Multifunctional voltage relay RN-263t from Novatek-Electro

After reviewing two phase-selection relays, I managed to arrange for a couple of interesting devices to be provided for review. One is aimed at the "household" market, while the other is more geared toward small businesses. Today I'm reviewing the first one, the RN-263T voltage relay.


There are some difficulties with the name of this device. It's usually called a voltage relay, but I think voltage monitoring relay is more accurate, although of course it's all just semantics. In any case, these devices are designed to protect the connected load from overvoltage and undervoltage (which is just as critical).
Some models can do more than just disconnect the load when the voltage goes outside the set limits: they can also introduce a delay before switching it back on (critical for compressors and the like), monitor the load current, display the current mains parameters, and so on.

On the manufacturer's website, the device is currently offered only in the Ukrainian market, but it's already available elsewhere in Russia — link.

So today I'm reviewing just such a device, so to speak: with blackjack and hookers.

It's packed in the familiar Novatek cardboard box, with inserts to keep everything from rattling around in transit.


The package includes the voltage relay and an instruction manual, which also serves as the warranty card.


The manual is quite detailed, covering both the technical specifications and the controls and settings.
You can view the full version here.


Technical specifications. Note the remark at the very end: as far as I can tell, the relay operates over a range of 130–450 V, but the display only shows values from 130 to 350; anything higher is simply treated as "a lot".


The manual also describes the error codes the device logs during mains faults. It also lists the warranty terms: the warranty period is 10 years from the date of sale, which is pretty good, especially considering the company has been around for a long time.


And here's the star of the review. It looks very neat.


For its specifications, the relay is quite compact: it takes up just 2 modules on a DIN rail. In the photo, it's shown next to a couple of circuit breakers.
Although, in fact, most relays of this type are now made in this size. Meander even has a version that takes up just one module. I'm comparing it with the one I have at home, which takes up 3.5 modules (a very awkward size).


The input terminals are at the top, and the output terminals are at the bottom. All the necessary markings are present, and you can also see the ventilation slots.
The wiring diagram on the side shows that the RN should be preceded by a 63 A double-pole circuit breaker. The manual specifies a type B breaker.
The clip is spring-loaded.


And now the main feature: the front panel.
There's a large display right away, showing voltage, current, and power. Three LEDs to the right make it easier to read, and the display also shows additional information.
The only button is below the display on the right. It switches between display modes and provides access to the logs and maximum current setting.
There's another LED on the left. It indicates whether voltage is present at the phase terminal, so if the relay contacts stick, it will stay lit even when the RN is in a fault state.
At the very bottom are three knobs for setting the delay before switching on and the minimum and maximum voltage thresholds.


When power is applied, the display first shows STA, which I assume stands for Start. Then it counts down the delay, after which the relay switches on and the device enters monitoring mode. The display then shows the last selected value. If the voltage is outside the set limits while the delay is counting down, the countdown starts over.
No-load consumption is about 1.5 watts, but this is a very rough measurement, so I wouldn't rely on it.

Pressing the button switches the display between input voltage, circuit current, and power. The last selected mode is saved and will be shown the next time the device is switched on.


There are three trimmers for setting the delay and voltage thresholds, but their readings are actually processed by a microcontroller. As you turn them, the value being set appears on the display, with fairly high precision. For example, I could easily set the voltage in 1–2 V increments.
Switch-on delay: 5–900 sec.
Minimum voltage threshold: 160–230 V
Maximum voltage threshold: 240–290 V


If you press and hold the button, after 6 seconds the relay switches to the trip log display mode, where you can view the reasons for the last five trips. The manual explains this, but in my opinion it could be clearer. I think the problem is that the display is too simple to show so much information using different symbols.


In addition to setting the delay and voltage, you can also set the maximum current. To do this, hold the button for more than 10 seconds. Unfortunately, there is no indication of how long you have been holding it, so you just press and wait. After 6 seconds, AEr appears (log display), but if you keep holding the button, the current protection setting is displayed after another 4 seconds.
Adjustment range: 1–63 A in 1 A increments. On older versions of the relay, the setting was in 1 kW increments; here it works out to 230 W increments, which is certainly more convenient.
Just out of curiosity, I set the trip current to 1 A and connected a 2.2 kW iron. I thought the relay would trip immediately, but it didn't: the relay waits 5 seconds, and only disconnects if the current hasn't dropped by then. After the overcurrent protection trips, the relay doesn't respond to anything; the only way to reset it is to disconnect power from the relay.
Of course, all this is described in the manual, but people here are used to reading it only after they've “broken everything” :)

The log recorded an overcurrent fault notification.


Before testing, I decided to take the relay apart and look at its construction. It's very easy to disassemble: there are two clips on each side.
The contact assembly is similar to the one used in circuit breakers and looks quite robust.

To take the relay apart further, you need to gently flex the plastic housing outward, since the top board is also held in place by several small clips.
I like the design overall: it's relatively simple, with no ribbon cables, wires, or loose parts.


Since the relay only breaks the live wire, the neutral contact is a solid busbar of a decent cross-section. The material isn't magnetic; only the clamping mechanisms are steel. The neutral wire that powers the protection relay's electronics is soldered to this busbar.


I found no issues with the circuit board either; the soldering is good and clearly not done by hand. Three 1206 resistors are connected in series in the voltage measurement circuit.
Current measurement is handled by an op-amp, MCP6021, and this left me a bit puzzled. The thing is, this op-amp is single-channel; the dual-channel version has a 2 at the end of its part number, and the quad-channel version has a 4. The exception is the 6023, which is also a single-channel op-amp. The dual-channel version has the same pinout as the familiar LM358 and similar parts, as well as the same circuit board layout, while the single-channel version is completely different. In short, I got confused here, and I'm more inclined to believe the board has a 6022 rather than a 6021.


The display and control board connects to the power and measurement board via a soldered pin header.
The power supply uses a capacitive dropper; in this case, it consists of two 0.33 µF capacitors connected in parallel.
One of the power supply filter capacitors is near the input terminals, with a BT169D thyristor next to it.
Near the output terminals are another pair of 470 µF, 25 V capacitors and an LM317 voltage regulator.


I happened to find a capacitor on the top board that clearly wasn't part of the original design. It happens; I don't see anything wrong with that.


A polarized relay, NRL709BC, rated for currents up to 80 A, is used as the switching element.


The device is rated for 10,000 loaded operations and 500,000 mechanical contact operations, i.e. switching on/off without a load.
However, the 63 A (14 kVA) rating applies only to a resistive load; with a more demanding load, it may be 10 times lower.


The relay manufacturer doesn't worry about this and assumes all loads will be resistive, but the service life specified by the manufacturer matches the figure given in the voltage relay's description.


The relay is most likely custom-made, and you can tell not only from the markings but also from the power contacts, which are specially designed for a particular enclosure model.


A temperature sensor is glued to the input phase terminal. The section of the terminal between the input and the relay also serves as a current sensor. This section is 26 mm long, and its resistance, measured with a YR1030, is about 0.04-0.05 mΩ, though this is a very rough measurement.


While the voltage relay was disassembled, I checked the contact resistance.
1, 2. The voltage drop at currents of 1 and 6 A, measured close to where the contacts exit the relay housing, was 0.862 and 5.156 mV, respectively. This gives a calculated contact resistance of 0.86 mΩ.
3. The total voltage drop along the phase line was 1.1 mV, giving a resistance of 1.1 mΩ.
4. The measured resistance of the current-sensing shunt was 0.05 mΩ.

So, the description of the relay used here specified a contact resistance of 1.2 mΩ. I measured 0.86 mΩ, which is of course very good, except for one catch. Let's bring in some physics and math and work out how much power will be dissipated in the voltage relay's phase line.
With a resistance of 1.1 mΩ and a current of 63 A, the voltage drop is 69 mV, so the power is 0.069 × 63 = 4.35 W, of which 3.4 W is dissipated in the relay. That's a lot; dissipating that much power in a cramped enclosure is difficult.
At 50 A, things look much better: 0.0011 × 50 × 50 = 2.75 W. Personally, I think this is the maximum continuous switching current.
You may ask how other manufacturers compare. I think it's pretty much the same, and in Novatek's defense, the design minimizes all possible contact resistance. In fact, the wires connect directly to the relay contacts, so it's almost impossible to do better, unless you find a relay with even lower contact resistance.


Later, I ran an additional test under conditions closer to real-world use: mains input voltage, 9 A load current, and an 8.7 mV drop across the contacts. This gives a total resistance of 0.966 mΩ along the phase busbar, which means that at 63 A the drop would be 60.8 mV, or 3.8 W. This is because mains voltage breaks through oxide layers more easily than the 5 V usually used for testing.


The schematic turned out to be a little simpler than before. The manufacturer provided it at my request, for which I'm very grateful. This had two immediate benefits: I didn't have to draw it myself, and it also rules out errors.

Here are a few key features:
1. Both half-cycles of the mains voltage are monitored, and the measurement input has an offset.
2. The "rough" supply is regulated by a thyristor rather than Zener diodes. This type of regulator isn't exactly "cutting-edge technology"; it's just very uncommon. In this case, it reduces power dissipation, especially when the load on this circuit is lower.
3. An optocoupler monitors for voltage after the relay, i.e. at the voltage relay output. The same optocoupler also controls the indicator LED on the front panel.


Control and indicator board schematic.



Now let's look at the power supply and voltage measurement accuracy.
The description stated that the minimum operating voltage was 130 V. In reality, the relay tries to operate at 80 V, and the microcontroller starts up normally at 95-100 V, but the voltmeter readings are slightly low and also unstable.


But at an input voltage of 102-104 V and above, the readings not only become stable but are also accurate.
The description also stated that accuracy isn't guaranteed above 350 V. In my tests, everything worked normally up to 400 V; above 402-405 V, the accuracy starts to "drift," and by the maximum input voltage of 450 V, the difference becomes substantial.

So, the voltage relay is not only calibrated very well, but also works accurately over a range of 105-400 V rather than 130-350 V. It's unlikely that voltage fluctuations in a household mains supply will be any greater.


More out of curiosity than as a test, I applied the maximum voltage I could—over 500 V—and the relay handled it without any problems.


During the test, I had to rig up a voltage booster for my LATR, and I noticed something interesting. The upper threshold on the voltage relay was set to 290 V. I applied 285 V, but the relay wouldn't start: the APV (automatic reclosing) counter kept resetting. When the input voltage dropped to 284 V, the relay switched on correctly and stayed on, since the 4 V hysteresis had kicked in.


I ran into some difficulties checking the current measurement accuracy. I simply couldn't get it under a proper load, so I'll have to think about how to do that in the future.
But I can already say that:
1, 2. The relay doesn't respond to loads of 15 and 60 W.
3. With a 150 (125) W incandescent lamp connected, the measurement works; the displayed current is one least significant digit lower than the actual value.
4. I connected a 400 W iron; same result.
5, 6. This is the most powerful appliance I could find at home, and it's also an iron. The result is similar to the previous tests: the reading is a little low, but not critically so. It showed 2.11 kW, which is about right for an input voltage of 230 V and a current of 9.1 A.


But it seemed to me that briefly applying 450 V wasn't all that interesting, especially since the specifications list this as the maximum continuous voltage—a level that's virtually impossible to encounter in a household power supply.
Even so, I applied slightly more than the specified 450 V and left the voltage relay like that for a while.


The test lasted 40 minutes. During the test, the voltage fluctuated between 450 and 455 V. The relay performed perfectly; the hottest part was the top of the case, at around 47 degrees. There was no smell of overheated components.


Later, I thought it would be a good idea to measure the temperature of the components inside, so I repeated the test.
The hottest components were:
1. The LM317 voltage regulator, at 47 degrees.
2. The thyristor in the pre-regulation stage, at 60 degrees.
Everything else was only slightly warmer than the surrounding air.


Next, a few oscilloscope traces.
First, the simplest thing I could check: relay switching is synchronized with the mains, and it always switches on just before the sine wave peak. This is a good way to switch heavy inductive loads, but it causes a substantial inrush current with capacitive loads.
It's hard for me to say whether this is the right approach, but in my personal opinion, if the relay is intended for apartments, it would probably make sense to design it for capacitive loads (TVs, computers, inverter air conditioners), since inductive loads (conventional refrigerators, transformer-based power supplies) are becoming less and less common.


Now for the switch-off speed tests. I'll start with switching off at the lower threshold.
The manual specifies a switch-off time of 7 seconds, but if the voltage drops below 145 V, the relay should switch off within 0.25 seconds.
0.25 seconds is 12.5 mains voltage cycles. I ran a lot of tests and can say that in practice the relay switches off in 12–15 cycles, or 0.25–0.3 seconds.
Below are six oscilloscope traces. The initial voltage was 220 V; during the tests, it was lowered to 140, 130, and 110 V. On the left are the results with the threshold set to 180 V; on the right, 160 V. Switching off is synchronized with the mains and always happens at the same point in the sine wave.


And the response to overvoltage.
I ran more tests here. I'll start with the following setup: the threshold is set to 240 V, and I apply 265, 280, 320, and 335 (340) V to the input.
1. 265 V: switch-off time of about 50 cycles, or 1 second, as specified.
2, 3. 280 V: switch-off time varied from 5 to 9 cycles, i.e. 0.1–0.2 seconds.
4. 320 V: switches off after 2.5 cycles, or 0.05 seconds. Sometimes, though, switching off took longer, apparently because the voltage was close to the threshold between operating modes.
5, 6. 340 V: the relay switched off after 1–1.5 cycles, or 0.03 seconds.



I set the upper threshold to 250 V.
1, 2. I applied 310 V; the relay switched off after 7 cycles, or 0.14 seconds.
3–6. With 340 V applied, the relay switches off after 0.5–1 cycle. It always switches off at the zero crossing, but this can be on either the positive or negative half-cycle.


Upper threshold: 280 V
1, 2. At 310 V, the relay switches off after 50–55 cycles, or 1–1.1 seconds.
3, 4. With an input of 340 V, it switches off after 1 cycle, or 0.02 seconds.


Above, you saw tests where the input voltage was abruptly changed from the baseline of 220 V. This was done very simply: I connected a winding from another LATR to the output of a 220–220 transformer, which let me set the voltage I needed and then add it to or subtract it from 220 V. The 220 V reading was due to the low power rating of the isolation transformer; in reality, it was around 223–225 V, with 230–235 V at the input.

At first, though, I simply adjusted the voltage on the fly with a Variac, so the resulting waveforms show a gradual (relatively speaking) increase in voltage. The response was similar to the one shown above, but I wanted to explain the switching behavior.
1, 2. The cutoff threshold is 250 V; applying 265 V gives a delay of about 1 second before shutdown.
3, 4. Applying 400 V gives a delay of just a few cycles before shutdown.

In fact, there are three response modes for overvoltage:
1. The input voltage is 5–30 V above the set value: shutdown after one second.
2. The difference between the input voltage and the set value is greater than 30 V: shutdown after 0.12 seconds.
3. The input voltage exceeds 335–340 V: shutdown after 0.5–1.5 cycles, or about 0.05 seconds. The manual specifies a threshold of 430 V for this mode, but for some reason I got a noticeably lower value.

In practice, all these modes work more or less as described, although the relay switches off differently in each case. Shutdown is synchronized with the mains in every mode, but:
In the first and second modes, shutdown occurs at the end of the first third of the positive half-cycle (photos 1 and 2).
In the third mode, it occurs during the pause, regardless of polarity, although in my tests shutdown more often occurred at the end of the positive half-cycle (photos 3 and 4).

Undervoltage cutoff is less critical in terms of risk to equipment, so the behavior here is much simpler: between 145 V and the set voltage, shutdown occurs after 7 seconds; below 145 V, it occurs after 0.25 seconds.


Now for the conclusions. First, a colleague, Ksiman, has already reviewed a voltage relay from this manufacturer, and at the time, along with the positives, there was a whole list of shortcomings. The manufacturer has addressed them. Here’s what has been improved:
1. There was no indication that the output was live. Fixed: the actual status is now shown by an LED on the front panel.
2. Current is measured at the relay’s contact busbar, which is more reliable, though I can’t yet comment on temperature compensation.
3. The ballast capacitors are now soldered to the circuit board instead of hanging by their leads.
4. The power terminals are now marked for both phase/neutral and input/output.
5. No overheating occurs during prolonged exposure to 450 V.
6. Shutdown time has been significantly reduced: at voltages above 340 V, it takes just 0.5–1.5 sine-wave cycles.
7. Relay switching on and off is synchronized with the mains.
8. No combination of button presses and adjustments can cause the relay to switch on in fault mode.
9. The current limit works in 1 A increments, or 220 W per increment.
10. The voltage relay measures voltages up to 400 V correctly, although the datasheet specifies a maximum of 350 V.
11. The automatic reclosing timer resets if the voltage goes outside the set limits while it is counting down.
12. The design has been extensively revised and is clearly more production-friendly. I was especially pleased that there are no longer any extra soldered or welded joints in the power circuits.
13. The warranty period is now 10 years instead of 5. The operating temperature range remains the same, from -30 to +55 °C.

Some of the issues I noticed remain. For example, the hysteresis is small—just 4 V—so with a weak mains supply and a heavy load, the relay may cycle on and off. Ideally, this should be an adjustable option.
There’s also a question about when the relay switches on. It’s debatable and depends heavily on the type of load, but the relay switches on at the peak of the sine wave.

That said, the controls are intuitive. You don’t actually need to press any buttons for the main settings, and the knobs are genuinely convenient. It’s also good that the manufacturer added memory for the last selected display mode. I liked the three-level shutdown delay based on the size of the voltage deviation, which should reduce nuisance trips.

Overall, the РН-263t has been extensively redesigned compared with the PH-260t, both in hardware and software, as well as in its construction. It’s essentially a completely different device now. I had almost no concerns about its stability or specifications. The only caveat is the maximum continuous current: my calculations suggest that 63 A is too much for relays like this when combined with high ambient temperatures, a tightly packed electrical panel, and prolonged operation at full load. In those conditions, I wouldn’t exceed 50 A.

That’s all for today. I hope you found this useful.