Search the Community
Showing results for tags 'active mixer'.
Hi there, I don't know if something like this has been proposed already but if it hasn't, here's a simple two-transistor virtual-earth summing amplifier for the mixer output of the MB6582. It's suboptimal in many ways because I wanted to keep it simple and cheap (and try to use as much as possible existing components value from the main project, e.g. BC547, 10uF caps, etc). I wanted to make it so that it did not require any modification of the existing hardware to work (including running from available voltages), and I also wanted to keep its PCB footprint small in case someone wants to add it to the MB6582 PCB. I think the schematics is pretty self-explanatory, the general idea here, as with all virtual-earth summing systems, is to basically have 0Vp-p at the summing point to reduce crosstalk between channels. The setup is a two-stage tansimpedance amplifier: T1 turns current at its input into voltage and is buffered from the output via the voltage-follower T2. The bootstrap capacitor C3 improves linearity and reduces bus residual. The existing mix resistors are pictured as RM1-RM8. This circuit has a lot of shortcomings, but since we're not exactly going HiFi here, it should still deliver much better results than the original passive summing system which had some severe crosstalk. It should also be fairly easy to add a volume control stage to that mixer output. Note: this is an inverting stage: the output phase is reversed. I wasn't entirely sure where to set the mark for the gain: the circuit will hard clip at around 4.8Vp-p input, 10Vp-p output. I've done some very basic (read: almost none ) testing with my MB6582 and it seems to do the job. Comments welcome. By the way, speaking of comments, I'd have two suggestions for the MB6582 PCB: Please add thermal relief on the ground plane: soldering the ground pins at safe temperatures, even with a 80W iron, is really not easy without them. While ENIG looks beautiful, it's really not DIY friendly either: HASL is much better in that regard: it has excellent component wetting. ENIG only has the upper hand if you have a lot of SMDs and need perfect surface level, which is not the case here. My 2c ! Cheers, T. mixer summing.pdf