Hopefully this gets the attention of a few resident experts on recording and PA stuff.
As a bit of an electronics geek, I've been building my own 8-channel mic preamp. A first cut seems reasonably good, but I've realised two things. First thing is that I can do a better job, knowing what I know now. And the second thing is that this is potentially something I could sell!
I've had a bit of a look into small production runs on cases and PCBs, and this looks like it could be workable. Before I commit myself to anything unworkable though, please can some resident experts check my assumptions on what features I should be including?
Main features:-
* 8 channels. * XLR in, TRS balanced/unbalanced out. Gold-plated connections, naturally. Incidentally, would anyone consider those dual XLR/TRS jacks worth having, or should I save a few quid by just using regular XLRs? * 12-way switchable gain. 0dB (in case of high-output DI use), 20dB, then up to 80dB in 6dB steps. More than 12 ways on a switch makes them incredibly expensive, sadly. Pots could be used, but (a) you can't easily match stereo channels, and (b) the preamp characteristics mean that most pots aren't accurate enough at the ends of their travel. * 48V phantom power, individually switchable for each input. * 3K3 input impedance. Will this be OK for most mics? (Checking on the web suggests it would be, but maybe someone with more experience could say for sure.) Also, would switchable input impedance be worth having, say between 3K3 and 150R or so? * Basic level indication LED - green for signal present, yellow for warning, red for overload. * +4dBu/-10dBV switch - doesn't change outputs, but alters warning/overload warning points. * Switchable warning threshold for setting desired headroom (-6dB or -20dB) depending on application.
Other design features:-
* First off, no EQ onboard. 8 channels doesn't give us enough front-panel space anyway! But my thinking is that a preamp should be entirely transparent, or as near to as possible, so if you want EQ then you do it downstream afterwards, because you can't get it back when you've taken it out. * Which leads me to transparency. My measure is THD+N. "Golden ears" ain't in it - I don't want "warm" sounds or anything like that. I'm a believer in Douglas Self's comment that adding distortion for "niceness" should come from a "niceness knob" somewhere, and isn't a desirable characteristic of the system as a whole. Once again, you can't get it back when you've taken it out. Sadly I don't have a meter for measuring THD+N - I plan on borrowing/hiring one later to get an accurate measure, but in the meantime I'm designing everything to get this as good as I can. So we're looking at op-amps, not tubes. * True balanced design - two balanced paths from input to output. * No electrolytic capacitors in the signal path. Polarised electrolytic caps are OK but you can't guarantee the polarity of input voltages, and bipolar electrolytics are dodgy, so it'll be polyester caps for input DC blocking. Outputs will use DC servos to remove any offset that creeps through. * No audio transformers - if capacitors are bad for linearity, transformers are worse! * Everything on one PCB. Trailing wires are the enemy when it comes to noise and reliability. * Toroidal transformer(s) for mains input, so no noise from that.
Cost-wise, I'm aiming for around £250-300. Maybe start off lower and see if they sell, and then check how long it takes me to put them together and cost it from there.