will work fine if implemented correctly.

If you use single ended, unbalanced pots you will most likely loose 12dB. How? If your mic pre's have solid-state balanced outputs, you will loose the inverted (cold) leg of that circuit going into the pot. Therefore you will loose half of your signal level: -6dB. The signal coming off the wiper of the pot then feeds the a/d. Presuming that the a/d has balanced inputs, your pot will only be feeding the non-inverting (hot) input, with the inverting pin tied to ground. You will only be driving half the circuit, loosing another 6dB.

The way out of this is to implement a balanced, bridged-t attenuator. Stepped attenuators sound better than potentiometers, and allow exact channel matching and repeatable steps. Also they maintain a good degree of circuit balance for CMR. The aforementioned Shallco's would be a good way to go. Recalculate the resistor values for 1dB (or even .5dB) steps. The attenuator impeadance will depend on the mic pre's output drive capabilites, the length of cable it will drive, and the input impeadance of the a/d.

You are likely to spend $140 per channel and many days assembling the 90 resistors onto each switch. Figure another $150 or so for connectors, chassis and misc. hardware. As you can see, this gets very expensive for any sizeable multi-tracking scenario.

Oops. Need to clear up my somewhat erroneous explaination.

Solid state balanced outputs feeding solid-state balanced inputs will loose 6dB when unbalanced (one leg open or shorted to ground). There are some solid state outputs that are designed to float like a transformer and deliver the same signal level when unbalanced. There is usually a penalty for the gain make-up, however, as the 6dB of forward gain is stolen from feedback, which otherwise would be reducing noise, distortion, etc.

The additional 6dB of lose could occur if your passive attenuator configuration is not implemented correctly. The balanced bridged-t attenuator avoids this when its Z is matched to the input Z of your a/d.

Cool.

This helps. Thanks!