Is the DAC calibrated?

No, with one exception: since firmware 4.6.0 the Supernova trims its I3C High-Voltage rail against its own measurement.

Supernova and Pulsar

The Supernova and the Pulsar have no general-purpose analog output. Their DACs set the target voltage rails (VTARG). How close a rail lands to the set voltage depends on the rail:

Rail	Range	How it is set	Measured error
Supernova I3C High-Voltage, firmware 4.6.0 and later	1.2–3.25 V	Closed loop: set, measure under load, correct	−7 to +2 mV across the range, in repeated sweeps on a revision C Supernova
Supernova I3C High-Voltage, firmware 4.5.1 and earlier	1.2–3.3 V	Open loop, fixed factory constants	1200 mV set gave 1075–1094 mV; spread ~120 mV between units
Supernova I3C Low-Voltage	0.8–1.2 V	Open loop, fixed constants	+8 to +11 mV, measured on one Supernova
Multifunction VTARG (both adapters)	1.2–3.3 V	Open loop, fixed constants	Supernova +13 to +18 mV, Pulsar −16 to −9 mV, measured on one adapter of each

“Fixed constants” means the same conversion for every unit, with no per-unit calibration, so other units can land differently. A set on the I3C HV rail takes about 550 ms on firmware 4.6.0 and later.

If the target needs an exact supply, measure VTARG with a multimeter, or power the target from a bench supply and use use_external_voltage().

Note

The rails can source current but not sink it. Lowering the voltage on a lightly loaded rail takes time: the voltage decays only through the load. Read the rail back before relying on the new value.

Nova

The Nova has one DAC output on its signal pins. It is not calibrated. Use it as a near-DC test stimulus, not as a waveform generator.

Related: Is the ADC calibrated?