Codes

NACK reasons

Every NACK reason code by number, the reason a fixture gives when it refuses a request, with what each means at the bench.

NACK reasons
27
Numbered in
E1.20, Table A-17
Version
0.12.0
Runs on
macOS 15 Sequoia or later, Apple silicon or Intel
Needs
An Enttec DMX USB Pro with its DMX/RDM firmware (major version 2), or, on a rig, an Art-Net 4 node with RDM on at its ports
Price
Free

A fixture that refuses an RDM request answers with a NACK and a reason code, numbered in E1.20’s Table A-17 and in the standards that extend it. RDMBench’s wire log shows the reason beside the refusal. Here is every code by number, with what it means at the bench, in RDMBench’s words rather than the standard’s.

  • 0x0000 NR_UNKNOWN_PID: The unit does not have this parameter. The most common refusal and usually harmless: a unit lists what it supports, and anything else should get this answer. It is a fault only for a parameter every responder must answer, and some units say nothing at all instead, which reads as a timeout rather than a refusal.
  • 0x0001 NR_FORMAT_ERROR: The unit heard the request but its data was the wrong length or shape for this parameter. From a well-tested controller it points at the unit’s firmware reading the parameter another way, often an older revision of the standard; from a hand-built request, check the data length first.
  • 0x0002 NR_HARDWARE_FAULT: The unit understood the request, but something inside it kept it from acting. The one refusal about the fixture’s hardware rather than the conversation: read its status messages and sensors, and try the same change from its own panel to see whether that fails too.
  • 0x0003 NR_PROXY_REJECT: Something between the controller and the unit, a node, splitter or proxy that answers RDM for the units behind it, turned the request away, and the unit may never have seen it. Put the unit on a direct line before reading anything into it.
  • 0x0004 NR_WRITE_PROTECT: The setting can normally be changed, but not right now. At the bench this is usually a lock you can see: a panel or menu lock, a password, or a setting held by the current mode or by a standalone program the unit is running. Unlock it on the fixture and try again.
  • 0x0005 NR_UNSUPPORTED_COMMAND_CLASS: The unit has this parameter, but not for this kind of request: most often one that can be read and not set, or set and not read. Not a fault; on some models the setting is changed only from the panel.
  • 0x0006 NR_DATA_OUT_OF_RANGE: The unit understood the request, but will not take that value: an address past what its footprint leaves, a personality it does not have, a sensor number past its count. Check the value against what the unit itself reports, not the manual, which may be for other firmware.
  • 0x0007 NR_BUFFER_FULL: The unit’s queue was full and it could not take the request. Usually brief: wait and ask again. If it lasts with nothing else on the line, the unit is not emptying its queue, which is its firmware.
  • 0x0008 NR_PACKET_SIZE_UNSUPPORTED: The request was bigger than the unit can hold. A controller should keep within what the unit takes; on an ordinary request it is the unit’s limit, not the cable, and worth a note to the maker.
  • 0x0009 NR_SUB_DEVICE_OUT_OF_RANGE: The request named a sub-device the unit does not have. Sub-devices are separately addressed parts of one unit, such as the modules of a dimmer rack; most fixtures have none. If the unit’s own sub-device count says it has that one, it contradicts itself.
  • 0x000A NR_PROXY_BUFFER_FULL: A proxy between the controller and the units has no room left for the replies and status messages it holds for them. The units behind it may be fine: ask more slowly, or take the unit off the proxy.
  • 0x000B NR_ACTION_NOT_SUPPORTED: The value was acceptable, but the unit cannot make the change as it is set up now. Look for another setting that has to change first, such as a mode the requested one depends on.
  • 0x000C NR_ENDPOINT_NUMBER_INVALID: A gateway term: the request named an endpoint, one of a gateway’s ports or universes, that it does not have. A fixture on a DMX cable sends this only through a gateway; check the port number asked for.
  • 0x000D NR_INVALID_ENDPOINT_MODE: A gateway term: the port is in a mode that does not allow the request, such as an input where the request needs an output. Change the port’s direction or mode on the gateway first.
  • 0x000E NR_UNKNOWN_UID: The device does not know the UID the request asked about, usually a gateway or proxy asked about a unit it has not discovered. Run discovery through it again.
  • 0x000F NR_UNKNOWN_SCOPE: An RDMnet term: the device is not in the scope the request named. It matters only on an RDMnet network, never on a DMX cable.
  • 0x0010 NR_INVALID_STATIC_CONFIG_TYPE: An RDMnet term: the device does not take the configuration type in the request. It matters only on an RDMnet network, never on a DMX cable.
  • 0x0011 NR_INVALID_IPv4_ADDRESS: Network setup: the device would not take the IPv4 address in the request, a broadcast or reserved address, say. Check the address, mask and gateway together.
  • 0x0012 NR_INVALID_IPv6_ADDRESS: Network setup: the device would not take the IPv6 address in the request. Check the address and prefix together.
  • 0x0013 NR_INVALID_PORT: Network setup: the device would not take the port number in the request. Port 0, and ports the device keeps for itself, are the usual cause.
  • 0x0014 NR_DEVICE_ABSENT: The sub-device or sensor is in the unit’s numbering but is not there now: a module unplugged, a sensor board missing or not detected. Worth chasing at the bench, since it can be a loose connector inside the unit.
  • 0x0015 NR_SENSOR_OUT_OF_RANGE: The request named a sensor past the unit’s sensor count. If the count the unit reports says that sensor should be there, it contradicts itself.
  • 0x0016 NR_SENSOR_FAULT: The unit says the sensor itself is faulty and gave no reading. A real lead at the bench: look at the sensor, its wiring and its connector before anything the sensor watches.
  • 0x0017 NR_PACKING_NOT_SUPPORTED: The request bundled several values into one packed message, which this parameter cannot carry. The controller’s concern rather than the fixture’s: ask one at a time.
  • 0x0018 NR_ERROR_IN_PACKED_LIST_TRANSACTION: Of several values sent in one packed message, at least one failed. The rest may or may not have taken, so read each back on its own.
  • 0x0019 NR_PROXY_DROP: A proxy passed the request on and lost the unit’s reply. The unit may have acted on it: read the setting back directly before assuming either way.
  • 0x0020 NR_ALL_CALL_SET_FAIL: A change sent to every sub-device at once failed on at least one of them. Read each sub-device back to find which; the others may have changed.