The input channel was the only prototype made, everything else is one of, including the quite complex master section. Almost everything is modular and on connectors. All pots are handwired and "soft" interfacing is used to connect the desk mechanically together, and this must be the reason that we had zero mechanical problems associated with connectors, switches, motherboards etc. Main signals use multiple paralel connector pins, and switches are double contact, even if they only perform logical switching in most cases.The frame is made of aluminium profiles, except the base is steel. I still remember how our old Soundworkshop desk started to crackle when somebody would lean on the armrest.
40 segment peak meters can monitor 4 different channel sources: source directly, line amp output, equalizer output and post fader signal. The scale is backlighted.
Patchbay is fully normalized. Aux sends are normalized to effects, or they can all can be routed to headphone mixer with a button press, without using patch bay at all. Above patch bay there are 24 microphone preamps with balanced outputs. They are normalized to line 1 input on the right (source) side of the desk, but can be patched as needed. They are provided "just in case" because most of the time we use external mic preams. Output drivers for the whole desk are located in a separate 19" enclosure where they can be swithced to different destinations by relays.

The desk is now 10 years old and it needs very little attention, except for cleaninig. During the first 3 months there were a few component failures, all of the same type. I discovered that a series of blocking capacitors had a failure rate of 0.15%.. Since 6000 of them are built into the desk it means that 10 capacitors died, but each time the damage was limited to a single channel, and usually to a small part of a channel only, so the rest of the desk, including the rest of the channel affected, operated normaly, except there was a bad smell from the burned protection resistor. A cracking pot here an there, a few dead light bulbs, and a couple of IC's is almost all that had to be cared for.

It took me four years to built this desk and everything was done on a computer, including 3D simulations for mechanical design. At that time I had a top end 386 40mHz computer. Larger PCB's had to be designed in two parts because I had only 8 mb of ram.