500 series module for DNR LM1894 (single-ended noise reduction)

[kmarksson]

Brand new to the forum, but I've lurked here a while. I've decided I want to move my dbx SNR-1 out of my outboard rack, but still wanted to apply single-ended noise reduction to tape sources via 500 series module. Since nothing like this exists already, I turned to the internet to see if there was an equivalent system that could help me do the job (the dbx 929 is the SNR-1 as a stereo 900 series module, alas). That's where I stumbled onto the National Semiconductor LM1894, the heart of the DNR systems of the 1980s: http://greyghost.mooo.com/dnr/

The datasheet for the LM1894 is fairly straightforward, but being relatively new to designing my own circuits there were a couple of pain points for me in getting this ready for being a 500 series module:

1. The LM1894 is expecting stereo unbalanced input signal(s)
2. It is necessary to drop the +16v power rail down to +8v to drive the LM1894 and LM3915 (metering)

Since the LM1894 sums both L & R to a single control path that feeds the low pass filter, I figured the unit would run fine with the IN+ connector only driving the left signal source (and only returning the left output to the OUT+ connector). In a perfect world, I could force-match the threshold level (managed by the 1k pot) on two separate units by passing control voltage via the LINK pin on the 500 connector for stereo operation—PIN 6 on the LM1894 (gain amp input) seems like it could be relevant for this ???—but my knowledge of electrical engineering is limited, and so instead I decided to add a jumper that passes IN+ between two units, driving the right inputs (PIN 2 on the LM1894) on both and making sure the summed L & R control signals are the same (prevents stereo drift). Using a dented pot then could make sure the threshold level on each stays matched.

For dropping the power down from +16v to +8v, I went with an LM317T DC voltage regulator. Here is where my limited EE knowledge causes problems. I'm wondering if I've done the right math for this section of my schematic and if the resistor cap values are adequate to safely reduce the voltage here (when calculating I determined that R2 should be 6.4k and R1 should be 1k to create the needed ~+8v output). Should the LM317T be connected to a heat sync as well?

I linked OUT-, AGND, IN-, and PGND together since the module is meant to be unbalanced, but I'm also unsure if that's the right/safe approach for handling the ground on this. Will passing the IN+ between two units via the jumper cause any trouble if I don't also chain ground between the units (I assume power ground is shared across all units in a 500 series chassis, but I truly have no idea).

Would LOVE for someone to check my work here, I've attached the schematic for the module and the datasheets for the respective ICs (and the voltage regulator). Any and all insight/advice would be greatly appreciated