Custom Build Mixing Console

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Joined
Jan 11, 2019
Messages
8
Hi guys,

As I always dreamed about building my own custom mixing desk we decided to start building one.
We started this around a year ago.
For the basic preamp we use the neve 1084 design. We like the sound of this preamp and have good experience with building this preamp.
We then devised a way to make Inline modules based on the 1084 and also added a balanced insert point. There is ofcourse a lot more to it then this.

The power supply is 3phase Full wave rectification.
This way it will have low ripple and the buffer capacitors will have an easy life and will last long.
A neve 1084/1073 wants a good power supply. The concept of this console is also to build something that will last and if something happens the session can continue and it can easily be fixed.
The raw rectified power supply is then split up in 4 seperate lines.
In the console are the smaller power supplies using lm317/lm337. there is one small power supply for each pair of channels. So if something fails the rest will keep on going.

Then the VU meters are set to 8dbu=0VU. Normally this should be 4dbu = 0VU but for some reason today's engineer like to record quite hot.....don't know why you would do that when you have 24bit adc....... but ok. Doing this we avoid that the meters are continuously pegged to the right....

So now some pictures:

first things first and we started out with a small frame prototyping.

test1.jpg


After this was tested and checked, the work started on the frame.

naked frame.jpg


And after some more work:

total frame.jpg


the work also started on the soldering some PCB's....


preamp.jpg


EQ.jpg



Elma Switches for the gain switch

elma.jpg


P&G faders

pgfader.jpg


after the channelfronts have arrived:

channelfronts.jpg


after the Faderpanels have arrived:

faderprints.jpg


The inside of 1 channel module:

inside.jpg


inside2.jpg


the power supply build:

ps1.jpg


ps2.jpg


Fader knobs and sifam meters have arrived

total1.jpg


VU.jpg


faderknobs.jpg


EAO switches have arrived

009.JPG


007.JPG


Subgroup/Master/Monitor and Talkback modules. (There will also be a VU meter module using the IN13 nixie bargraph to display peak or VU)

013.JPG


The monitor section

016.JPG


This is where we are now. It took us around 8 months to get where we are now.
We design are own pcb's, frontpanels are made by schaeffer
Still waiting on the mixbus PCB's and potmeters for the balance and aux/cue's
 
That is a HUGE project. I love the way you have used standard 19 inch rack mounts turned horizontal for the basic module fixings. Very innovative.

Very well done.

Cheers

Ian
 
That's great! Wow..

Did I see right? Do you have a hi-z input on each channel?
 
warpie said:
That's great! Wow..

Did I see right? Do you have a hi-z input on each channel?

No this changes the impedance from the MIC input by wiring the input transformer in series or parallel just like the 1073/1084 neve did.

The line input is a little bit different then a standard 1084.
We used a 40db pad in front of the mic transformer and wired the elma switch for the gain needed.
We did use a line input transformer for the return channel.
It saved us 1 transformer (X32)
The problem was mostly the available space.
1 channel now has already 4 transformers.....

Also by doing this the same connector can be used for mic or line input so no need to patch.
We did make a  switching system so that if you put the input on "LINE INPUT" you can't turn on the phantom
Of course you can still make mistakes when in "MIC" input.....
However the gear we build is mostly transformer coupled and doesn't mind phantom.....
And if I make an output using Line drivers it is made to withstand phantom. (all professional gear should withstand phantom in my opinion except ribbon mikes without stepup transformer...)

The width of the channels is exactly the same as the height of 1U.
We choose this because it's easier to make a frame and we can make standalone versions of the channels if needed.
Change the frontpanel to Horizontal. and it can be used as an inline preamp.
 
:+1 on use of  U19 .  Looking clean and well thought out! :eek:
Really digging how open the frame is as well.  Looks very serviceable.

Mind I ask how much  Front Panel Express put you back?  I hope per-unit costs were reduced for you, given the size of the order.
I'm not sure if I should go with them or a local cnc shop for raw cuts and do laser etching myself.

Oh also, how are you handling logic? Are you avoiding use of backplanes?

Picts are so inspiring. Thanks for the fuel!
 
boji said:
:+1 on use of  U19 .  Looking clean and well thought out! :eek:
Really digging how open the frame is as well.  Looks very serviceable.

Mind I ask how much  Front Panel Express put you back?  I hope per-unit costs were reduced for you, given the size of the order.
I'm not sure if I should go with them or a local cnc shop for raw cuts and do laser etching myself.

Oh also, how are you handling logic? Are you avoiding use of backplanes?

Picts are so inspiring. Thanks for the fuel!

If you get 30+ pieces of one front panel you will get something around 30-40% discount so that helps a lot.
Have to look it up.

For the logic I took it as simple as possible.
There is a mute signal that is active-low so if any of the channels goes "solo" the mute signal is pulled to ground.
when this happens al the other channels will mute unless they are also in "solo".
the switches are momentary and flipflops + schmitt trigger handles the input from the switches.
A relay switches the signal.
The relay and leds are controlled using a ULN2003.
So only one digital signal runs trough the console for the muting.
The logic has a seperate power supply also

Then on the main section the same principle is used for the subgroups

If you  need more info
Just ask ;)
 
ruffrecords said:
How did you do the buses? They were quite something in the originals.

Cheers

Ian

When I was researching this part I found that neve used some kind of voltage summing ("passive summing" as people sometimes call it) in a transformer and then an amplifier providing the voltage gain.

This way of mixing does have drawbacks also. When the source impedance changes, the summing changes.
So if you change the balance control of 1 channel the whole mixing changes. You can get away with this using larger mixing resistors but then noise can be a problem.

Another option is Current summing ("active summing" )
The impedance at the summing amp input is kept as low as possible and the summing amp is a current to voltage amplifier.
now if the source impedance changes the rest of the mix doesn't change.
Problem for the summing amplifier is that the gain changes when you change the mix controls.
So it would be better to use a "Current feedback" amplifier as the frequency response doesn't change that much when the gain is changed.

Another nice thing is the use of a transformer in the mixbus. The mixbus can float and can be balanced as well.

so after some thinking A cae up with the following

The bus is returned to an input transformer. This transformer has 2 input windings. Normally you put these in series or parallel.
This time the bus is returned on only 1 primary winding.
Feedback is taken from the output of the preamp stage using a resistor to the second primary winding.
The 2 primary windings are the same so the gain can be calculated using the feedback resistor and the bus resistors.
The input transformer is also included in the feedback loop so response is quite good.
There is one thing to be carefull with.
Because of phase shift you could have an oscillation and I needed to increase some coupling capacitors.
I will post  some more info about this when it's up and running.
I also wanted to try something different that nobody had tried as far as I know of

I did however keep the bus impedance as constant as possible so if this approach is not optimal I can just use voltage summing.
The balance pot is a relatively low value so that it's not to much loaded from the mixbusses.
Becauseof this the Neve output stage get loaded with around 600ohm like they where meant to be loaded with.

Hope it's a little bit clear as it's not that easy to explain in short.
 
Input transformer method of summing....interesting.  How have you managed DC on signal bus?

There's people around here way smarter than me that might want to better understand or indicate things to consider with your  idea.  Good luck! But like you said, if it does not work, you can always revert to the Neve way.

I've been warned about the cumulative effect of using large bus resistors, but hopefully my low channel count will keep things quiet enough.

Please keep the picts coming, as you progress, if you don't mind, and the testing of the summing.

And thanks for mentioning FPE discount, I'll look into it.
 
Vintage Audio Projects said:
When I was researching this part I found that neve used some kind of voltage summing ("passive summing" as people sometimes call it) in a transformer and then an amplifier providing the voltage gain.

That is basically correct. Voltage/passive summing gets an unwarranted bad press because of the reasons you mentioned (most of which are not an issue). But what I meant with my question was how you built the buses physically. Neve went to a lot of trouble to screen them for example.

Cheers

Ian
 
ruffrecords said:
That is basically correct. Voltage/passive summing gets an unwarranted bad press because of the reasons you mentioned (most of which are not an issue). But what I meant with my question was how you built the buses physically. Neve went to a lot of trouble to screen them for example.

Cheers

Ian
The busses are run on a PCB (8channels per PCB) but the ground seperates each bus as a sort of screening.
However the interconnecting of the PCB's is done using screened cable.
Also the connections from the switches (LR and Subgr switches) are done using screened cable.
There will be NO flat cables in the desk
If we have to use wire to connect signal lines they are screened (only one side so no ground loops)

The reason we used the PCB's is also that we can implement the mute logic on the mixing PCB (seperated from the analog mixing part)

The busses are also balanced/floating so this also helps a lot.
 

Latest posts

Back
Top