#teardownit

The reflectometer's capabilities in terms of maximum distance and damage detection accuracy are determined by the sensitivity of the amplifier and some other important parameters.

Amplifier sensitivity

The sensitivity of the reflectometer, along with the pulse amplitude, is one of its most essential characteristics. It determines the maximum operating range of the device. It is crucial if you're going to check cables with high signal attenuation. Technical documentation often skips this parameter or describes it very vaguely. In an ideal scenario, sensitivity should be characterized as the input voltage when the waveform on the instrument display is contained between the top and bottom edges of the screen (i.e., “x mV for full-screen deviation”). Vertical sensitivity is sometimes measured in decibels. The decibel value is relative and has no meaning unless given a reference level of 0 dB. With this data, the amplifier's sensitivity can be calculated, as each 6 dB step doubles the gain. Based on the parameters mentioned above—pulse amplitude and amplifier sensitivity—it is possible to calculate the maximum overlapped line attenuation, which also serves as a criterion for assessing the quality of TDR. The maximum overlapped attenuation (amax) is defined as the line attenuation when the deviation of the vertical beam is at least one-eighth of the full screen. In this case, amax is calculated using the formula:

‍ Amax = 20 lg8 + 20 lg(Upuls/ampl), - Amax is the maximum attenuation, - Upuls — pulse amplitude under load Zo, - Vampl — amplifier sensitivity for full-screen deviation.

‍ It should be mentioned that this method can only be used to compare reflectometers from different manufacturers. It will be impossible to accurately calculate the maximum attenuation overlapped by the reflectometer using this method since the attenuation in metal cables is frequency-dependent. Therefore, different duration pulses will correspond to different Amax values. The covered method for estimating overlapped attenuation better suits optical TDRs. Optical fibers have frequency-independent attenuation, and therefore, in optical time domain reflectometers (OTDR), the pulse amplitude and sensitivity of the photodetector are usually not taken into account; they are "integral" parameters of the device. A parameter called "dynamic range" is introduced instead, i.e., the insertion loss of the line, at which the signal-to-noise ratio SNR = 1 for a certain duration of the probing pulse. To widen the dynamic range of OTDR, one needs an increase in probing pulse power and the receiver's sensitivity, as well as a very specific set of digital processing algorithms developed by the manufacturer.

Resolution and accuracy

[illustration from viavisolutions.com]

Before discussing the issues related to resolution and accuracy, some basic concepts should be defined. Display resolution is the interval between two consecutive dots on the screen. Fault detection resolution is the minimum distance between two consecutive faults to be visible as separate faults on the reflectometer display. Sampling precision is just the precision with which samples are carried out. The term "fault localization precision" also speaks for itself.

Fault localization precision 

The accuracy of finding the fault in the vast majority of real-world cases is limited by the availability of reliable information for the cable being tested and not by the sampling precision or display resolution of the reflectometer. Firstly, the pulse propagation coefficient can only be known with an accuracy of a few percent, and on top of that, it is affected by temperature changes. A 1% error in setting the pulse propagation coefficient leads to a 1% error in determining the distance. Secondly, the limited information on the actual cable route, in turn, limits the user’s ability to find a point along the cable trace marked by the reflectometer as the location of the fault. After all, when tracing a cable to localize the damaged spot, it is nearly impossible to account for all the peculiarities like safety loops, different laying levels, specific terrain, etc. The resolution of fault detection is also influenced by the pulse duration.

Sampling precision and display resolution

Sampling precision and display resolution are not the same thing. The device may have a high level of sampling accuracy, but its display may have poor resolution, and vice versa. For example, consider a device with a display resolution of 100 ns. It can detect damage with an accuracy of +100 ns, regardless of the precision of the sampling generation every 100 ns. Similarly, the device may have a display with a very high resolution (say 0.1 ns). Still, if the sampling precision is insufficient, then in this case, it won't be possible to accurately determine the location of the fault. One should remember that when searching for a fault location with a reflectometer, high display resolution, and sampling precision are not always necessary. Even if the reflectometer allows one to find the damage locations within 1 cm, due to the cable routing environment, this level of accuracy is not enough for fault localization; you simply won't be able to find them on a real cable. Thanks to modern clock generation techniques, many OTDRs have sufficient accuracy and resolution of sampling that they can be used in most real-world applications. A few other important characteristics are tied to the parameters we've covered. We are talking about measurement limits (maximum and minimum distance) and the possibility of scale adjustment.

Maximum distance

Modern technologies make it easy for manufacturers to achieve an almost unlimited range for their devices. However, the sensitivity of the device does not guarantee actual detection of damage at the specified maximum distance. The best way to determine the range of a device is to test it with a cable. The OTDR's distance measurement range and cable length should be gradually increased until a critical point is reached, i.e., a state when the OTDR cannot detect the open end of the cable. When comparing different reflectometers in such a way, it is necessary to use pulses of the same duration.

Minimum distance

Suppose a minimum operating distance of 10 m is specified for one device and 20 m for another. In that case, one shouldn't immediately jump to conclusions. Both devices likely have the same display resolution. Still, the second one has a physically larger display or is less capable of adjusting to the same resolution. Therefore, the key characteristic is the minimum display resolution, not the minimum working distance.

Scaling

Have you ever made a processor? I did. Took me just 12 microchips and a clock generator.

The processor can execute several commands, and programs can be written for it. The program code and input data are entered using micro switches, and the state of the output register is displayed by four LEDs, according to the number of bits.

At first glance, even this circuitry may seem very complex. Still, it consists of simple modules interconnected by data and control buses. Understanding each module's operation and interactions is relatively easy.

The primary function of a processor is to perform arithmetic and logical calculations. Therefore, our processor is naturally built around an arithmetic-logical unit (ALU).

As an ALU, we will use a full four-bit binary adder SN74HC283N. This chip is asynchronous; it adds two four-bit numbers and outputs a five-bit number (a four-bit number and a carry flag if there is an overflow).

If you are unfamiliar with the binary numeral system, it's high time to learn it. It is precisely the same as the decimal we're used to; its base number is 2 instead of 10.

0000b = 0, 0001b = 1, 0010b = 2, 0011b = 3, and so on until 1111b = 15.

If one adds 15 to 15, the resulting number is 30. The corresponding binary number is 11110b. That is 14 plus 16, or 1110b plus 10000b.

In other words, the carry flag here means the number 16, 2 to the fifth power. The SN74HC283N adder considers the presence of a carry flag at the input and outputs a carry flag to the output, which is why it is called a full adder.

So, we have a microchip with two inputs for four-bit binary numbers: A3..A0 and B3..B0.

Why does binary digit numbering start with zero? The digit's index is the power to the base; in our case, the base is 2.

0100b = 4 - 2 to the second power. 1000b = 8 - 2 to the third power. 1001b equals 2 to the third power plus two to the zero power, 8 + 1 = 9.

Any number to the zero power is equal to 1.

The microchip has a four-digit sum output S0..S3 and a carry flag output. It also has overflow input, but we won't use it, so we'll simply ground it. Thus, this input will always be a logical zero.

What can we do with the adder? We can add two binary numbers and receive their sum. We have a carry flag that indicates whether the adder has overflowed. We can add a binary number to 0 by entering 0000b (simply four 0 bits) into one of the inputs. The output will be the same number. You'll see in a moment how this is useful.

So, we've already got ourselves an ALU—not that capable, but enough for us to use. Next, the processor needs registers.

Registers are internal memory cells of the processor meant for direct access. From these exact registers, four-bit numbers arrive at the inputs of the ALU, and the resulting sum is written to them.

The register is based on a synchronous latch. Our asynchronous adder does not save data; its output momentarily pulls its pins low or high, corresponding to the input values.

(To be more precise, it happens almost instantly, but the performance of the SN74HC283N chip is so much greater than the clock frequency of our processor that we can ignore this "almost" part.)

Unlike the asynchronous adder, our synchronous flip-flop changes the state of its outputs only when it receives a clock pulse and is subject to "enable" signals at its control inputs. It will keep its state until the signals at the control inputs are permissive and a new clock pulse arrives.

We use a total of four identical 4-bit SN74HC161N registers in our processor. Strictly speaking, they are not just registers but much more sophisticated multifunctional chips—synchronous binary counters.

The SN74HC161N chip has a 4-bit input P0..P3 and a 4-bit output Q0..Q3, a carry flag output, a CLK clock input, and four control inputs.

The inverting asynchronous reset input ¬RST sets all 4 latches to zero whenever this input is driven low, regardless of the state of all other inputs.

A low level on the inverting parallel enables input ¬PE allows writing to register Q0..Q3 of data directly from input P0..P3. This will occur on the positive-going edge of the clock.

Parallel enable input is crucial for CPU design. This input allows us to connect all four registers' data inputs P0..P3.

This creates a parallel bus, and the data from it will be read and then written to Q0..Q3 output of the register with ¬PE input set low on the positive-going edge of the clock.

It is possible to set ¬PE input low for multiple registers at once. But we do not need to do it for our processor since it is supposed to output the result of any one of the four available commands to its dedicated corresponding register. We'll elaborate on it later in the text.

The data bus is marked as S0..S3 in the large diagram. The data goes there directly from the output of our ALU, the binary adder.

In addition to the data bus, our processor will have a reset line and a clock line. In other words, we'll connect the reset and clock inputs of all synchronous data-storing chips together.

Four LEDs are connected to the OUT Q0..Q3 register output, so we can see the binary number stored in this register.

In addition to reset and clock inputs, the SN74HC161N chip has count-enabled inputs CET and CEP. Both of these inputs must be high for the number at the output Q0..Q3 to increment at each positive edge of the clock pulse.

The two general-purpose registers, A and B, as well as the data output register OUT, have their CET and CEP inputs connected to the ground. Thus, counting is disabled, and the chip operates only as a latch register.

The program counter PC register connects these pins to the positive power terminal. Thus, the contents of this register will increase by 1 with each clock pulse while the ¬PE input is high.

If the ¬PE input is driven low, the program counter value will not increase by 1 in this cycle, and the PC register will store the number from the data bus instead.

This is useful for jump operators. After all, the program counter determines the address in program memory for the processor to execute the command in a given cycle. Here is the diagram of the program memory module.

The output of the PC register is the address bus A0..A3. It is 4-bit and thus can address 16 program memory locations.

The program memory controller is the SN74HC154N decoder. Fifteen of its sixteen outputs ¬Y0..¬Y15 are pulled high. One of the outputs is low, corresponding to the binary number A0..A3. This determines which program memory cell is read in a given cycle.

The program memory cell is an 8-bit microswitch. Bits with switch contacts closed are pulled low through diodes. And those with contacts open are pulled high through 10 kOhm resistors.

SN74HC540N is an inverting buffer. As a result, D0..D7 will have logical ones where the microswitch contacts are closed and logical zeros where they are open.

The LSBits D0..D3 store the value, the second binary operand for our ALU adder.

The MSBits D4..D7 contain the operation code, the actual command for the processor. The first operand for the ALU is selected by this very command.

If the SN74HC154N decoder is a 4 by 16 demultiplexer, then the SN74HC153N is, on the contrary, a dual 4 by 1 multiplexer. We need two of these dual multiplexers to load a single 4-bit operand into the ALU.

If A = 0, B = 0, the ALU will receive data from register A. If A = 1, B = 0, then the operand will be the contents of register B, and if A = 1, B = 1, then logical zeros.

Finally, when A = 0, B = 1, the operand is taken from the 4-bit microswitch, the input register.

Where do these SEL A and SEL B signals come from? As for SEL B, it's simply the fifth digit from the program memory cell, D5. D4, D6, and D7 proceed to the program decoder.

This decoder is put together based on two logical gate ICs. SN74HC10N is a dual 3NAND gate, and SN74HC32N is a quad 2OR gate.

The control line of the SEL A multiplexers is connected to the output of the OR gate. At the same time, its inputs receive bits D4 and D7 of program memory. If at least one or both are equal to one, then SEL A will be set high; otherwise, the output level will be low.

So, we can indicate where the operand for the ALU comes from using bits D4, D5, and D7. We must also let the processor know which register to store the additional result.

A low level on line ¬LOAD0 selects register A. ¬LOAD0 = D6 | D7, meaning register A will be selected for the result if both D6 and D7 are zeros.

We can already start understanding what some of the opcodes will do.

0000 - ADD A, Im. SELA = 0, SELB = 0. The contents of register A plus the immediate value from D3..D0 is loaded into register A.

0001 - MOV A,B if Im = 0, or A = B + Im. Loading the contents of B into A with the ability to add immediate value.

0010 - IN A with Im = 0 or A = IN + Im. Entering a number from the microswitches into register A, again with the ability to add a direct value. Any instruction has this opportunity since our ALU is an adder.

0011 - MOV A, Im. Loading the immediate value into register A. It is added to zero since SELA = 1 and SELB = 1.

A low level on line ¬LOAD1 selects register B for the result of the operation. ¬LOAD1 = ¬D6 | D7, meaning that to select register B, D6 must be high and D7 must be low. Therefore, the following four opcodes are obtained.

0100 - MOV B, A with Im = 0 or B = A + Im. 0101 - ADD B, Im. 0110 - IN B or B = IN + Im. 0111 - MOV B, Im.

A low level on the ¬LOAD2 line uploads the result of the operation to the OUT register, i.e., to the LEDs. ¬LOAD2 = ¬(¬D6 & D7) D6 should be 1, and D7 should be 0.

1000 - OUT B at Im = 0 or OUT = A + Im. 1001 is the same as 1000 because SEL_A = D4 | D7. 1010 - OUT Im. 1011 is the same as 1010 because SEL_A = D4 | D7.

Finally, selecting the PC register (the operators of moving the program counter pointer to the desired program memory address) corresponds to the bit values D6 = 1 and D7 = 1.

It gets somewhat more complicated from here. Let's take another look at the command decoder scheme.

As you can see, there is also a synchronous D-trigger U1B of the SN74Н74N chip in the processor schematics. I absolutely love this chip; it allows one to do many interesting and useful things.

In this case, the trigger remembers whether the adder overflowed in the previous cycle. Just like our four counter registers, it is clocked by a common CLOCK line and reset by a common RESET line.

¬LOAD3 = ¬(D6 & D7 & (¬C | D4))

This means that when D6 = 1 and D7 = 1, writing to the PC register will occur only if one of the two (or both) additional conditions are met.

If D4 is low, the transition will only occur if there is no adder overflow in the previous cycle. This way, we have a conditional transition.

1100 - JZ B at Im = 0 or JZ (B + Im). 1110 - JZ Im.

If D4 is high, an unconditional jump occurs.

1101 - JMP B with Im = 0 or JMP (B + Im). 1111 - JMP Im.

So, we've looked through all 16 possible commands on our tiny processor and are ready to test them out. But the processor still needs a frequency generator.

Our clock is straightforward. It is made from two inverters of the SN74HC14N microchip. The S130 switch allows one to select the clock frequency: 1 clock cycle or 10 clock cycles per second.

Using the S132 switch, one can toggle between clock-driven and step-by-step execution of the program with a push of the S129 button, which is also equipped with an inverting buffer thanks to the SN74HC14N chip.

Now, let's try to write a program for our small processor.

0000 OUT B 1001 0000; display the contents of register B with the LEDs 0001 ADD B, 1 0101 0001; add 1 to B 0010 JZ 0 1110 0000; if there was no overflow, go to address 0 0011 IN B 0110 0000; read the number from the micro switch 0100 OUT B, 1 1001 0001; return this number + 1 0101 JZ 11b 1110 0011; if there was no overflow, go to address 11b 0110 MOV A, 0 0011 0000; so that the count starts from zero once again 0111 MOV B, A 0100 0000 1000 JMP 0 1111 0000; unconditional jump to address 0

This program will output numbers from 0000 to 1111 to the LEDs and then output the number entered by the microswitches plus 1 until 1111 is reached. Then, the adder will overflow, and the program counter will return to the beginning of the program.

On line 0110, we could just write MOV B, 0, which means register A wouldn't be used in the program. And we want all our registers to be used.

Suppose one does not clear the contents of register B before jumping to the beginning. In that case, it will still contain 1111 and immediately cause an overflow in line 0001. As an outcome, instead of binary numbers 0000 through 1111, we will see LEDs switching from 1111 to 0000 and back.

As you can clearly see, the program works exactly as intended. This means that we have not made any mistakes along the way.

Sometimes, a case exists for using multiple computers from a single desk. Several servers could be quietly grinding numbers in the background until they need to be accessed by the user for configuration or maintenance.

Sure, buying several sets of keyboards and mice is always easier. Still, it's smarter to use a specially designed unit, a KVM switch like the one we have here, the INRIKS SW4041KVM.

Features

INRIKS SW4041KVM is a 4-to-1 HDMI KVM switch. That means one can connect 4 PCs (HDMI outputs and USB ports) and one set of monitor and USB HID devices like a mouse and keyboard for the user.

This user (should we call him "operator"?) can switch from one PC to another with an IR remote or a button on the unit. This selects what computer is connected to the operator's monitor and peripherals.

In addition to these controls, there's RS232 for connecting to an external control system like Crestron.

The maximum resolution for an HDMI signal is 4K at 60 Hz.

Visual overview

INRIKS SW4041KVM is shipped in a simple cardboard box with a brief description on a sticker. The switch is carefully placed in a plastic tray, and all the accessories are under it.

The box contains:

- HDMI KVM switch

- Power supply: 5V/1A

- User manual

- IR remote

- Mounting ears and screws, 1 kit

- Grounding screw

- Terminal block for RS232

- USB A to B cables, 4 pcs

The device kit is very complete by today's standards. It includes nearly everything needed, even four USB cables. The only exception is the HDMI cable, which has to be bought separately.

The back panel has the following connection ports:

- power input, 5V

- RS232 terminal block connection

- 4 groups for PC connections, each with an HDMI input and a USB-B port

- HDMI output for the operator

- 2 USB-A ports for keyboard and mouse

Front panel

- 2 USB-A ports, just in case you need them

- IR window for the remote

- LED indicators for active input connection

- button to cycle between those inputs

- power LED

The RS-232 port allows the switch to be connected to various control systems to automate processes. The included manual has all the commands for controlling the switch from an external automation controller. Everything is clearly listed, along with the connection parameters needed.

A complete mounting kit with 2 brackets and 6 screws is also included. Those brackets are installed flush with the bottom panel of the case and are useful for mounting on a flat surface.

It is worth mentioning that the case is made of steel, and there is a grounding screw, among other accessories. So the unit can be safely installed in a server rack.

We've found that the paint is scuffed around the mounting point inside the case. This should be done intentionally to provide a better case-to-ground connection. Still, it is rather odd to do the paint scuffing manually.

Disassembling the case, we found it thick, substantial, and solid. Port cutouts are exactly where they need to be, with no shifts or bends.

Looking closely at the PCB inside, there's nothing interesting there. Everything is soldered neatly and accurately. There are a few sloppy spots on the backside, but nothing to worry about. I'm just mentioning those to be somewhat fair.

Testing

To test the INRIKS SW4041KVM switch, we prepared a set of three laptops, a monitor, a keyboard, and a mouse.

Everything's connected, and we have an image on our screen a couple seconds after powering up the switch. The mouse moves as it should; there is no input lag or video distortion. That is exactly the point of this device; one can't tell if there's an active device between the desk and the computer.

We've tried to switch inputs with the button and the remote. The only difference is that you can pick any channel with the remote, and the "SWITCH" button just cycles through them (1, 2, 3, 4, 1, 2, 3,...)

When switching inputs, there's a few seconds of delay. That is for connecting your monitor to the next PC and syncing USB devices. This behavior is normal and should be considered; it happens every time you switch HDMI inputs.

This is not a drawback for a reasonably priced device. We've tested one with instant switching before, a multiviewer 4 to 1 for CCTV applications, and it had a lower 1080p maximum resolution. In this price range, customers must pick either 4K@60 with no input lag or 4 FullHD connections simultaneously.

Then we went further and tried connecting a USB headset to our workstations. The idea is that we use those laptops for work, and we need sound devices along with a mouse and keyboard.

And the results were mixed. The headset worked except for the LED indicators for the PC connection. We suspect that the wireless base unit does not recognize our laptop, connected through the switch, as a computer for some odd reason. Apart from that, everything worked fine; the sound goes both ways.

To wrap it up, we've tried disconnecting HDMI input cables and seeing if the KVM switch works as just "KM" without "V." The third time was no charm; the switch went to standby mode. So, it switches only between active HDMI inputs and does not care if an active USB connection is plugged in.

Kinda verdict

HDMI KVM switches save space and simplify wiring for very particular cases. INRIKS SW4041KVM is an excellent device for connecting one set of peripherals to multiple computers and infrequently switching between them.

RS232 is definitely a plus to automate the switching process.

Unlike bridges, which can be used to measure the distance to the defect along the cable route, pulser fault locators search for the cable fault location, i.e., determine where exactly the defect is located. At the same time, cable parameters or temperature do not affect its localization accuracy.

The primary purpose of pulser fault locators is to localize defects in the protective insulating sheath of the cable, due to which the shield and/or conductors are shorted to the ground. The signal from the generator is applied between the defective conductor and the earth to find such locations. As a result, a closed circuit is formed (generator-conductor-defect-soil-generator).

Current flowing in the ground is concentrated in the area of the earth electrode and fault. However, the current flows at different depths along various paths with the lowest resistance between these points. A sensitive voltmeter can detect the potential difference in any area of the ground due to current flow. It is such a device that is used in a pulser fault locator as a receiver. The signal is picked up by two probes located a small distance from each other (usually mounted on an A-frame). The receiver probes are stuck into the ground to reduce transient resistance during signal acquisition.

Depending on the condition of the ground, pulser fault locators detect defects with a resistance of 0.5-2 megohms. For successful localization, it is necessary that as much of the current as possible flows through the fault into the ground and through the ground to the earth electrode. Therefore, the generator should be connected to a high-quality earthing system.

AC voltage or DC voltage pulses are used as the test signal. In the first case, the receiver only measures the signal strength between the two probes (potential difference). In the second case, in addition to the signal strength, the polarity of the signal can also be used to determine the position of the receiver's probes relative to the defect (before or after the fault). But the AC signal also has an attractive feature - it allows you to trace the cable.

In any case, the signal strength is minimized when the probes are positioned strictly above the defect and symmetrical about it: if the fault is located exactly below the probe (midway between the two probes of the A-frame), the signal strength will be zero. It is this fact that is used to localize the defect.

For error-free fault location and refinement of the cable trace, the probes should be placed in a plane perpendicular to the trace and again find the point where the signal will be minimized - accuracy can be as low as 0.1 meters.

The signal levels are maximized at these locations since the current is concentrated near the earth electrode and the defect. If the distance between the earth electrode and the fault is large, the signal may not be detected because it is too weak. The distance to the defect should be estimated using a bridge or TDR to reduce the search time beforehand. This will allow you to start the search close to the fault instead of walking along the trace from its beginning.

It is essential to remember that the signal's amplitude at a certain distance from the earth electrode is equal to the amplitude at the same distance from the defect. This can significantly accelerate the search - using a receiver calibrated in this way allows not to react to minor signal fluctuations on the trace. Using a point located on the other side of the earth electrode at the same distance from the cable as the earth electrode as a reference is usually recommended.

Suppose any conductors (pipelines, armored cables, etc.) run close to the cable and parallel to it. In that case, the return current, choosing the path of least resistance, will flow through them rather than through the ground. This significantly reduces the signal level and makes it difficult to localize the defect.

The same problem occurs if the localizable defect is located in one of several cables from the same point. In such a case, the capacitance of the shield-to-ground circuit can be quite significant since the cable shields are usually connected. The large capacitance of the shield-to-ground circuit leads to the fact that the leakage of the test signal through the capacitance of the shields into the ground becomes comparable to the leakage of the test signal through the defect resistance.

For the same reason, all conductors in the cable under investigation that are grounded under normal circumstances and are therefore capable - instead of grounding - of carrying current back must be disconnected from grounding.

The generator earth electrode should be installed as far as possible from the cable route in a perpendicular direction to eliminate the influence of parasitic circuits running parallel to the cable. The grounding must be as good as possible (with the lowest possible resistance).

Suppose several defects are detected in the cable. In that case, their magnitude is estimated by the leakage current value of each of them - if the cable has deviated perpendicularly from the cable route when taking measurements, the more significant the defect, the greater the distance at which it will be detected. It should be noted that in the presence of several defects located close to each other (which often happens when the cable sheath is damaged or it gets wet), strong defects can "mask" weak ones. Therefore, after the localization of the fault and its elimination, the cable should be rechecked.

When the cable is under a hard surface (asphalt, concrete), to reduce the transient resistance between the probes and the surface, the surface should be moistened by placing pieces of foam rubber soaked in slightly salted water on both probes. However, this complicated method is only used when the size of the hard coating is sufficiently large.

The situation is more manageable if the cable is across a small, hard-surfaced area (e.g., crossing a road). To locate the cable, you can "extend" the probes so that they are outside the pavement. An additional earth electrode is connected to one of the probes with a wire, placed on one side of the pavement, and the other probe is inserted into the ground on the other side to do this. The next step is to find the position of the probe where there is no signal and locate the fault in the middle of the segment between the probe and the second earth electrode.

If the cable is installed along a hard-surfaced area (e.g., under a roadway), a strip of soil along the roadside can be used for the feeler gauge.

1. Our equipment fully* supports** the HDMI standard***

HDMI is the most basic standard for digital video transmission.

"Acceptable" quality is HDMI 1.4, "good" quality is HDMI 2.0. All standards above 2.0 are redundant, at least for now. Also, HDMI at a lower quality than Display Port wins in reliability, which is why professional equipment uses HDMI. By 2016, 100% of new TVs with resolutions higher than HD had an HDMI connector.

The HDMI 2.0 standard has many parameters:

- Resolution, 4K@60 or 5K@30 or 8K@30 (4:2:0)

- HDR support for 1080p@60, 4K60 (4:2:0)

- Widescreen aspect ratios 21:9, 2:1

- CEC, unified control signal

- Audio, up to 768kHz, 24-bit, 32 channels

- 3D

- Ethernet

🧐 In practice, compliance with the HDMI 2.0 standard is guaranteed to mean: 4K resolution at 60 frames and 4:2:0 color subsampling; audio up to 192 kHz 24-bit 2 channels. All other parameters are optional, but these two are sufficient for professional use.

Good HDMI equipment manufacturers list all supported parameters and all technologies in the specifications to avoid surprises during installation.

2. Our equipment fully* supports** HDR technology***

HDR - high dynamic range. This technology allows the display to increase brightness locally. HDR is often used in bright night scenes in movies or when the sun, fire, or explosion is shown on the screen.

In technical terms, HDR is a wider color range, extra bits for extra brightness. Color is 8 bits; a signal with HDR requires 10 or 12 bits. Most budget segment custom displays use a conversion: it takes HDR 10 bits and converts it to 8.

To support extended dynamic range, all elements of the system must work with it:

- A source, computer, or media player with HDR support at the output end

- The content, video file, or web player must contain HDR information

- Cables, extenders must transmit 10-12 color bits

- Display with 10–12-bit HDR support; specific HDMI port on the display must support this mode

- The display must have a peak brightness many times higher than the nominal value; this is the point of HDR technology.

🧐 The "HDR-compatible" label on the equipment does not mean that it supports this standard. It's a marketing ploy, meaning the device takes an HDR signal as input and quietly converts it to regular 8 bits. The source sends HDR, but the HDR-compatible device simply ignores the extra 2 bits of brightness.

🧐 HDR 10 bit (8bit + FRC) is also an example of a marketing ploy. The sentence has a cherished number 10, but it stands for "8 + FRC". FRC stands for Frame Rate Control, meaning that the monitor receives a 10-bit HDR signal and plays back 8 bits digitally enhanced. The effect is like HDR, suitable for inexpensive TVs, but is not HDR.

Good manufacturers do not specify HDR support for equipment where it is emulated or limited to compatibility only. Therefore, the choice of HDMI devices with HDR support is quite narrow.

3. Our extenders transmit the video signal* without delay**

Video signal transmission is always delayed. A 0 ms transmission delay is physically impossible, but it is unnecessary in practice. During transmission, time is lost to encoding/decoding in the extender.

Human reaction time is on the order of 250 ms, and reaction time of cyber athletes reaches 100-120 ms. The minimum perceived latency a human can hear and see is 15-30 ms.

The monitor and the extender give the delay or input lag from the computer or set-top box to the monitor. Consumer segment monitor introduces 30-70 ms of delay; gamer and professional models are closer to 10-30 ms.

🧐 A "latency-free" HDMI extender is a device that transmits a signal faster than 30ms. The main consumer quality of lag-free video transmission is the ability to work comfortably with a mouse or trackpad. For this purpose, it is enough to keep the transmission speed within some known limits.

Where manufacturers specify "latency free", it is there. Still, it is less than all other equipment in the chain and will not be noticeable (e.g., 8ms).

4. Our equipment fully* supports** 4K***

4K resolution is a marketing designation for all resolutions around 4 thousand pixels wide. It's usually 3840×2160 resolution, precisely four (2×2) FullHD screens, and a 16:9 aspect ratio.

16:9 is an excellent usable ratio for a monitor, but they're changing it both ways now. Laptops with modern interfaces are moving toward "square" 16:10 and 3:2. Home TVs show us movies and clips in "wide" theater format, and that's 21:9 or 18:9. All these resolutions can be "about 4,000 pixels" wide, too.

🧐 Resolution is always counted in pixels, not in marketing terms. Strictly speaking, 4K is only 3840×2160, not 4096×2160. This is the resolution "4K" is meant in the HDMI standard.

Good manufacturers list all supported resolutions clearly and precisely in the specifications list, with frame rates for each resolution.

5. Our extenders transmit video without modification* or conversion**

The HDMI standard involves connecting the source to the receiver with a cable. This cable must be made to a specific HDMI standard, with HDMI connectors, between devices with HDMI ports. Extenders, converters, active cables, or scalers do not exist in this formula. These devices came later to extend the standard's capabilities.

- An HDMI cable is 19 individual conductors and pins on a connector. A twisted pair network cable is only 8 conductors.

- A passive HDMI 2.0 standard cable can theoretically be 30 feet long; LAN cable for IP networks is limited to 400 feet.

- Twisted pair typically supports speeds up to 10Gbps, HDMI 2.0 is 18Gbps for 4K resolution.

Signal conversion is inevitable when transmitting HDMI through an extender. Even installation equipment standards that advertise "perfect quality" as HDbaseT convert the HDMI signal at the transmitter and decode it at the receiver.

🧐 All active extenders compress the HDMI signal from 18Gbps to more acceptable values. The transmission accuracy is determined by the quality of the codec algorithm. If the conversion retains ≥95% of the 4:4:4:4 3840×2160 signal, the difference will not be noticeable when viewing.

Nowadays, electronic homebuilders on microcontrollers, especially Arduino, are very popular. The microcontroller makes it easy to make many interesting and useful things.

I am even more interested in assembling different devices on simple logic circuits. They are easier to buy when there is a shortage of semiconductors, and they are not susceptible to software failures, hacking, and firmware corruption. And most importantly, their principle of operation is more visual, which gives me aesthetic pleasure.

Today I will introduce readers to four simple devices. Three of them are based on flip-flops, and the fourth one, which we will start with, could use a trigger.

Voting machine

The majority rule applies when people vote to accept or reject a proposal. The decision is made if more than half of the participants vote in favor.

If there are three voting participants, it requires two or all three to press the buttons to answer in the affirmative. This is the logic of this little electronic toy.

The 74LS08 chip contains four AND logic elements with two inputs. The output of each element will be 1 if 1 at both inputs. Otherwise, it will be 0.

The circuit uses three of the four 74LS08 elements. The inputs of each are connected to one of three possible pairs of buttons. They are pulled up to the ground with resistors, so the input will be a logical 0 unless the button that connects the input to plus power is pressed.

The CD4075 chip contains three OR logic elements with three inputs. The element's output will be a logical 1 if there is a 1 on at least one of the inputs or two or all three.

If any button pair is pressed, one of the inputs of the 74LS08 will be a logical 1, and the output will be a 1. The transistor will open, the buzzer will sound, and the LED will light.

In the case of an actual voting machine, it would be desirable to lock the button so that the voter does not have to hold it down. For this purpose, lockable buttons or toggle switches could be used.

Or you could use flip-flops - devices with two stable states that remember their state until a signal comes that changes this state.

Standby light with touch switch

But first, let's study a monostable multivibrator, aka a single vibrator. This device has one steady state, which exits under the action of the input signal, gives a certain pulse duration to the output, and returns to the steady state again.

This property can organize lighting in some small room or cabinet. The economical red LED consumes very little current and indicates the sensor location.

If you touch the sensor, the red LED turn off, but the five bright white LEDs turn on. Depending on the adjustment resistor, they will shine for a few seconds to a minute or two. After that, the white LEDs turn off, and the red LED turns on again.

The basis of this circuit is one of the most popular chips among DIY enthusiasts - the NE555 integrated timer.

The two main components of the NE555 timer are an RS trigger and two comparators. The comparator is essentially an operational amplifier, or more precisely, an operational amplifier mode.

When the voltage on the "+" input is higher than on the "-" input, the comparator output will have a high voltage level, i.e., almost plus power. And if, on the contrary, the "-" input voltage is higher than the "+" input, the output will be low - almost minus power. That is practically zero/ground because our power supply is unipolar.

The RS-Trigger is a circuit with two steady states, a logical 1 and 0 at the output. The one on the S - Set input switches the output to a high state, and the one on the R - Reset input switches the output to a low state. Thus the trigger is a memory cell.

The NE555 timer inputs are high impedance, which allows you to connect touch buttons directly to them. A touch button is a metal plate, or a metalized surface on the board, as in our case, where you can touch it with your finger.

The touch of a finger on our sensor connects the capacitance and a human body inductance to pin 2 of the NE555 timer. The capacitor C1 protects this input from interference to prevent false positives.

The variable voltage induced on the human body at the high impedance input of comparator 1 will cause the comparator to switch between high and low logic whenever the instantaneous voltage value passes through a value equal to a third of the supply voltage.

One switch of comparator 1 to logical 1 is enough to set the trigger to 1. Pin 3 supplies plus power to the anodes of the bright white LEDs. They turn on, and the small red LED, which glows when its cathode is minus power, turns off.

At the same time, the transistor, the open collector of which is connected to pin 7, closes. This transistor is designed to discharge the time-setting capacitor C3. The capacitor starts charging through resistors VR1 and R2.

When C3 is charged to 2/3 of the supply voltage, comparator 2 will go into a high state, going to the reset trigger input. The C3 charging time, and consequently, the LEDs glow, depends on the VR1 resistance, which we can adjust, and is in the range of 3 to 130 seconds.

The transistor will open and almost immediately discharge C3. At output 3, i.e., the cathode of the red and the anodes of the white LEDs will be a logical 0. The white LEDs will stop shining, but the red one will turn on. And our standby light (red LED) will wait for the next touch of the sensor.

LED candle (nearly)

This electronic candle does not simulate the Flame flicker, but you can turn it on with a lighter or match, and turn it off with a whiff, just like a real candle. The infrared photodiode responds not only to the flame but also to any household appliance IR remote control.

Here the trigger of the CD4013 chip is used in the already familiar RS-trigger mode. In fact, it is capable of more, as we will see next.

Transistor Q1 amplifies the signal of the infrared photodiode D1. With the trimmer resistor R3, which is the collector load of the transistor, the sensor sensitivity can be adjusted, and the capacitor C1 suppresses high-frequency interference.

The amplified signal comes to the trigger input S. The trigger goes into the state of logical 1, and our candle - LED D2 - lights up.

The LED is not connected directly to the chip output, although it could handle such a small current, but through an inverting dc amplifier on transistor Q4.

It is inverting because the logical 1 at the base of the transistor turns into a logical 0 at the collector. The collector is connected to the LED anode, and with a logic 0 at the anode, a current will flow through the LED.

The microphone signal is amplified by two transistors, Q2 and Q3. They form a pure class B amplifier, which means that when there is no input signal, both transistors are completely closed, and no current flows through them.

Class B audio amplifiers distort the sound signal very much, so class AB is used instead in audio equipment to save power and class A for quality. But in our current device, it is not required to play the audio signal but to generate a pulse on the R input of our trigger to turn off the LED when blowing into the microphone (blowing out the candle).

The positive pulse from M1 through C2 will open Q2. The minus relative to the emitter will appear at the Q3 base, a PNP transistor that opens just by a minus. On the open collector of Q3 will be a plus; this is just the low logic level that will reset the trigger and blow out the candle.

The combination lock

We promised to tell you more about the CD4013 chip, and this electronic lock uses its functionality.

The circuit is based on 4 synchronous D-flip-flops, i.e., the two chips. Each CD4013 contains two such triggers.

The data input - D - U1A is always a logical 1 because this input is connected to plus power. But the synchronous D-trigger passes the input state to the output only when it receives a clock pulse on the C-Clock input. The rest of the time, the trigger keeps the output state that it remembered earlier.

U1A receives the synchronization pulse from button S1. Next comes a flip-flops chain, transmitting a logical 1 to LED D1, the light of which means the opening of the lock.

Remember we said that the CD4013 output current is enough for the LED? Here the LED is connected directly to the output of the chip.

Flip-flops U1B, U2A, and U2B are synchronized by buttons S4, S7, and S9, respectively. Therefore the secret code is 1479. This is the sequence in which the buttons must be pressed to open the lock.

Buttons S2, S3, S5, S6, S0 reset flip-flops U1A, U1B and U2A. After pressing any of them, the code must be entered from the beginning. But they do not close the open lock because they do not affect the last flip-flop.

To close the lock, press the button S8, which resets the U2B or S9 flip-flop. Because with the opening of the lock, the inverting output U2B through the logic inverter Q3A resets the previous three flip-flops. Pressing S9 will transfer a logic zero from the U2A output to the U2B output and close the lock.

The CD4011 chip contains 4 NAND logic elements. It would help if you connected all its inputs together to use the NAND as an inverter. The capacitor C1 protects the reset line of the flip-flops when the lock is opened from false flip-flops.

Demo video

The ARC (Audio Return Channel) function in HDMI has helped to remove unnecessary communication between TV and AV receiver, soundbar, or home theater. What does the ARC feature mean for twisted-pair HDMI extenders, and how can you use it?

We will review the INRIKS EX4076E twisted-pair HDMI extender with ARC function (and other useful features).

Main Features

The INRIKS EX4076E extender provides a maximum 4K60Hz video transmission range of up to 230 feet over the CAT6 twisted pair.

The transmitter has an additional HDMI pass-through output for connecting a local TV. The receiver has a S/PDIF digital optical audio output.

The kit provides two-way IR signal transmission to control a remote TV on the transmitter side or a video source (e.g., TV receiver) on the primary TV side.

External review

The INRIKS EX4076E is shipped in a plain gray box. The pictograms on the label show the device's main features.

Inside the box are the transmitter, receiver, and manual. The transmitter and receiver are packaged separately in tactile sachets.

Under the top layer are additional accessories: power adapters, brackets, IR emitter, and IR receiver.

The front and back panels of the transmitter and receiver indicate that we have a complex long-range HDMI signal transmitter. This kit gives the installer additional flexibility in connecting additional devices.

The transmitter has a pass-through HDMI output (e.g., connecting a local TV without an additional HDMI splitter). You can read more at this link:

The kit has a two-way IR signal transmission function. You can read more about this function by following this link:

The transmitter and receiver have In and Out inputs for connecting the IR blaster and IR receiver.

But only one pair is included. The manufacturer assumes that you will transmit the IR signal only one way. Another strange thing for us was the massive size of the IR blaster and IR receiver.

Perhaps these accessories have better signal reception and emission characteristics. But it won't be easy if you need to do a covert installation.

An IR blaster and receiver photo of another INRIKS HDMI extender kit is below for comparison.

The good thing is the removable brackets are included. The brackets are made to attach more flexibly to various vertical and horizontal surfaces.

The receiver has a S/PDIF digital optical audio output on one side and a S/PDIF / ARC switch on the other.

The S/PDIF output connects external audio devices with an optical cable (receiver, amplifier, soundbar, etc.).

S/PDIF mode.

The receiver extracts audio from the HDMI signal from the transmitter and sends it to the S/PDIF output.

ARC mode.

The S/PDIF output will return audio from a TV connected to the receiver (if the TV supports the HDMI ARC function).

Myth. Thanks to ARC, the sound from the TV is transmitted over twisted pair towards the transmitter (for some installations, this would be very useful). It is not!

Devices teardown

The receiver has a thick housing with ribs to improve cooling.

The PCB is locked into the case tightly without any gaps. The front side does not have any problems. But the back side has a minor patina at the soldering points.

Testing

We connected the transmitter and receiver with a 230 ft CAT6 twisted pair for the first test. The connecting cables for the laptop and TV were 6 ft long.

The video started streaming to the TV without any glitches immediately after the connection. We replaced the twisted pair cable with a CAT5e 230 ft cable. To our surprise, the video appeared on the TV. The manufacturer specifies a range of 230 feet exactly when using CAT6 cable, and this is very important.

But after some time, the video signal disappeared for a short time when using the CAT5e cable. Therefore, we do not recommend its use and advise you to follow the manufacturer's requirements.

For CAT5e cable, the length should be no more than 170-180 feet.

Instead of a conclusion

In the previous article, we reviewed the basic parameters NEXT and FEXT. The Attenuation to Crosstalk Ratio (ACR) is used when evaluating structured cabling. This parameter is equivalent to the signal/noise parameter concerning the transient effects at the near end of the NEXT, i.e., it serves as an estimate at the receiver input for the undergone line attenuation of the signal and for noise from the transient effects at the near end. Quantitatively, ACR is expressed as a logarithmic measure of the difference between NEXT and cable attenuation. If, for example, the ACR value is 10 dB, this means that the NEXT power of the interference at the receiver input will be 10 times less than the power of the useful signal, i.e., the signal-to-noise ratio will be 10.

Suppose the communication system operates in single-cable mode, and the signal levels at the outputs of the transmitters at points A and B are the same and equal to 0 dB. If the line attenuation at frequency F denotes by Ak, then at transient attenuation NEXT at the same frequency signal levels, Pc and transient noise Pp at the input of receiver A will be Ak and NEXT.

Then ACR = Pc - Pp = NEXT - Ak.

The practical meaning of the ACR parameter becomes more precise if the frequency characteristics of the attenuation of a symmetrical pair (A), transient interference (NEXT), and the parameter (ACR) are presented on the same graph. The frequency at which the values of attenuation and NEXT are the same (in this case, it is 100 MHz) determines the upper limit of the operating frequency range. At frequencies above the boundary value, the power of the interference NEXT exceeds the signal strength.

Another cable system indicator, the Equal Level Far End Crosstalk (ELFEXT), has the same physical meaning as ACR. The only difference is that ACR is associated with NEXT, while ELFEXT is associated with FEXT. The ELFEXT parameter becomes critical when several transmitters of the same system are transmitting in the same direction on pairs located in the same cable.

In this case, ELFEXT = FEXT - Ak.

In addition to the ACR and FEXT parameters, two additional parameters are used - PS-ACR (Power Sum ACR) and PS-ELFEXT (Power Sum ELFEXT), which consider the real influence on this pair of all other cable pairs.

ASYMMETRY (IMBALANCE) LINE

Asymmetry is a transmission parameter, as it is determined by the pair parameters and affects its bandwidth, and an influencing parameter, as it affects the transitions between other pairs.

Each symmetrical line must be balanced concerning the ground in a certain way. Two types of asymmetry are distinguished depending on the current, DC, or AC.

DC asymmetry is estimated by the relative value of the difference of resistance of the cores of a symmetrical line and should not exceed 1%. The presence of the resistive unbalance of the line, equal to the difference of its core resistances measured at AC current, can be interpreted as the inclusion of an additional low-pass filter with the resistance of the longitudinal arm dR. In addition to the resistive component, the longitudinal unbalance of the line in the general case contains a capacitive component; it can arise, for example, due to accidental crossing of the cores of different pairs in the connection points of cables. This component can be interpreted as the transverse capacitance of the additional low-pass filter mentioned above.