Mike powers his Apollo Computer Buffer Unit (Patreon)

Here is a short video of the first power up of Mike's AGC Buffer Unit. Mike explained that they abused an MECL logic gate to use it as an analog differential amplifier ("all logic circuits are made of analog parts", as Don Vonada would say).

From Mike's update:

The core of the problem is the way they're abusing the MC306 chips to form differential receivers. This works great if the two inputs are working as intended -- the chip correctly produces an inverted and noninverted output matching the differential input. However, if both inputs are left floating (i.e., I have nothing hooked up to those pins on the 91-pin connector), then of the outputs from the chip both sit at about 670mV -- or logic high in ECL. This completely breaks the logic of the box, which (rightfully so!) assumes that a signal and its inverse are not high at the same time. To the point where it just fails all of its consistency checking across the redundant command inputs and just clocks whatever out to the AGC. Seems like a pretty critical design flaw....
Unfortunately, this means I'm pretty much stuck until I get a proper mate for that big connector.