Fullscreen HTML5Med Res (??MB)HTML4

Cable U S B Tester It’s frustrating when a USB device doesn’t work, and you don’t know if it’s a problem with the device itself or the cable. This is a huge problem if, like us, you have a drawer full of USB cables and don’t know which ones are good or provide power only. Bad cables can also cause intermittent problems. Now there is an easy way to test all manner of USB cables; this USB Cable Tester is so handy, we think you will find it indispensable! Part 1: by Tim Blythman T here is not much worse than an intermittent fault when it comes to checking and diagnosing faulty gear. It’s worse if it is due to a dodgy cable because you can never be completely confident that you have ruled out other problems. So it’s crucial to be able to test cables for this reason. These days, a lot of gear connects with USB cables and not just when it’s attached to a computer. Practically all mobile phones use USB for charging, and they’ve also found many niche uses due to their ubiquity, even for devices like shavers and toothbrushes. So we’ve designed a USB Cable Tester that can check practically all standard USB cables. If you’re like us, you probably have a mix of the latest cables (such as USB-C) and a good number of older types (such as miniand micro-USB). The USB Cable Tester will test any cable with either a USB-C or USB-A (2.0 or 3.2) plug on one end and any USB-C or USB-B plugs (such as 2.0, 3.2, micro or mini) on the other end. With some basic adaptors, you can also test common variants such as OTG (‘on-the-go’) cables and non-standard cables, such as those with USB-A plugs at both ends. 28 Silicon Chip This device is compact and automatic. Simply plug a cable into the appropriate sockets, and it immediately gives you an assessment. You will know straight away if the cable is suitable for your purpose. Testing The USB Cable Tester performs two primary tests. Initially, the various conductors in each cable are tested for continuity at low current. This test can pick up whether, for example, a given lead has the appropriate internal data connections for USB 2.0 or USB 3.2, or whether it can carry power only. It can also detect internal short circuits which can interfere with normal operation. The Tester can also perform a high-current test on the VBUS and GND leads to establish how much current the cable can handle without dropping excessive voltage. Checking the ability of the cable to carry current is arguably the most useful test, as it allows detection of the most subtle and intermittent faults. These are the faults where the device seems to operate normally but fails when a burst of current is needed. The Australia’s electronics magazine device resets due to its supply dropping out and might even immediately start working again. Devices like portable hard drives, which often require significant current, are especially prone to this problem. None of these tests characterise the high-speed data performance of the cable; much more specialised equipment is needed to do this. Still, these tests are performed very quickly and can be used to give a very fast ‘go/ no-go’ assessment on a cable. With the rise of the Right To Repair movement, we think that the USB Cable Tester will become indispensable in places like Repair Cafés. We shudder to think how much good gear has been discarded due to having a faulty USB cable. Background Before delving in, you might like to look at some recent articles we have published. The June 2021 article on The History of USB (siliconchip.com. au/Article/14883) describes the connectors and wiring that the USB Cable Tester needs to work with. That would be a good article to read if you’re interested in understanding and repairing USB cables. siliconchip.com.au The front panel gives access to the six USB sockets: two downstream facing ports (DFPs) at left and four upstream facing ports (UFPs) at right. To test a cable, you plug one end into either of the DFPs and the other end into one of the UFPs. It does the rest automatically. The July and August 2021 issues also included articles on How USB-C Power Delivery (USB-PD) Works (siliconchip.com.au/Article/14919), the operation of USB-PD Chargers (siliconchip.com.au/Article/14920), and USB-PD Triggers (siliconchip. com.au/Article/14996). USB power delivery is a relatively recent addition to the USB standards and is not something our unit tests; these power delivery features are usually built into devices rather than cables. Both this article and the USB Cable Tester use USB 3.2 to refer to any cables that you might know as USB 3.0 or USB 3.1, since the USB 3.2 standard replaced (and is backwards compatible with) both USB 3.0 and USB 3.1. This is a similar situation to the way that USB 2.0 encompassed and replaced USB 1.0 and USB 1.1; it's now common to refer to devices compatible with these as USB 2.0. Design Before delving too deeply into the circuit details, we’ll mention some of the design considerations that we made along the way. We designed the USB Cable Tester to be economical to siliconchip.com.au build, easy to use and robust enough for regular use. While it certainly would be possible to do this job without a microcontroller, that would entail a complicated design. Add in the fact that the nature of the test results are often more than a simple numeric result or basic binary go/no go, and a microcontroller is an inevitable part of the circuit. With that in mind, we’ve used a 40-pin PIC microcontroller. Any fewer pins would require a multiplexer or switch, adding complexity and cost. Rather than fall back on one of the old-fashioned 40-pin micros like the PIC16F877, we’ve decided to get with the times and use its modern descendant, the PIC16F18877. The microcontroller displays the test results on a 20x4 character LCD, allowing simple ‘human-readable’ assessments to be delivered. Thus the USB Cable Tester can be used by even those with no electronics experience. The low-power features of this new microcontroller mean that a power switch can be omitted. This may seem like a small saving, but it’s one fewer part to consider during design and construction and shaves a few dollars off Features & specifications for the USB Cable Tester 1. Test just about any USB cable 2. Current pulse tests at 100mA, 500mA and 1A 3. Downstream facing ports can accept USB-A (2.0/3.2) or USB-C (3.2) 4. Upstream facing ports can accept USB-B (2.0/3.2), USB-C (3.2), Micro-B (2.0/3.2) or Mini-B (2.0) 5. Reports faults with individual cable ends (eg, plug with bare wires or detect OTG cables) 6. Can differentiate between power-only, USB 2.0 & USB 3.2 cables 7. Will report short circuits, open circuits and other faults 8. Reports voltage drop and cable resistance at usable currents Australia’s electronics magazine November 2021  29 Fig.1: like the PCB, much of the schematic is taken up by the 26 resistors that isolate the microcontroller from the USB sockets. In the unlikely event of a ‘live’ USB cable being plugged in, they will afford some protection to the microcontroller and whatever is at the other end. 30 Silicon Chip Australia’s electronics magazine siliconchip.com.au the cost. The USB Cable Tester simply sleeps between uses, sipping a tiny 30μA from the battery. It runs from three AA cells which will last for many years with the USB Cable Tester sitting on the shelf. The 4.5V nominal supply voltage means that no regulator is needed; another part (and more money) saved. The device is housed in a compact 140mm x 110mm x 35mm instrument case, about the smallest that would fit everything. This means that it is sturdy and looks the part, too. Some parts we could not skimp on. The USB Cable Tester uses robust USB sockets, which cost a bit more but are paramount to the longevity of such a tool. We doubt that any other device will have USB leads plugged and unplugged so frequently. We have aimed to use through-hole parts to allow the parts to be easily obtained and the USB Cable Tester to be easily assembled. Still, it contains a handful of SMD parts for various reasons, primarily certain types of USB sockets. Many of these sockets are only available in that form. Circuit details Refer now to Fig.1, the complete circuit of the USB Cable Tester. CON3 and CON4 at upper left are the downstream facing ports (DFPs) – you can equate these to the ‘host’ ports from before USB-C. But since USB-C cables are end-to-end symmetrical, a new distinction needs to be made. CON3 is a USB-A 3.2 capable socket, while CON4 is a USB-C socket (which by nature supports USB 3.2). CON3 will also accept older USB-A 2.0 cables since it is designed to be backwards compatible. CON5-CON8 are the upstream facing ports (UFPs), analogous to the ‘device’ socket before USB 3.2. CON7 is USB-B 3.2 and, like CON3, can also accept a USB-B 2.0 plug; leads with these plugs are sometimes called printer leads due to printers being one of the few items large enough to fit such a port. CON6 is a USB-C socket and is accompanied by a USB micro-B socket, CON8. Like CON7, it can accept either a USB 2.0 or USB 3.2 plug. Finally, CON5 is a USB mini-B socket, which is only available in a USB 2.0 version. The various pins from CON3-CON8 are connected to one of 26 1kW resistors. To reduce the number of pins siliconchip.com.au Australia’s electronics magazine November 2021  31 that are needed, some pins are joined. For example, the GND pins of CON3 and CON4 are connected to the same resistor. That is because these pins perform similar functions in each connector and have no reason to be connected by a cable. They are functionally equivalent as far as the USB Cable Tester is concerned. This means that the USB Cable Tester does not know whether the cable is plugged into the USB-A or USB-C socket, but that isn’t necessary for checking cables. As we noted earlier, a 40-pin microcontroller does much of the work. IC1 is a PIC16F18877 8-bit enhanced midrange microcontroller. It’s one of the cheapest 40-pin microcontrollers available at the moment. There is a slightly more inexpensive version with less flash memory, but given the ongoing chip shortages, we’ve decided to standardise on the part with more flash. 26 of IC1’s pins are connected to those 26 1kW resistors, and these pins are used to probe the connectivity of the cable being tested. For the most part, IC1’s GPIO (general purpose input/output) pins are interchangeable. We use one feature that is not present on all the available pins, and that is the interrupt on change (IOC) feature. The ports that do have this feature have been wired into the downstream facing ports. Without using IOC, we would have to wake up the microcontroller periodically to test whether a cable is connected. This feature automatically wakes it up as soon as any connection is made between the downstream and upstream ports. This made laying out the PCB slightly more complicated but allows IC1 to use the deepest sleep mode available, thus saving the most power when the unit is idle. This circuitry is used to probe any pin combination between the upstream facing port and the downstream facing port. We’ll explain how that works in more detail in the software section below. Current affairs The cable current-carrying capability is tested by sending a brief burst of power through the VBUS (5V) and GND wires of the cable under test. 32 Silicon Chip The completed USB Cable Tester photographed from the front and rear. This shows that all connections are made from the front of the case. One tactile switch is raised so it can be more easily accessed through a hole in the lid. Since practically all the GPIO pins on IC1 can act as analog inputs for its internal ADC (analog-to-digital converter), we can probe the cable at several points to see how much voltage is dropped between them. Up to 1A is supplied by a circuit based around Q3, a P-channel Mosfet. Q3, L1, D3 and the 10μF capacitor form a fairly standard buck (stepdown) regulator arrangement. When Q3 is switched on by a signal Australia’s electronics magazine from the microcontroller via the 220W resistor, current flows from the battery positive through L1, charging the 10μF capacitor. When Q3 switches off, the inductor’s magnetic field collapses, causing current to continue flowing to the capacitor, through the path provided via D3 and into the 10μF capacitor. As in any other buck regulator, the duty cycle at Q3’s gate determines the voltage that the capacitor charges up siliconchip.com.au to but with the proviso that Q3 is a P-channel Mosfet and thus is on when its gate is pulled low. A 10kW resistor between Q3’s gate and source keeps it turned off when it is not being driven. The test voltage is applied to the cable by three of the contacts of RLY1 and RLY2. One set of contacts connects VBUS of the downstream facing port to the positive end of the storage capacitor. The second set of contacts connects the GND of the downstream facing port to the 220mW shunt resistor returned to circuit ground, used to measure the current. The third set of contacts connects VBUS and GND at the upstream facing port, which is necessary to complete the circuit. Note that current flows in the same direction as it would under regular use. It’s important to realise that none of the USB GND connections are connected directly to the circuit ground during this test. They are connected to either end of the 220mW shunt resistor but only when the relay contacts are closed. The 1kW resistor across the 10μF capacitor discharges it when the buck regulator is not running. This is mainly to reduce the current flowing when the relay contacts open, reducing relay contact wear. Both relays are controlled by N-channel Mosfet Q2, which sinks current from the battery through both relay coils when its gate is brought high by the microcontroller. A 10kW resistor keeps the Mosfet off when it is not being driven, and 1N4148 diode D1 absorbs the back-EMF from both coils. As we noted, power is derived from three AA cells, giving a nominal 4.5V. A separate battery holder is wired into CON1. This feeds the 1000μF bulk bypassing capacitor, Q3 and powers the buck regulator and relay coils. Schottky diode D2 feeds from the battery into microcontroller IC1’s supply, bypassed by 1000μF and 100nF capacitors. These also provide power to the LCD. The diode means that the microcontroller’s supply does not dip during the brief bursts of current draw during cable pulse testing. Display LCD1 is a 20x4 character LCD panel that has its supply fed directly from pin RD6 of IC1. The signal from RD6 is also connected to the gate of Q1, siliconchip.com.au Parts List – USB Cable Tester 1 double-sided PCB coded 04108211, 130mm x 102mm 1 green double-sided PCB coded 04108212, 134mm x 30mm (front panel) 1 laser-cut acrylic bezel to suit LCD [Cat SC5970] 1 140mm x 110mm x 35mm plastic instrument case [Jaycar HB5970, Altronics H0472] 1 3xAA battery holder with leads (CON1) 1 5-way pin header (CON2; optional, for ICSP) 1 USB-A 3.2 socket (CON3) [Würth Elektronik 63.2213.200] 2 USB-C sockets (CON4 & CON6) [Würth Elektronik 632723.20011] 1 Mini-USB socket (CON5) 1 USB-B 3.2 socket (CON7) [Würth Elektronik 69222103.200] 1 Micro-USB 3.2 socket (CON8) [Würth Elektronik 69262203.200] 1 2-way pin header (CON9; optional, for calibration ➊) 2 2-way pin headers and jumper shunts (JP1 & JP2; optional, for calibration ➊) 1 20x4 LCD module (LCD1) [eg, Jaycar QP5522] 1 16-pin header, 2.54mm pitch (for LCD) 2 1A telecom relays, 5V DC coil (RLY1, RLY2) [eg, EA2-5NU, Cat SC4158] 2 6mm tactile switches (S1 & S2; optional, for calibration ➊) 1 100uH 12x12mm SMD inductor (L1) [eg, Bourns SRR1280-101MCT] 4 M3 x 15mm machine screws 8 M3 hex nuts 8 No.4 x 6mm self tapping screws or M3 x 6mm machine screws 2 6-way stackable headers (for mounting LCD) Semiconductors 1 PIC16F18877-I/P ➋ microcontroller, flashed with 0410821A.HEX (IC1) 2 2N7000 N-channel Mosfets, TO-92 (Q1, Q2) 1 SUP53P06 or IPP80P03P4 P-channel logic-level Mosfet, TO-220 (Q3) 1 1N4148 signal diode (D1) 2 1N5819 1A schottky diodes (D2, D3) Capacitors 2 1000μF 6.3V electrolytic 1 10μF 16V electrolytic 1 100nF 63V MKT or 50V ceramic Resistors (all 1/4W axial 1% metal film except as noted) 1 10kW mini horizontal trimpot 4 10kW 28 1kW 1 220W 1 100W 1 220mW 1% 2W M6432/2512 SMD ➊ The USB Cable Tester will work fine without calibration, so these parts are optional. Still, see the text next month for information about how S1 can be used during regular operation. ➋ IC1 can also be a PIC16F18875 programmed with 0410821B.HEX. Either the I/P or E/P variants will work. A kit is available from the Silicon Chip Online Shop Because of the current semiconductor (& component) shortage, we we concerned that our readers might not be able to build this project. At the time of publication, several of the key components are difficult to source. So we made the decision to purchase all the parts in advance and make a kit available. This not only ensures you can build it, it also greatly simplifies getting the parts. The kit (code SC5966) will come with everything needed to build a complete unit as shown here, except for the case (available from Jaycar & Altronics) and the three AA cells (which are easy to get). The initial price is $110 + postage ($99 + postage for current subscribers) although that could go up if the initial batch sells out quickly. See the shop listing on page 90 or on our website at siliconchip.com.au/Shop/20/5966 Australia’s electronics magazine November 2021  33 microcontroller’s reset line (which is usually pulled up by a 10kW resistor) and circuit ground. Since the calibration menu is only displayed just after a reset, pushing S2 is a simple way to reset the microcontroller and enter calibration mode. S1 is connected between PGD and circuit ground. When it is pressed, it can trigger the IOC interrupt noted earlier and can thus be used to wake up the USB Cable Tester without plugging in a USB cable. Software This main screen is shown when the USB Cable Tester is idle and doesn’t have a cable plugged in. The battery voltage and a countdown timer (until the unit sleeps) are shown. which switches the LCD panel backlight cathode via a 100W resistor. A 10kW resistor also holds Q1 off when the micro is not driving the pin. Thus, when RD6 is low, LCD1 and its backlight are both off. When RD6 is taken high, LCD1’s internal controller is activated and its backlight is switched on. This means that the USB Cable Tester can completely shut off power to the LCD when IC1 is in sleep mode. Six more of IC1’s pins are connected to LCD1 to control it in four-bit mode. This makes the best use of the available pins without needing a separate I/O expander chip. CON2 is an optional in-circuit serial programming (ICSP) header for programming microcontroller IC1. The PGD and PGC pins are also used for USB cable sensing, so a USB cable must not be connected during programming. The PGD pin is also connected to CON9, a two-pin header, via a 1kW resistor. CON9’s other connection is circuit ground. This interface is used to connect to the transmit pin of a TTL-serial interface such as a USBserial adaptor which can be used to enter a dedicated software interface for calibration. No receive pin is provided. Instead, two-way communication is achieved by displaying data on the LCD screen 34 Silicon Chip during the calibration process. Test points TP1, TP2 and TP3 are provided for calibration. These connect to circuit ground, the positive microcontroller supply and the positive end of the 220mW shunt, respectively. JP1 and JP2 are also used only for calibration. When bridged, JP1 connects the upstream and downstream facing VBUS lines. Similarly, JP2 connects the upstream and downstream facing GND lines. When fitted, they leave only the relay contacts and shunt resistance in the current test circuit. Thus, the resistance of the relay contacts can be measured and entered into the calibration settings. This value is then subtracted from cable readings to give a true value. S2 is also intended to be used for calibration. It is connected to the The PIC16F18877 is a reasonably well-equipped microcontroller, and we’re using several of its internal peripherals to provide the features needed. The software loaded into the chip starts by initialising several of its internal peripherals. This includes setting most of the I/O pins as inputs with internal pull-ups, used to sense cable connectivity. It also sets up the UART (serial) receiver and PWM output for the buck converter, plus the seven I/O pins associated with the LCD. Timer (T0) is configured to fire an interrupt every 262ms (approximately four times per second). This is a reasonable rate for quick screen updates while still allowing the display to be legible. The timer is used to display a startup screen for around seven seconds. During this time, if an ESCAPE character is received on the UART, the calibration is started and a menu is displayed on the LCD. The calibration runs until either the microcontroller is reset or a Ctrl-C code is received on the UART. Otherwise, the UART is disabled after seven seconds, and the main ‘idle’ screen is displayed. A subroutine is called after 10 seconds of the idle screen to put the USB Cable Tester into low-power sleep mode. The Tester automatically runs tests as soon as a cable is plugged in. This known-good cable is identified as USB 2.0 compatible with no problems and a voltage drop of 116mV at 1A. Australia’s electronics magazine siliconchip.com.au GND GND VBUS DP DM TXP1 TXM1 RXP1 RXM1 TXP2 TXM2 RXP2 RXM2 1 VBUS DP 1 2 DM 2 3 TXP1 RXP1 3 4 5 TXP2 6 5 TXM2 RXM2 4 4 3 RXM1 RXP2 4 3 TXM1 5 6 6 5 6 Table 1: this data is stored in the microcontroller as arrays of 18 bytes, making up 144 bits (18 x 8). These correspond to the connection combinations that might be detected. It is compared with the data gathered during cable testing. This involves shutting down the peripherals mentioned earlier and setting low all the pins associated with the LCD. This reduces the quiescent current as it avoids leakage from any floating input pins. The upstream facing ports are pulled to a low level, and the downstream facing ports remain as inputs with pullups. Thus, any cable plugged in will pull one or more of the downstream facing port pins low. The IOC flags are set to allow a pin change to wake up the micro. Just before engaging sleep mode, the pins are checked one more time; if a cable is detected, sleep is bypassed. While unlikely to occur with so many pins, it is possible for a pin change to be missed, hence the reason for the double-check. When a pin change is detected (which could include a press on S1), the micro wakes up and initialises all the peripherals again before returning to the main idle screen. Whenever the micro is awake, it uses the timer to perform tests about four times per second. The results of the test dictate what is displayed. The idle screen is shown if no connection is detected; this also displays the battery voltage and a countdown timer until sleep occurs. The tests work simply. Each pin is typically set as an input with a pull-up. One at a time, in turn, each pin is pulled low and the states of the other pins are tested. The wires in the USB cable connecting pins in downstream and upstream sockets result in other pins being detected as being low. The tests are done in three phases. One phase simply checks for connections between the pins associated with the downstream facing port. A second phase checks the upstream facing port. For the most part, these should show no connections, except perhaps for the cable shield and USB-ID pin. USB-ID is grounded on USB OTG cables to indicate that the equipment connected to what would normally be a ‘device’ needs to behave as a ‘host’. Depending on how the cable is wired, there might also be a connection between the cable shield and ground. Any other connection within an A faulty cable is quickly identified; in this case, the GND wire is detected as open circuit (1-, Opens:GND) and naturally, it has no useful current-carrying capacity on its power lines. siliconchip.com.au Australia’s electronics magazine upstream or downstream port likely indicates a cable fault. So if one end of a cable is plugged in, any of these sorts of problems that are detected are displayed on the LCD screen. The third test phase is a complete ‘matrix’ analysis of every combination of downstream facing port pin and upstream facing port pin. This is turned into a cable-specific signature that is compared with a list of signatures corresponding to known cable types. Some cable types have multiple signatures. For example, the reversible nature of USB-C means that there are two equally valid signatures for a USB 3.2 cable. Table 1 shows what connections are expected for each cable type. An exact signature match means that the cable is a known type and displayed as such. An inexact match is shown as the nearest match and the differences are detected. For example, the LCD might indicate that a USB 2.0 cable is detected, but with the D+ line open; such a cable may be suitable for a power-only application but will be no good for data transfer. A simplified version of the decoding would work as follows. • Power-only cable: just the red points in Table 1 detected. • USB 2.0 cable: the red and mauve points are detected. • USB 3.2 cable (Gen 2x1): as for USB 2.0, plus any one of the four remaining groups of connections. • USB 3.2 cable (Gen 2x2): as for USB 2.0, plus either all the green points or all the orange points. November 2021  35 How we decided on which USB sockets to use We’ve spent a great deal of effort to make sure that the sockets we are using for the USB Cable Tester are durable and functional, as well as being hand-solderable. The latter is actually quite a tricky problem, especially for the USB-C parts. USB-C packs a lot of pins into a small connector. Since there are two rows of pins in the connector, breaking them out into two rows at the PCB makes sense. But having two rows of PCB pins will mean that the ‘bottom’ row cannot be surface-mounted, as there would be no way to access them from above. They’d be covered by the ‘top’ row of pins. Since one row of pins goes through holes on the board, soldering them will be slightly easier. But we think these are the finest pitch through-hole and SMD parts that we have used in any project. You’ll need to have the correct gear (including a magnifier and a syringe of flux paste). Therefore, soldering these parts is the trickiest part of constructing the USB Cable Tester despite our best efforts. Fortunately, the fullsize USB-A and USB-B parts are simple through-hole devices. We looked at utilising pre-built USB breakout boards, but they would have substantially increased the size of the final unit and cost quite a bit more too. The mini-USB socket, CON5, is a part we’ve used many times before. Since there is no USB 3.2 variant of this connector, a standard USB 2.0 part is adequate. The micro-USB part is small too, but not much different from the mini-USB socket. They both only have a single row of pins. The good news is that you can use the circuit itself to test that the sockets are soldered correctly. We’ll go into more detail during the construction, but briefly, we can use the existing hardware and logic to probe for any shorts in the socket soldering. A short in the socket soldering will appear to the USB Cable Tester like a fault in the cable, even if it only occurs at one end. Thus, we will advise an unusual order of construction, so that the USB Cable Tester’s microcontroller can run its tests during construction, well before it is complete. That way, you can take your time and check your work both visually and electrically to ensure that you end up with a functioning USB Cable Tester. A12 A11 A10 A9 A8 A7 A6 A5 A4 A3 A2 A1 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 GND RX2+ RX2− VBUS SBU1 D− D+ CC1 VBUS TX1− TX1+ GND GND TX1+ TX1− VBUS CC1 D+ D− SBU1 VBUS RX2− RX2+ GND GND TX2+ TX2− VBUS VCONN SBU2 VBUS RX1− RX1+ GND GND RX1+ RX1− VBUS SBU2 D− D+ CC2 VBUS TX2− TX2+ GND B1 B2 B3 B8 B9 B10 B11 B12 B12 B11 B10 B9 B8 B7 B6 B5 B4 B3 B2 B1 B4 B5 B6 B7 The two rows of closely spaced pins used in USB-C type plugs and sockets demand a tight pin pattern on the PCB. The part we have chosen will be the most challenging part of this project to solder, and we doubt there is anything easier to hand-solder available. Any additional connections detected beyond these constitute some form of short-circuit fault. When a cable is detected, the current pulse test is also performed once every five seconds. This is only done periodically to reduce battery drain and relay wear. The first test is done about half a second after detection, to allow time for the cable to be fully inserted. For this test, the micro sets all the pins associated with connectivity testing as inputs and closes the relays to complete the power circuit. The reference for the ADC is set to the 1.024V FVR (fixed voltage reference). Being a 10-bit ADC, each digital step then corresponds neatly to 1mV. The micro ramps up the PWM signal to Q3 while monitoring the voltages at various points along the VBUS and GND wires of the cables, including just 36 Silicon Chip above the current measuring shunt, which allows the test current to be determined. The ADC is sampled 16 times at four points over several PWM cycles to compensate for the relatively high amount of ripple in the applied voltage. At 100mA, 500mA and 1A, the voltages are stored. If the measured voltage rises above 1V (at any point in the cable) at any time, the test is cut short. The 1.024V reference used for these measurements puts an upper limit on what can be meaningfully measured. Another reason for cutting the test short is that it avoids a high load on the batteries. With a fixed 1A output, there is actually a greater load on the batteries when a high resistance cable is tested; this part of the circuit behaves much like a current source. In any case, a cable dropping Australia’s electronics magazine anywhere near 1V is not going to be of much use. The USB Cable Tester then displays the results from the highest test reading, including voltage drop and calculated cable resistance. When the cable is unplugged, the USB Cable Tester returns to the idle screen and counts down its timer to enter sleep mode unless another cable is plugged in for testing. Next month Next month we’ll describe the construction, calibration and use of the USB Cable Tester. We’ll also describe how the USB Cable Tester can check its own construction and assist with finding soldering faults in the SMD USB sockets we are using. See the panel for more information about the sockets and why such a feature will be handy. SC siliconchip.com.au