Chroma 6345 electronic load mystery part identification

[KrudyZ]

The part in question is in an SOT23-3 package with ABG printed on it.
The reference designator starts with a Q, so it's unlikely to be an IC.
Its function should be that of a current mirror, like a BCV62 with its emitters connected together and brought out on a single pin.
From what I can gather without a schematics, it is used to mirror a current proportionate to the high side sense resistor voltage drop to the low side.
Does anyone know what it might be?

[MagicSmoker]

Probably a PNP BJT. See the (mercifully short) list here:

https://alltransistors.com/smd-search.php?search=abg

[KrudyZ]

The 2SA1312 would fit package size and marking.
Here is the schematics I backed out. It looks like a normal PNP actually works in this place and it doesn't require a current mirror as I originally thought.
V1 and V2 are floating 5V supplies.
The OPAMP is an Elantec EL2075.
V3 is the voltage of the external UUT, e.g. a 20V power supply.
V4 is the sense voltage seen across the high side sense resistors. There are also low side sense resistors elsewhere, but they are only used for the MOSFET drivers to control the current they pass individually.
R1 is referenced to the low side of the UUT where the power MOSFETs are doing their work loading it.Their are five of these circuits in parallel, each sensing across a separate set of sense resistors that go to different pins of the load connector, but they are all shorted together on the load card.
With no voltage on V4, R1 gets about 128uA. This gets reduced by 27.5uA/mV on V4, so it's running out of head room at about 5mV sense voltage. The actual sense voltage is twice that since it comes in via a 50 Ohm source terminated cable.

This electronic load was meant to be used for load testing VRMs for motherboards and came with customized load cards (which I don't have). It is able to drain 300A, with a slew rate of 150A/us, so it requires very low inductance connections to the UUT.