Relay module with Ethernet, USB and (unexpectedly) RS485 control

Today I’m reviewing a rather interesting and useful device—if it weren’t for a few “buts” that somewhat spoiled my overall impression of what is, generally speaking, a pretty decent piece of hardware.
This is an eight-channel relay module that can be controlled remotely over Ethernet.


It all started when the DFrobot store offered to let me review one of their products of my choice. I picked something that I thought could actually be useful. There were plenty of products to choose from, but most of them are already familiar to many people and aren’t particularly interesting. Besides, most of them aren’t really in my field, since they require programming skills, which I don’t have.

And so I received a rather nice black box containing the whole kit. It looks very presentable.


The package includes:
1. RLY-8 controller
2. Power supply
3. USB cable
4. Ethernet patch cable.


1, 2. The Ethernet and USB cables are nothing special, except perhaps that the USB cable is the proper type often used to connect printers.
3, 4. The power supply is rated at 7.5 V and 1 A. It has a flat-pin plug, and the output connector is the standard 5.5/2.1 mm size.



Since the package contents aren’t particularly interesting, let’s move on to the controller.
Specifications from the store page:
MCU: STM32
External power: 7–23 V DC
PoE power: 44–57 V DC
Maximum switching current of the outputs: 277 AC-10 A/125 VAC-12 A
Number of relays: 8
Network configuration: DHCP/Static IP Address
Control methods: Ethernet/USB
Default IP address: 192.168.1.10 (Port:2000)
Working Temperature: -30–85℃
Size: 145 mm * 90 mm * 40 mm
Weight: 530 g


The module is 145 mm wide, equivalent to 8.5 modules in terms of the dimensions used for modular electrical equipment.


There are two clips on the back for mounting on a DIN rail. All well and good, but the clips are very stiff. Installing the module isn’t particularly difficult, but removing it is another matter.


The device is noticeably shorter than standard modular equipment, so it will sit deep inside an electrical panel. The photo compares it with a standard Schneider residual-current circuit breaker.


All the connectors for power and the devices being switched are located on the long sides of the enclosure. I should point out the proper, high-quality power terminals, of the so-called “elevator” type, where the wire is clamped by a U-shaped mechanism.


The top of the enclosure indicates what each connector is for. It also has indicators for each relay and for power.
There’s also a SYS LED and, next to it, a hole for the factory-reset button.
Unfortunately, the device doesn’t show its current status. There are only indicators for relay activation, and if the relays aren’t on, all you can tell is that power is connected. The SYS LED should indicate this, judging by the logic, but I’m not sure.


Power can be supplied through three inputs:
1. Ethernet, via PoE.
2. The power connector, from 7 to 23 V. It’s odd that they didn’t make it 24 V, the industrial standard for devices like this.
3. USB.

Control can be via either Ethernet or USB, in any power/control combination.


There are two versions of this device:
1. Ethernet+USB - link, price: $59.50
2. Ethernet+RS485 - link, price: $79.90

The first version is the one shown in my review, and honestly, I just don’t understand what the extra $20 is for.


There’s also a version with Ethernet-only control and no enclosure - link, price: $83.00


And two more versions without an enclosure:
1. USB+external power - link, price: $77.50
2. Controlled and powered via USB, link. Price: $83.00

And once again, I can't understand the pricing. In my opinion, these devices aren't all that different from one another, yet the one I'm reviewing is considerably cheaper.


Anyway, let's see what's inside. To do that, remove the four screws holding the cover in place.


Here's something that's both a plus and a minus for the designers.
The plus: they used machine screws rather than self-tapping screws.
The minus: they're ordinary screws, so if you turn the cover over, they'll fall out. It's especially "nice" when that happens while you're installing it in a panel.


My first impression of the board: nice and clean.


It looks like the store commissioned the board and the device as a whole, rather than simply reselling it, since the store's name is printed on the board.


The PoE power converter has galvanic isolation, as well as diode bridges to protect against reverse polarity and a small inductor to reduce interference.


The converter is built around an MP6001 PWM controller MP6001, with a rated power of up to 12.5-15 W.


The power regulator for the separate input uses an MP2307 PWM controller MP2307 from the same manufacturer as the one in the PoE converter.
As far as I can tell, all three power inputs (PoE, the separate connector, and USB) are converted to 5 V. There's a fuse on the 5 V line, primarily to protect the USB power line.


Earlier I complained that the maximum input voltage is 23 V. Unfortunately, that's indeed the limit, imposed by the PWM controller.


Ethernet connectivity is handled by a DP83848VV controller DP83848VV.




Everything else is handled by an STM32F107RCT6 microcontroller STM32F107RCT6 from STMicroelectronics. It's a 32-bit chip based on an ARM Cortex-M3 core, running at 72 MHz. It got me thinking: not all that long ago (by my standards), computers had less powerful processors.




The loads are switched by Hongfa HF3FA relays HF3FA with 5 V coils.


The maximum current is 10 A for the normally closed contacts and 5 A for the normally open contacts.


I was a little surprised to find this chip. It's a TP75176E TP75176E, an RS485 transceiver that isn't listed for this version of the device.
There's an in-circuit programming connector next to it, as far as I can tell.


According to the datasheet, it's a pretty decent transceiver.


What's more, its outputs are routed to the contacts under the USB connector. So the boards aren't just universal in terms of connection configuration; the only difference is which connector is soldered on. You could actually desolder the USB connector, solder on a standard one, and get an RS485 port. What's more, the device covers aren't marked USB or RS485, so they're universal too. Which raises the question: what exactly are you paying an extra $20 for when you buy the RS485 version of the controller?
And that's not all: it turns out the transceiver chip isn't just present on the board—it actually works. I tested it with a USB-RS485 converter.


Since I've turned the board over, I'll mention another drawback. From above, I was pleased with the neat assembly, but when I looked underneath, my hair stood on end...


This isn't just flux that wasn't cleaned off; there are literally mountains of it. It looks like they soldered with ordinary rosin, and used a lot of it. You should smack people over the head with the very soldering iron they used to solder all these connectors and terminal blocks.


That wraps up the inspection; let's move on to the software.
When I connected it via USB, the driver installed automatically, but just in case, there's a link to it on the store page.


Next, launch the application, which is also available for download from the product page. In fact, the link redirects through a couple more pages, but the download works without any problems.
The only snag is that on my main computer running Windows 7, the software takes about a minute to launch! On my Windows 8 tablet, though, it starts almost instantly.


Then click the Com Control button to open the program's main window. Here you can select the required COM port (its number appeared when I first connected the device, and you can also find it in Device Manager) and configure access over the local network, for example by setting a static IP address.


In my case, I just clicked Open Com and connected to the controller.
Control is extremely simple: you can only switch the different relays on and off manually, in any order. When the software restarts, it reads the data from the controller and displays the current state. But if you reboot the controller itself, everything resets to zero, unfortunately...


In the last window, 3 of the 8 relays were selected, and those same relays switched on in the controller.


The controller can be powered from the same USB port, so no power supply is needed in this case. It draws about 100 mA with everything off, and up to 550 mA with all the relays activated.


Network setup is really straightforward. In DHCP mode, everything happens automatically: the software finds the controller and displays a line with its name to make it easier to identify.
If needed, you can set the address and name manually, but keep in mind that the device cannot work through the “cloud.” This is both an advantage and a drawback. The advantage is security; the drawback is less flexibility and the need to configure port forwarding on the router, ideally with a static IP address, although there are workarounds for that too.


The main window here is a little simpler than in COM port mode, but there’s no difference in how you control the device. Let’s try switching on a few relays.


Looks like it works. It’s connected to a switch with proper PoE — link.


As I mentioned, the software is very simple, but the manufacturer provides a full description of the protocol used to “communicate” with the controller, in two versions: the “standard” one


and a JSON protocol that was completely unfamiliar to me. So, besides its main function, this device could also serve as a kind of learning tool for beginner programmers who want to explore this sort of thing.


I found a YouTube video where someone connected this controller to some kind of media server (judging by the software) and configured its operation there. The description is very basic, of course, but it gives you a good idea of how it works.


But my joy was short-lived. I switched on all 8 relays and stepped away for a while; just a couple of minutes later, the controller rebooted.
After investigating, I found that with all 8 relays on, the converter responsible for PoE power overheats. The PWM controller overheats and goes into protection, but then switches back on almost immediately.
According to the thermal camera, the PWM controller's case temperature reaches 125–127 degrees. The same thing happens with seven relays on, but after about 5 minutes.

I strongly suspect that the converter—or, more specifically, its transformer—is incorrectly designed, causing the PWM controller to operate under excessive load, since the controller itself doesn't draw much power. Power works perfectly with all the other inputs; the problem only occurs when powered via PoE.


For comparison, the temperatures on the left are with PoE power and with a regular power supply. I think the difference is pretty obvious.


Video version of the review.


Let's sum things up.
In my opinion, this product is a 50/50 proposition. On the plus side, it has a good design, fairly high-quality components, a wide choice of connection options (Ethernet, USB, RS485), the option to turn the $60 version into a fully featured model that's even better than the one sold for $80, plenty of control channels, and an open protocol. But it also has its drawbacks: the PoE power converter overheats, it doesn't save the last settings, and the included software is primitive.

This controller on AliExpress — link, and the RS485 version — link.


That's all from me. I hope this information was useful and might come in handy for someone.