Light,
under control

One non-dimmable constant-voltage supply, one Notrix chassis, and your existing load runs. The supply feeds the dimmer's input; every load lands directly on a channel output. No dimmable driver, and no smart switch per load.

Anywhere serviceable and ventilated: a mechanical room back-board, a service void, or a joinery cupboard. Keep 40 mm clearance above and below the heatsink and avoid mounting directly above the supply.

Because all control lives at the chassis, mounting it within reach saves every future service call from needing a ladder.

Usually. Existing switch legs become permanently-live feeds and the low-voltage control run is pulled through the same conduit. No new boxes, no plaster cuts. This is how the Ottawa side-split in our stories was done in a weekend.

A five-load room is typically an afternoon: mount, terminate, power up, assign channels, set the calibrated ceiling per channel, then record scenes. The setup guide walks the sequence step by step.

Add the connected watts on that chassis and add 20% headroom. A Trek carrying five 42 W runs (210 W) wants a 250 W supply. Each chassis in a multi-unit job gets its own supply sized to its own group.

On the Vort, yes. Its channels are split into independently-fed groups, so part of the chassis can run at 12 V and the rest at 48 V — or any other combination — with a non-dimmable supply per group.

Aurora and Trek run all channels from a single supply voltage. Mixed-voltage jobs on those models take one chassis per voltage.

6 A per channel on every model, with a chassis total of 18 A (Aurora), 30 A (Trek) and 60 A (Vort). Channels can be paralleled for a single heavy run.

Dimmable drivers carry the dimming electronics per fixture, and their price scales with wattage. A plain constant-voltage supply costs roughly a third per watt, and you need one per group rather than one per load — which is where most of the saving on the calculator comes from.

1% and it holds there. The 16-bit engine resolves the bottom of the curve finely enough that a twenty-minute sunset ramp shows no visible stair-stepping.

A per-channel maximum you set at commissioning, so "100%" means the level the space was designed at rather than the driver's absolute limit. It also keeps mismatched fixtures tracking each other across the whole curve.

Yes — the chassis exposes each channel as a dimmable load to a C4 director over the local network, with scenes callable by name. No adapter hardware.

Yes. Channels are assigned by fixture type at commissioning: single-channel white, paired tunable white, or grouped RGB/RGBW. Mixed assignments on one chassis are normal — that's the point of mix-and-match loads.

SNMP is supported for uptime, per-channel state and supply health, so a chassis can sit in the same monitoring dashboard as the rest of the building's kit.

Over HTTP on the local network, with a token generated at commissioning. Endpoints cover channel level, scene recall, schedule state and an event stream for live changes.

Full request and response shapes are in the reference: download the API reference.

Yes — a single write to a channel's level endpoint, expressed as 0–100 percent. Fade time is an optional parameter on the same call.

The chassis accepts continuous streams for live faders; the reference documents the recommended update interval so a controller doesn't outrun the fade engine.

Almost always the supply, not the dimmer: check it is a constant-voltage unit at the fixtures' rated voltage and that total load is under 80% of its rating. Undersized supplies sag under the inrush of a fast fade.

Set each channel's calibrated ceiling against the other rather than trusting nominal output — batch variation in tape is common. Once ceilings match, the whole curve tracks.

Hold the recessed button on the chassis for ten seconds. Channel assignments and scenes clear; firmware stays. Export your configuration first if you want it back.