Stuff I'm working on:
a) Headphone amplifier with delay, for live-sound use. Handy when the console is 100' away from the FOH stacks. Normally, when you cue up something in the cans, you hear the headphones and the stacks. When you're far away from the stacks, the echo between the 'phones and the PA is annoying. Woody Nuss suggested it.
I've got layouts finished for two versions. One is deluxe: line in on combo XLR-TRS jacks, line out on XLR and TRS, headphone out on TRS, and an output level control. Expensive in small quantities. The less-deluxe version has line in on the combo jack, no line out and no level control. Both versions have a 20-x2 LCD and a rotary encoder for the user interface. Delay time can be set in feet, meters and milliseconds. The byte-wide 512k memories mean about 1.5 seconds of delay ... which oughta be long enough. Rather than putting a 1/4" TRS for headphone input, I put the combo jacks, figuring the console probably has control-room outs on 1/4" or XLR. You can always use an insert cable from the headphone out if necessary.
A Xilinx Coolrunner CPLD and two async SRAMs are the delay engine. There's a few pages of Verilog that implements the delay. It's basically a homework assignment for a student learning Verilog. Not at all difficult. The CPLD talks to the CODEC over I2S. The sample rate is 96 kHz simply because that's what the converters can do. Samples are 24 bits wide. A Silicon Labs 8051 handles the user interface. I use a 16VAC wall wart for power.
It's all too expensive. Who wants to pay more than $100 for this? Besides, once digital consoles take over, it'll be free.
b) I've had this idea for a one-rack-space wireless receiver cue box, basically a big switcher with a headphone jack that lets the wireless wrangler cue up any of 24 (expandable in groups of 24) wireless receivers. Good for troubleshooting, especially when the wireless wrangler is NOT the monitor or FOH engineer.
I use Phoenix connectors for the I/O because I assume that a rack full of wireless receivers is custom-wired at the shop before the rack gets sent out on tour for a year. These connectors use a whole lot less space than 24 XLRs. Anyways, the design is basically done and a layout started. Basically, an SiLabs 8051 runs the user interface, scanning buttons and lighting LEDs and enabling analog switches that select which input(s) will drive the headphone out. The usual sort of mic preamp chips do the inputs, and trimpots are used to set the gain for each channel.
c) I'd like to do a "digital snake" using the TI PGA2500 preamp chips in the stage box, with remote-control gain and a fibre interface. I bet I could do the muxing of the ADC data in an FPGA, and that FPGA drives the fibre PHY directly. I know this is a solved problem ...
Posted by Andy at December 30, 2005 12:33 AM