OB-215 Digital I/O Module by Novatek-Electro

Another module from Novatek. This time, we're looking at a device for reading data from various types of sensors and using it as part of automation systems, with the option to transfer the data to a PC.
Inspection, teardown, tests, conclusions.

For residents of Ukraine, it may be easier to buy this module on prom.ua: link.
As always, I'll start by providing a link to the Russian version of the site, since the link in the title goes to the Ukrainian site and will most likely be unavailable to people in Russia. In Russia, I found it at one store for 2940 RUB.

In very simple terms, here's what this module is for.
It can read data from a single analog or digital sensor and send it to a PC or another device over the MODBUS protocol via the RS485 interface.
The module also has a built-in relay with changeover contacts. It can be configured to switch when the measured parameter changes, and manual control is also available.

A quick note: the picture is from another site, and the folks there got a little carried away. The device cannot (at least for now) display absolute humidity, level, illuminance, or pressure readings. It can only show them as measured voltage or current, with the corresponding scaling for the actuator.


The manufacturer says the module can be used as:
− a remote DC voltage meter (0 – 10 V);
− a remote DC current meter (0 – 20 mA);
− a remote temperature meter with NTC (10KB), PTC 1000, PT 1000, or D18B20 digital temperature sensors;
− a temperature controller for refrigeration and heating systems;
− a pulse counter that stores the result in memory;
− a “pulse relay” with a switching current of up to 8 A;
− an RS-485 to UART (TTL) interface converter.

The OB-215 provides:
− control of equipment via a relay output with a switching capacity of up to 1.84 kVA;
− monitoring the state (closed / open) of a contact connected to the
“dry contact” input.

If you want the details, I'll do my best to cover them in this review. As always, I'll start with the packaging: nothing unusual here, just a small box with the manufacturer's logo.


The package includes the OB-215 module and a manual with connection and setup instructions.


The manual is quite substantial. It's also available in digital form, and I'll quote from it below. As usual, the manufacturer offers a ten-year warranty on its product.


Device specifications.


The design is conventional, with a standard width of one module, and it's intended for installation on a 35 mm DIN rail.


The module is slightly shorter than a standard circuit breaker, but that doesn't prevent it from fitting in a regular electrical panel.


1. Power and RS485 interface terminals are at the top.
2. Measurement inputs and relay output contacts are at the bottom.
3. As usual, there's a wiring diagram on one of the side panels. However, with so many sensor options, it only shows the connections in general terms. For the details, you'll still need to consult the manual.
4. The front panel has only indicators for power, relay status, and RS485 communication.


There are so many sensor connection options that the manufacturer had to split the wiring diagram into two parts.
Analog and pulse sensors are on the left, digital sensors on the right.


The module's enclosure consists of two halves.


In the past, separating them was never a problem, but this time the manufacturer apparently decided to troll us make the construction more robust and added glue to some of the pins. At least some of them broke.


The construction is pretty typical for Novatek: several boards stacked on top of each other, with neat assembly.


I also had no complaints about the soldering quality, although some components were clearly soldered by hand. But one thing that's definitely missing is a protective coating on the board, which has become almost standard for devices aimed at the industrial market. I'm not talking about fully potting it, but a polyurethane conformal coating would have been quite enough. Even plain old Plastik-70 would have improved the board's resistance to environmental exposure.


1. The lower board has a relay, as well as the power converter's capacitors and inductor.
2. In the middle is the RS485 interface transceiver SN65HVD11
3. The top board has only the status LEDs.
4. Between the middle and top boards you can see the rail-to-rail op amp TS922


1. The whole thing is controlled by an STM32F071C8T6 microcontroller. That's a little unusual, since Novatek more often uses Atmel products. I had to desolder a couple of capacitors to get to the microcontroller; it's not easy to spot at first glance.
2. The NT75 relay is made by Ningbo Forward Relay. The same one is used in the RN-119. It's rated for currents up to 16 A, but Novatek specifies a maximum current of 8 A.
3. The lower board carries the power regulator and the relay control components.
4. There are two power regulators: a TPS5430 switching regulator with an input range of 5.5–36 V, and a linear 3.3 V regulator. Everything here is very generously rated.


1. The device starts up at 7 V, drawing about 28 mA, and shuts down when the voltage drops below 6 V.
2, 3. At 12 V it draws 20 mA, or 58 mA with the relay on.
4, 5. Raise the voltage to 24 V and the current draw drops to 13–34 mA, respectively.
6. Since I have a suitable Mean Well power supply module, I used it for the rest of the tests.

By the way, I think this is the first Novatek module that doesn't run on mains voltage, and it reminded me of a little story. Some time ago, I helped an acquaintance in the US develop a simple device. The spec called for an external power supply. Since there was plenty of room inside the enclosure, I quite reasonably asked why we shouldn't build the power supply in, perhaps mounting it on the same board as the rest of the electronics.
He replied that in that case the device would have to be certified for electrical safety. With an external supply, you can simply buy a certified power supply, and if anything goes wrong, the power supply manufacturer is the one who has to worry, since the device itself runs on a safe voltage.
That's essentially the same situation here. Since it connects to sensors and a PC, it's best to choose the power supply carefully.


Since the module has no controls, all “communication” with it happens either through the software or by sending individual commands over the MODBUS protocol.
So I downloaded and installed the software, and was in for a pleasant surprise. Judging by the look of it, Novatek had decided to redesign the software. It didn't take long to see that this was clearly a step in the right direction.


But first, you need to connect to the module. There's a “Program settings” menu for that, and I was surprised twice more: first, to see that you can save the connection port instead of having to select it again each time; and second, to find a device search button.


Sure, I already knew the port number, but it was nice to click “Search” and almost immediately get a response from the connected device, showing the port and its settings. Then I could click that line, confirm, and have everything set up automatically.
Of course, network device discovery isn't anything new—for example, it's been standard for IP surveillance systems for years. But I'd been waiting so long for Novatek to add this that it was genuinely nice to see.



A quick explanation of what's shown in the program window.
At the top left, you can see the relay output status and the reason it switched off or on. The measured value or pulse counter is also displayed here.
Below are the relay controls: force it on or off, or switch it on for a fixed 200 ms. If the relay was off, activating this function turns it on for 200 ms and then off again. If it was on, it turns off after 200 ms.
At the bottom left are the relay selector (OV-215 or OV-216), connection status, firmware version, operating mode, additional commands (reboot, reset settings, reset counters, and change the counter value), supply voltage, MODBUS ID, and buttons to connect to the device and open the software settings.

On the right are the module's operating mode settings. Below them are buttons for saving or loading previously saved settings, as well as buttons for uploading settings to and downloading them from the device.

A brief explanation here: the software doesn't have the usual "Apply" button. First, you set the required options, then send them to the device. When connected to the device, the software automatically retrieves its settings.

Overall, the software is intuitive, and I don't think anyone will have trouble using it. Credit to the developers, though as always, I'd like to see more. I'll get to that in the conclusions.


If the connection is lost, the device will automatically try to reconnect. Also, if the device isn't connected when you launch the software and click "Connect", the software will connect to it automatically once the device is plugged in.


Available module settings. One thing to note: the operating mode determines which settings are available. For example, in "pulse input" mode you can configure pulse counters, while in "digital sensor" mode you can adjust the temperature calibration and sensor type.


Input and output configuration modes and their associated options.

1. Pulse counter. Works with a "dry contact" or "open collector". The module has an internal pull-up to the power supply, with a voltage of 2.8 V and a current of 0.5 mA.
2. Logic/pulse input. Essentially the same as the previous mode, but pulses aren't counted. The relay can operate in toggle mode or mirror changes to the input signal, with debounce support in either case.
3. Voltage measurement. The device measures input voltage up to 12 V. You can set the relay to trigger when the voltage goes outside the specified limits.
4. Current measurement. The same capabilities as voltage measurement, but based on current, up to 20–25 mA.
5. In voltage and current measurement modes, you can convert the measured value to make it easier to work with the results.
6. Analog temperature sensors. Choose from three types: NTC, PTC1000, and PT1000. You can apply a correction and link relay control to the measured temperature, for example, to implement a thermostat or indicate when the temperature exceeds a set value.
7. Digital sensor. The same capabilities, plus the option to select the sensor type.
8. Choose from five types: 18B20, DHT11, DHT21, DHT22, and BMP180.
9. Interface converter. Here you set the communication parameters for a device connected to the module input via UART; the output converts the signal to RS485.


As an example, here's how to link the relay's trigger thresholds to changes in voltage. A brief explanation: voltage is measured in 0.01 V increments, up to a maximum of 12.00 V. The settings use whole numbers, so 12.00 V is entered as 1200, and 5.67 V as 567.

1–3. The upper threshold is set to 800 (8 V), the lower to 600 (6 V), and the mode is "open contacts".
In this mode, the relay turns off when the voltage rises above 8 V, and turns on when the voltage drops below 6 V. Between 6 and 8 V, it retains its previous state. Put simply, the relay switches on/off at 6/8 V, with 2 V of hysteresis.
4–5. Here I set the same threshold for switching on and off: 8 V. That gives zero hysteresis, so the relay predictably responds to even the slightest change in voltage.
6. After changing the thresholds from 800/800 to 800/798—that is, adding 0.02 V of hysteresis—the operation became more stable.

Once everything is configured, you don't really need the software anymore. The relay will operate on its own, responding to the set thresholds.


Current measurement mode works the same way, but here each setting unit corresponds to 0.01 mA. For example, 798 = 7.98 mA.


The pulse-counting mode is self-explanatory, so I'll just explain the option for converting the pulse count into counting units.
In some cases, one pulse doesn't equal one counting unit. For example, if you connect it to an electricity meter rated at 6400 pulses per 1 kWh, you can set the conversion to 6400 to 1 and see the result in kWh. You can also set it to 64 to 1 and get the result in 0.01 kWh.
Likewise, you can use conversion with any other units if needed.


A natural question is bound to come up: can the counter count without being connected to a computer?
Yes. For example, I applied pulses to the input, then disconnected the computer while continuing to send pulses. When I reconnected the computer, the counter showed the total number of pulses received.


A few measurements.
Early on, I noticed that the module can measure its supply voltage, which is quite useful, but its measurement accuracy is clearly not very good.
For example, with 24 V at the input, it showed 24.9 V. Then I checked across the entire range and found that it displays the correct value only when the input is 12 V. At other voltages, it underreads or overreads by about 0.1 V for every 1 V.

Of course, this feature is more of an auxiliary one, but even so, the measurements are rather crude. I think the developers should take a look at the calibration.


Voltage measurement accuracy over the 0.01–12 V range: measured values in blue, actual values in red.
Across the specified range, I measured an error of no more than 0.5%, compared with the stated 1%.


Before checking current measurement accuracy, I took one additional measurement.
During the test, I noticed the voltage present at the time and found that it takes about 10 V to produce a current of 20 mA. So the module continues to operate over the same voltage measurement range, but with an additional 500 Ω load.


Current measurement accuracy over the 0.01–20 mA range. The stated error was 1%, and my results were about the same when calculated across the full range, though just barely. Still, accuracy at some values seemed rather poor to me; I would have liked it to be better.


I tested temperature measurement accuracy the same way I did in my review of the RPM-416 data logger. I took several resistors with different values, measured their resistance, then used the table to calculate the equivalent temperature for a PT1000 sensor.

The resistor value is shown in green, the calculated temperature in red, and the measured temperature in blue. The stated error is 1 degree; in reality, it was closer to 0.5 degrees.


There’s no point in checking temperature measurement accuracy with digital sensors, since that’s determined by the sensor itself, so I simply checked that they worked. The temperature was room temperature; I warmed the sensor slightly with my hands while connecting it, and the result was quite reasonable.


It’s worth noting that this device is best thought of not so much as a standalone module, but as a device that can work as part of a system with a MODBUS controller, even allowing parameters to be changed “on the fly” by the central controller’s program.

Conclusions.
I tested most of the device’s capabilities, and overall they fit the stated specifications—some with room to spare, others just barely. Either way, I have no complaints here. I also had no particular concerns about the design or the quality of the components used. Some are name-brand, others are more basic (the relay, for example), but there’s plenty of headroom in the specifications, so that’s fine too.

But I’m left with a somewhat subjective impression, and I’ll go into that in more detail here.
First, of course, the price. I understand that this device is aimed at the industrial market, with its own quirks and requirements, and that $40 for this kind of functionality, build quality, and especially the warranty is considered quite reasonable.
But the market for this kind of “home” device has been growing rapidly lately. People are building “smart homes” and want to monitor everything, and in those situations the price matters more. In my opinion, the hardware doesn’t justify the asking price, and ideally the price should either be cut by a factor of 1.5–2 or the functionality should be expanded. It’s a niche device, so the problem is that to lower the price, you need to increase sales, and to increase sales, you need to lower the price. Something like that.

Now, about the functionality and circuit design.
The software was a pleasant surprise: device discovery, easy configuration, and a lightweight interface. Overall, it’s good. But why not add logging to the software as well? I doubt it would have much impact on the price. Notifications could also be added, which would be excellent.

I also have a few suggestions for the hardware. For example, I’d really like to see at least galvanic isolation on the RS485 interface. It’s not very expensive, but it’s useful. Ideally, the power supply should be isolated too, which could be done by installing a ready-made module, though that might reduce the operating voltage range.
Why? Simply so you don’t have to worry about ground loops or interference on the sensor and power supply lines, or similar problems coming in through the 485 interface, or a surge frying a bunch of devices. The point is that a device can behave differently “in the field” than it does “on the bench,” and it’s better to be safe than sorry than to troubleshoot glitches at a site that could be hundreds of kilometers from home. Especially when you’re dealing with a device that handles analog signals.

While preparing this review, I also came up with an idea for a similar device, but one that might be better suited to the same “smart homes.” The same kind of module in the same enclosure, but without relays, and with two pulse inputs and two analog inputs instead.
You could connect two sensors from water meters and two temperature sensors to such a device, then monitor water consumption and temperature. And if they add graph plotting to the software, it’ll be even better.

All in all, it’s an interesting device with plenty of room for development. It could be just as useful in the consumer market as in industry. That’s all for now; I hope you found this useful. Share your ideas for expanding its functionality.