Integrate with a PandA

A PandA exposes a plain TCP server on port 8888 (control) and port 8889 (data capture). Several client libraries build on this interface:

Approach	When to use
Direct TCP	Low-level scripting; quick one-off checks
PandABlocks-client (Python)	Python applications; preferred for new integrations
fastcs-PandABlocks (EPICS / Tango)	Control-system integrations via EPICS or Tango

Direct TCP¶

Connect to port 8888 to send commands and read responses:

nc <panda-hostname> 8888

See the PandABlocks-server commands reference for the command syntax.

Connect to port 8889 for the data capture stream:

nc <panda-hostname> 8889

Press Return; you will see OK. After arming PCAP (in the web control or via port 8888), data streams until you disarm. See Tutorial 2 for a worked example.

Python client (`PandABlocks-client`)¶

PandABlocks-client provides both a blocking and an asyncio client.

Install:

pip install pandablocks

Blocking client¶

The BlockingClient is the simplest entry point:

from pandablocks.blocking import BlockingClient

with BlockingClient("my-panda") as client:
    responses = client.send(["PCAP.ARM=1"])

Data capture to HDF5 / numpy¶

PandABlocks-client includes helpers to read the binary capture stream and write it to HDF5 or convert it to numpy arrays. See the PandABlocks-client documentation for the PcapHdf5Writer and related utilities.

EPICS and Tango via `fastcs-PandABlocks`¶

fastcs-PandABlocks wraps PandABlocks-client and exposes all PandA attributes as EPICS PVs or Tango attributes via the FastCS framework.

Refer to the fastcs-PandABlocks repository for installation and configuration instructions.