1. Power
Choose a PoE source or the specified external supply.
Knowledge · GigE cameras
Your GigE camera is connected, but the software cannot find it or no image arrives. Check power first, then IP addresses and acquisition settings. This guide takes you from a wired connection to your first saved image in Hikrobot MVS.
1. Power
Choose a PoE source or the specified external supply.
2. Connection
Check link speed and assign compatible camera and PC addresses.
3. First image
Start continuous acquisition in MVS, then test the production trigger.
A PoE injector is optional. GigE carries the images; power comes from a compatible PoE source or a separate power connection.
Camera → PoE switch → PCThe switch supplies power and forwards images. Check the PoE standard, per-port power, total power budget and uplink speed.
Camera → vision PC PoE portA suitable VC5000 or VC2000 variant supplies power through its camera port. Check the exact configuration.
Camera → injector → PCAn injector adds power when the network port does not supply PoE. The injector needs its own suitable supply.
Camera → Ethernet → PCUse the specified power connection when powering the camera separately. Check voltage, connector and pin assignment.
Record the exact camera model, its power requirements and the Ethernet speed it supports. A standard PC network port will usually not power a camera. Power over Ethernet (PoE) carries power and data through one cable, provided the camera and the supplying port or injector are compatible. Check the per-port allowance and the switch's total power budget.
If suitable PoE support is unavailable, use the camera's specified external supply and connector. Select cabling for the link speed, length and installation. Repeated movement in a cable carrier requires a cable designed for that duty. Once connected, check the negotiated link speed reported by the network adapter.
For the first test, connect one camera to a dedicated PC network port. An additional adapter may be needed if the existing port serves the company network. USB Ethernet adapters also depend on the USB connection and their driver, so validate them under the intended load.
Assign different IP addresses to the camera and its PC port within the same subnet. The subnet mask defines which addresses belong to that local network. The example below is for an isolated test setup.
Example for an isolated camera link: PC 192.168.10.1, camera 192.168.10.2, subnet mask 255.255.255.0. First check that the PC is not already using this subnet.
An isolated camera link does not need a default gateway. Keep the company-network gateway on the appropriate company-network adapter. With several directly connected cameras, give each separate adapter and its camera a different subnet to keep routing unambiguous.
Automatic addressing is another option. DHCP obtains an address from a server; supported link-local configurations can establish local addresses without one. Check the addressing method at both ends. A temporary camera IP assignment may be lost after a power cycle. Document the final addressing method and verify that it survives a restart.
Install an MVS release and driver supported by the operating system and camera. Hikrobot describes MVS as a tool for discovering devices, changing settings and acquiring and saving images. Use the official download page and the documentation for that release; available controls can vary.
Find the camera in the device list and match its serial number to the physical label. Use the camera software's network configuration function to correct its IP address if necessary, then open the device. If access is denied because it is already controlled elsewhere, close the other acquisition application.
Select a modest frame rate and a supported pixel format. For this standalone test, disable external triggering and select continuous acquisition, often called free run. Start acquisition. If frames arrive but appear black, check the lens cap, aperture, lighting and exposure time. Save an image and reopen it before moving on to the machine's trigger configuration.
If the camera is absent, start with power, link indicators, the cable and the selected network adapter. Then inspect addressing and the software installation. Discovery alone does not establish that the image stream can reach the host.
If the device is listed but cannot be opened, check the subnet and whether another application holds control. If it opens without supplying frames, confirm that acquisition has started and that the camera is not waiting for a trigger. Check firewall rules for camera traffic on the intended adapter using the driver documentation, rather than permanently disabling protection for the entire PC.
For incomplete or missing frames, temporarily reduce the frame rate and check packet size throughout the connection. A successful lower-data-rate test helps direct investigation towards capacity, buffers or host load. The required production rate still needs its own test.
When uncertain, start with a packet size every device supports, typically 1500 bytes on a standard Ethernet setup. Larger packets, called jumbo frames, can reduce packet-processing work. The camera, adapter and any switch must all pass the selected size. Adapter menus may count frame bytes differently from the camera's packet-size parameter.
Check the hardware before copying a jumbo setting from an example. Change one setting at a time and watch incomplete-frame and packet-resend counters. Additional receive buffering can absorb short bursts, but sustained overload still requires more capacity or a lower data rate.
Run the final resolution and pixel format at the highest required trigger rate while inspection, storage and result communication are active. Agree on a test duration and acceptable missing or incomplete image count. Where every product requires inspection, define how the machine handles a missing image.
Measure trigger-to-image and trigger-to-result time. Include reconnection, power interruption and PC restart tests. Store camera settings, addressing and driver and software versions with the machine configuration so replacements can be commissioned consistently.
Budget for the network adapter, power equipment and cabling as well as commissioning and maintenance. Compare these costs with current manual inspection time and measured losses caused by missed images or stoppages. Payback needs evidence from the actual process.
Consult the GigE manual for power, network configuration, packet size, packet interval and pixel format. Available features depend on the camera model and software release. Check the exact model manual for timing features.
A practical benefit of VC-series industrial PCs is built-in PoE on supported models. One Ethernet cable carries image data and powers the camera, removing the need for a separate PoE injector. Choose the model and power budget to match your cameras.

Industrial vision PC with PoE options for multiple cameras. Compare network ports and PC configurations by variant.
Four PoE ports listed for MV-VC5525-256G70-2T and MV-VC5715-256G70-2T.

Vision controller with four GigE ports. Select a VC2143 variant for integrated PoE; VC2140 variants do not include PoE.
VC2143: 20 W total PoE loading according to the product specification.
Share the camera model, PC and adapter details and required frame rate. For an existing fault, include what the software reports and which settings have already been tested.
Discuss your camera setup