[rbuskov]

I'm thinking about bulding a line mixer based on the EzTube mixer design, although with a very different configuration.

I would need 16 mono channels with the follow features:

- Balanced line input (with +4 dbu/-10 dBV switch)
- Direct out (post fader)
- 6 mono and 2 stereo aux sends (post fader)
- Pan
- Level (with 100 mm fader and bypass switch for easy unity gain summing)

The master section would be very basic:

- Aux levels (6 mono/2 stereo)
- Stereo mix level
- Balanced outputs for all 12 mix busses

Obviously, I will not be able to use the (otherwise very clever) EzTube module design for this project. Adapting the Eztube circuits and designing PCBs should not be a problem but I'm slightly concerned about the metal work, which can be very time consuming (and expensive) in my experience.

Does this project make sense? Are there other DIY projects I should take a look at before I dive in? Any good ideas for the metal work?

Thanks,
Rasmus

[ruffrecords]

Before I say anything else, just one question - there is no EQ, correct?

Cheers

Ian

[rbuskov]

Before I say anything else, just one question - there is no EQ, correct?

Cheers

Ian

Correct.

[ruffrecords]

OK, then there is no reason not to use the Twin Line Amp card (TLA) as the basis. Each board would be able to handle two inputs. SIgnal flow would be:

10K:10K transformer -? fader -> 1/2 TLA -> (AUXes/PAN and direct out (via transformer)

+4/-10 switch would work by altering the TLA gain

For 16 total inputs you need only 8 TLA boards  plus some for the buses.

How much in hand on the fader do you plan on?

Cheers

Ian

[rbuskov]

How much in hand on the fader do you plan on?

Good stuff... not sure I understand the question though. Are you talking about cost?

[ruffrecords]

Good stuff... not sure I understand the question though. Are you talking about cost?

Sorry, I meant gain in hand.  Like mixer faders typically hove zero marked 10dB from the top so you can add up to 10dB gain.

Cheers

Ian

[rbuskov]

OK

10 dB would work, 15 dB would be better.

[ruffrecords]

OK

10 dB would work, 15 dB would be better.

No problem, just a matter of setting the TLA gain appropriately.

You say you want 6 mono and two stereo AUX sends (all post fader) which is a total of 10buses plus L&R. It is also 10 knobs plus the PAN.

On a single channel will you ever use all 6 mono and both stereo AUX sends at once?

Cheers

Ian

[rbuskov]

No, I would be unlikely to use more than half the auxes on the  same channel at any given time.

[ruffrecords]

No, I would be unlikely to use more than half the auxes on the  same channel at any given time.

So the 6 mono aux sends could be replaced by 3 pots and 3 switches to select one of two buses for each aux.
Let's say w do something similar for the two stereo aux sends converting them into one pot/pan combo and a bus selection switch. Here's a block diagram of the channels:



Cheers

Ian

[rbuskov]

Wow... thanks for your help, Ian. I see how building with your TLA's could simplify this project a lot. I hope you sell PCBs?

Now, the switches on the auxes will obviously reduce cost, but is that the only reason for your suggestion? If so, I would prefer to put a pot for each aux on every channel.

Perhaps I should explain my thoughts around this project...

I produce electronic music and will be using the mixer in the composition/sound design process. I rarely mix more than 4-6 synths at a time, but I like to use full channels for FX returns and use the auxes to "chain" effects. For example, I might send a synth to a flanger on send 1 and to a delay on send 2, and then decide to send the delay return channel to the flanger as well, so the repeats will go through the flanger twice. It's a creative thing and it's important for me to be able to experiment without having to rearrange the FX sends.

It's also worth noting that many of the effects I use are actually guitar pedals, and most of my synths are somewhat noisy, so for the above scenario I'm not overly concerned with noise. I also plan to use the mixer for summing during mix down, and here I obviously do care about noise.

The idea is to add "summing" switches to the channels that bypass faders and pan pots, and sends the signal directly to the stereo buss at unity gain (left buss for odd channels, right buss for even channels).

I would use decent (but not super high end) faders and pots, decent output transformers for direct and aux outputs, and then invest in some REALLY nice transformers for channel inputs and stereo out.

This way, I can have a ridiculous number of aux sends AND high end summing in a single mixer without breaking the bank

(Well, I might break the bank a little, because those hight end transformers are EXPENSIVE).

[ruffrecords]

There is probably not a lot of cost saving because decent switches cost about the same as decent pots but there is a change of loading on the output of the TLA. Basically, the minimum load the TLA can drive is 2400 ohms. The more aux send pots you hang on this output the sooner you run out of drive capability. If you look on the DIY tab of my web site:

http://www.customtubeconsoles.com/diy

and scroll down to the EZRubeMixer folder, in there you will find a sub-folder called MixingAndRouting in which is a pdf file all about  mixing and routing using the TLA and how to calculate the loading on the TLA output in terms of equivalent 50K loads.  The TLA can drive a maximum of 21 equivalent 50K loads without clipping or losing headroom. Using this you can work out that 6 mono aux send plus two (panned) stereo ones , a main bus pan and a lightly loaded direct out represents just over 23 equivalent 50K loads so it is just over what the TLA can comfortably drive.

Using auxes switched in pairs  this total drops to just under 14 equivalent 50K loads and you can afford to drive a few more bridging inputs from the direct out.

Cheers

Ian

[rbuskov]

I see. I had some troble following your math until I read the document but it all makes sense now :-)

Let me think about this for a bit and get back to you.

[rbuskov]

OK, here's an idea: A friend of mine suggested that I remove the direct outs from the individual channels and instead add 8 subgroups with outputs. This way the load will not only be lower but also  more predictable, since I don't have to worry about several bridged inputs on the direct out.

This design would suit my workflow just fine... in fact, it might actually be an improvement now that I think about it.

The channels would then have 3 pan controls and 6 aux sends, which adds up to 22 equivalent 50K loads if I use 50K aux sends. This could be lowered to 19.6 by using 100K aux sends instead, but I wonder if that would really be necessary? In the  document you write that the amp can drive 21 equivalent 50K loads with all sends cranked up to max, which I am dead sure will never happen.

Does that make sense?

[ruffrecords]

Does that make sense?

The design advice I give in the document is conservative. As you say, the chance of all controls being in the heaviest load position at once is small so the probability of actually exceeding 21 x  50K loads is very small. Even then, the TLA will actually drive loads lower than this but just not up to its normal +30dBu.

Do the sub groups replace the original L/R bus?

Cheers

Ian

[rbuskov]

I would want to sum the subgroups to a stereo bus, and I guess I might as well add an option to switch channels directly to the stereo bus rather than a subgroup as well then.

[ruffrecords]

I would want to sum the subgroups to a stereo bus, and I guess I might as well add an option to switch channels directly to the stereo bus rather than a subgroup as well then.

OK so basically it would be 16:8:2 with no directs and an option to route a channel straight to the 2. That seems OK to me. You just need to ensure the routing switch allows you to only route to one pair of  buses at a time.

Cheers

Ian

[rbuskov]

Fantastic

If I understand correctly, I can build the mixer using your TLA2 "across the board", so to speak (i.e. on channels, subgroups, stereo master and auxes)?

Also, I assume that VE mixing would be better than passive?

[ruffrecords]

Fantastic

If I understand correctly, I can build the mixer using your TLA2 "across the board", so to speak (i.e. on channels, subgroups, stereo master and auxes)?

Yes
quote]
Also, I assume that VE mixing would be better than passive?
[/quote]

Debatable but probably yes.

On the panned buses it will improve crosstalk.

On the mono auxes there is no crosstalk to worry about and with 16 sources the interaction between controls will be negligible. If you used passive mixing on just the mono auxes you could save three TLAs (and 1.5A of heater current and about 30 something mA of HT current).

You will need eight TLAs for the channels. eight for the sub-groups (VE mixing), two for the stereo auxes and three for the mono auxes (passive mixing). I make that 21 in total.

Cheers

Ian

[rbuskov]

Thanks, Ian.

I see how this project could be viable and I'm sure that the mixer would sound absolutely wonderful. Still, I'm going to explore some more options before I dive in.

Line DI Based On Reel Card/Cards?

Started by Unbound The Lab

Transformer based Line/DI?

Started by SWAN808 The Lab

6 input line mixer with Capi 2 stage line amp and ACA-Bo

Started by eternal slang Mixers/Monitoring Systems

WTB: Line mixer in UK

Started by untune Black Market