DIY mix-buss?

[JohnDavisNYC]

just wondering if anyone knows about the feasability of this... i was thinking it would be great to be able to build my own 16 channel mix buss... kinda like the dangerous music one, but with less features... just 16 channels (8 L/R pairs) into a stereo output, no volume, no pan, no master fader. possibly rip off the design of a classic console buss or something (neve, api, etc). any thoughts? is this a completely crazy and imposible DIY concept?

john

[Scodiddly]

It's dead simple, actually. All you need is a reasonable buss amplifier, which can be done with an op-amp or two. Or a FET, or a tube... simpler than a mic preamp, most likely. Especially if all the inputs are fed from the same sort of source (DA ouput?) and you don't need any volume/pan controls. For that situation it's just a buss amp for each buss and a handful of resistors.

[EasyGo]

Maybe check out the Scott Hampton JFET Mic Pre article a couple issues ago. He has the circuit for the 'gain module' and shows you different applications for using it, one of which is a circuit for a line mixer with a transformer.

[JohnDavisNYC]

hmmm... sounds cool. i'll check out that DIY article... anyone know how i might be able to copy the mix buss of a neve 80 series? what sort of buss amp do they use? i just want stereo (L/R out) so would that technically be two mono busses?

cheers,
john

[soundguy]

Ive been working on a mixer for a LONG time. Dont expect to do it in a weekend.

To my knowledge, the 80 series used a 1272 for the gain make up on the mix out. The buss was passive and it was shielded in this aluminum extruded tube. The thing about those consoles which is or isnt unique depending on what you are comparing it to but certainly worth considering is that the bus was mic level which is why you seee a 10468 on the front of a 1272. As with all neve stuff that Ive seen there is documentation ad naseum on how the buss was constructed but where you place your bussing resistors is really going to depend quite a bit on the enclosure that you put it in and all that.

You also need to decide wether you want to do an active or passive buss, each has its own advantage and disadvantage.

NY Dave has been a treasure trove of info on maintaining constant impedance which is critical if you go with a passive buss. do a search, you'll turn up lots of info.

so yes you are talking two mono busses. And btw, everyone seems to think the dangerous is passive and its not, its completely active with opa's for buffering.

Roll music is making a passive buss box that has a mic level out which, if you have some pocket change, would be a lot easier instead of having to fuck around learning how to do it if that isnt so important to you. All you do is plug it into your mic pre and go.

silly me, I just got a real console, I should have just gotten one of those.

dave

[jspartz]

What are the pro's and con's of passive vs. active? I have been thinking about this type of a project myself. It would be a great article for TapeOp!

Jason

[leigh]

Roll music is making a passive buss box that has a mic level out which, if you have some pocket change, would be a lot easier instead of having to fuck around learning how to do it if that isnt so important to you. All you do is plug it into your mic pre and go.

I checked that out online, they call it the RMS216 Folcrom. Pretty straightforward idea.

The only confusing thing in their description is saying it uses a "standard 8-channel DB25 connector" on each channel. Why? Why not use XLR or balanced 1/4"?

Leigh

[JohnDavisNYC]

wow, thanks soundguy! the reason i wanted to make it myself, was one, to understand what was going on in there... also, i'd like to design it if possible to have some character something that can saturate nicely if i'm mixing rock or something, not sure if the Rolls box would have a particular character... i'll search for the info from NY Dave about impedence, as well. what do you recon are the respective characteristics of active vs. passive?

i basically just want the output of this box to go into a monitor selector/router for assigning to a mixdown deck, CDR, etc. if i don't want to use a mic pre for makeup gain, am i better off with an active buss?

cheers,
john

[soundguy]

oh man, just go do a search...

I think that for what I have come to understand whatever pros and cons you read about can be applied to a 48 channel bussing arrangement, but when you are talking about 16 channels, I dont know...

for one, an active buss uses at least one more opamp per channel. This can be either a plus or minus depending upon what your school of thought is in terms of noise and color and of course how much of an issue you want to make out of constant impedance on a passive bus (which for the featureless design originally sugggested shouldnt in theory amount to much of an issue). You can get all fancy on an active bus and wire in a bunch of different summing opamps on switch that are not present in a passive design, so thats one advantage if you want to take the bus to the end of the line in tweakerville. Of course then you have the noise of an additional opamp which would have to be less than the thermal noise of the 16 bussing resistors and the noise of the final gain make up amps in order to really become a good argument IMO. I think for a 16 channel thing its not so much a big deal, but thats my opinion and Im an ass, so... Go read fred forsell's take on his website, he wrote a very informative article which you might have to stare at for a few days before it makes sense if this is the first time you've thought about this stuff. I had worn out the paper I printed that thing on before I halfway got it...

dave

[JohnDavisNYC]

just checked that Roll box... looks kinda cool... but 1000usd for a passive box... plus a good pre... starting to get a bit pricey... is it going to cost close to that much to build one? will i save much my doing it myself?

j

[JohnDavisNYC]

i'll go do a search... thanks for your help!

cheers,
john

[soundguy]

why dont you want to use a mic pre for makeup gain?

All a mix bus is in the passive case, is a bunch of resistors which wind up with an insertion loss value depending on how many channels you have feeding the buss. This means that for every channel you add, the level gets lower, so you need more makeup gain to get it back to line level. Keep in mind this is a)basic terms coming from b)a guy that halfway understands it, but I figured Id attempt to answer your question anyway.

In the case of an 80 series neve, it IS using a mic pre to get it to line level (despite what you will read about the 1272 not being a mic pre fantasy voodoo).

The only difference between a mic pre and a line amp is the input impedance of the amplifier and the way the input transformer is loaded to perform effficiently in the circuit. So, in the case of neve, the only difference between a 1272 and a 1271 is the input transformer, its the same amplifier and the same ouput amplifier (which are both found in a 1073 with that same amplifier and output transformer as well as one additional gain stage for above 40 something dB, but the 1272 isnt a mic pre, please...) In the case of api, you had 325's working as the line amp which was transformerless input and then the same circuit more or less with a 1:8 transformer on the input and its now a 312 and its a mic pre. There are many other examples of this. You build an amplifier so it can drive a output transformer to provide you with line level. then you chose a transformer for the input to handle the level and impedance that you need to amplify with the amp you just designed. Its more complicated than this, but thats the building block idea.

A mix buss will not color the sound so much as the amplifier that you chose to amplify your buss level to line level. Regardless of which design you chose, you will ALWAYS need an amplifier to be the last step in that chain. With that folcrom box, its just the box with the resistors, you need to supply the amplifiers. If you think $800 is too much to pay for a handful of resistors in a box, go build it yourself. So far Ive been figuring out how to do it for more months than I care to comment on, so if your time is worth anything, its worth it to pay someone for their research into figuring it out and putting it together for you and making sure it works and promoting the product name so you can resell it later and get some money back (good luck selling your DIY for the cost of even parts). different conversation.

Anyhow, what you need to do is figure out a buss, figure out the amp (your color here) and then the matrix after that. Its simple, but you have to experiment to get it right and if you add panpots and shit like that, the placement of the resistors in the box gets pretty critical and you really need to swap out values depending on how many channels, everything that I have tried with my bussing experiements effects everything else, its all connected, so it can take a while to do it right, and when you pick a different amp to feed the bus, you have to start from scratch...

If you are gonna do this though, start thinking in terms of amplifier and impedance and forget the term mic pre and all that. Its all just amplifiers, transformers and impedance.

dave

[NewYorkDave]

Hey gang, I guess this is why my ears were ringing. John, yours sounds like a simple design challenge. Essentially, you're looking for two 8-channel summing networks in a box. Do you plan to use this with, say, a pair of your favorite mic preamps for the make-up gain?

Since impedance DOES play a big part in the design of a passive mixing buss, my first question is: what do you intend to put INTO the mixer (I'm assuming it's the outputs of your D/A converter) and what impedance is said source rated to drive? If your source can drive a low impedance without degradation in performance, that simplifies the design considerably. If you're not sure, post the model name and number and I can probably look up the specs.

[Scodiddly]

For the simplest mix buss I'd recommend an op-amp.

Basically you set up the simplest inverting amplifier circuit, say 10K resistor on the input and on the feedback loop. The actual inverting in becomes a "virtual ground"... which means you can just add more inputs with more 10K resistors, and there isn't really an issue with the number of inputs affecting the overall gain.


http://www.allaboutcircuits.com/vol_3/chpt_8/8.html
has some examples of opamp mixer circuits. The last diagram is what I'm referring to.

[NewYorkDave]

"Virtual ground"... well yes, kind of. Real op-amps are not as well-behaved as the textbooks would have you believe. No op-amp has unlimited open-loop gain, and the gain falls with frequency. Also, the number of inputs DOES affect the gain at which the op-amp is running; or rather, the gain the op-amp needs to drive its inverting input to 0 volts. It's useful to consider an op-amp summer as a simple inverting stage with a feedback resistor and an input resistor composed of all the input resistors in parallel. That makes it easy to figure the actual gain required of the summing stage in order to achieve its "unity gain" output.

The op-amp summer works well, but is not without its perils (stability can be an issue). Beware the tendency of electronics tutorials to present the op-amp as a perfectly-behaved and predictable gain block. The ideal op-amp may be, but the ideal has not been achieved yet in real silicon.

The attraction of the passive mixing buss, besides its lack of the currently unfashionable (deservedly or not) op-amp active summer, is that it allows the user to employ any number of devices for the make-up gain and add the desired "flavor" that said device can impart.

I just signed up for a cheap file hosting account; and when I have it up and running, I'll post a simple 8-channel stereo summer that will NOT cost anywhere near $800 to make unless you gold plate all the parts