DSO150 Oscilloscope Module

As mentioned above, as I was rootling around on the manufacturer’s website [1] I noticed the DSO150 oscilloscope kit (see Figure 2). This device is a development of the DSO138, using the same display and microcontroller. However, this model is only available in kit form with an enclosure. The whole user interface apart from the switch to select the input coupling is driven using four buttons and a rotary encoder. The interface is straightforward and it is hard to see how it could be improved. This kit also comes with the SMD components already mounted. There are just a few leaded components that need to be soldered to the board and then the unit can be put together and tested. The instructions are of a high standard and none of this will present any difficulty to a hobbyist with a little experience. Because the device is supplied in kit form it is not subject to the various standards and regulations that apply when ready-assembled units are sold. The lack of a CE mark is not really a problem, as a manufacturer can always add it later to a finished unit.

pic1 7 DSO150 Oscilloscope Module

I spotted the biggest shortcoming in the unit immediately after assembling it, when I first turned it on: the zero level shifts with changes in the power supply voltage. Since once again a circuit diagram is supplied it was easy to find the cause: unfortunately in the DSO150 the linear regulator to stabilize the supply for the input amplifier has been omitted! The offset for the ADC input is set using just a Zener diode, which does not provide a very well-regulated voltage. As before, you get what you pay for. Fortunately, however, there are leaded series resistors in the positive and negative supply lines, and it is a relatively easy job to replace these with 78L05 and 79L05 voltage regulators. The supply for the Zener diode is now moved to the regulated 5 V rail, and the zero level is stable.

As in the DSO138 the vertical offset is done in software in the microcontroller. The full-scale drive amplitude is 640 mVPP, and so only about 20 % of the ADC’s available input range is used. Compared to the DSO138 interference is much less visible on the DSO150, and occurs less frequently. It is also possible to add an external trigger input to the DSO150.

The unit is an excellent gadget for experimenting with and could make a good secondary oscilloscope for a hobbyist. After a bit of modification to the power supplies it works perfectly acceptably (see Figure 3), especially when you bear in mind that it only costs about £25 (US$30) including shipping from China. There is a chance that you will end up paying VAT and other import charges on top of this, as its cost may exceed the duty-free threshold, at least in Europe.

www.elektormagazine.com

10. June 2019 by sam
Categories: Dev Kit | Comments Off on DSO150 Oscilloscope Module

DSO138 Oscilloscope Module

As Figure 1 shows, the DSO138 module consists of a circuit board with a color LCD attached. Soldered to the board are all the user controls (slide switches and pushbuttons) and sockets for signal input and power. On the left-hand side you 55can see that the selection of input coupling and vertical scale is done in the simplest possible way using slide switches. Unfortunately the quality of these switches (at least in the unit I received) is so abysmal that changing the gain setting often results in the complete disappearance of the signal. The buttons on the right control other functions, and the user interface is generally well-designed and intuitive.

pic1 6 DSO138 Oscilloscope Module

The color LCD panel has a resolution of 320×240 (QVGA) and the visible area has a diagonal of 2.4 inches (6.1 cm). Some low-end desktop scopes, much more expensive than this device, have a display with the same resolution, and so the device should be good enough at least for my application. Happily the device comes with a circuit diagram, which is more than you get with many items of test equipment these days. As a result it was easy to see how to add an external TTL trigger input. The module operates from a 9 V power supply, drawing around 100 mA. A simple inverter circuit generates a negative rail, which, like the positive rail, is stabilized using a linear regulator. The microcontroller, an STMicroelectronics STM32F103C8 based on an ARM Cortex-M3 core, is powered at 3.3 V via another regulator.

The design uses the fast ADC built in to the microcontroller. It offers a resolution of 12 bits and an input voltage range from 0 V to 3.3 V; the middle of this range, 1.65 V, corresponds to the zero position on the display. However, the display is already clipped when the unit is fed with a sine wave of much less than the maximum possible amplitude. This, in conjunction with the high resolution of the converter, allows the vertical offset to be adjusted. In other words, the vertical offset is not implemented in the analog preamplifier circuit, but instead is subtracted from the sample values in software.

The fact that only a relatively small part of the ADC’s input range is used obviously makes the device more susceptible to interference and spikes, and these sometimes also affected the horizontal deflection. I tried to fix this by improving the decoupling of the circuit at various points, but without success. ‘You get what you pay for’ as the cliché has it, but in this case you do get quite a lot of hardware for your money, and its minor deficiencies can be tolerated. The DSO138 is not just available as a ready-made module: if you fancy doing some soldering you can also get it slightly more cheaply in kit form, with the SMD components already fitted.

10. June 2019 by sam
Categories: Dev Kit | Comments Off on DSO138 Oscilloscope Module

The PGA2311

For those who are interested, in Figure 3 we have drawn the internals of the PGA2311. It is immediately clear that we have two (because it is stereo) amplifiers, the amplification of which are set by a set of (serial) control inputs. The control signals (CS, SCLK, SDI and MUTE) come from microcontroller IC1.

A brief remark about the connection of ZCEN (Zero Crossing Enable). The PGA2311 has a zero-crossing detector;
the idea behind this is (when this function is enabled) that the change in volume only takes effect after the next positive-going zero-crossing of the audio input signal. In this way spurious audio artifacts as a consequence of the change in volume can be minimized. On the circuit board ZCEN can be connected via a jumper to either +4.7 V or ground, so that you have the choice whether to enable or disable this functionality.

For a more detailed description of the inner workings of the PGA2311 we refer you to the datasheet [1]. Between the microcontroller and the PGA2311 we provided test points on the circuit board for the serial control signals (CS, SDI, SCLK and MUTE).

The input and outputs go via resistors of 47 Ω (R1 through R8) to two different connectors. A pair of so-called stack-through-connectors (K2 and K3) connect the inputs and outputs of the DAC and the volume control (in the same was as is done for the power supply).

The inputs and outputs of IC2 also go to two 3.5-mm audio connectors (K4 and K5) on the circuit board. This means that the unprocessed DAC signal is available using a 3.5-mm jack (K5) or via the cinch connectors on the DAC board. The output signal from the volume control is available on K4.

When the board is used as an ‘independent’ volume control, K5 is the input for the PGA231 and K4 is the output. The 47-Ω resistors protect the outputs from capacitive loads and create a separation between the various connectors.

pic1 5 The PGA2311

www.elektormagazine.com

10. June 2019 by sam
Categories: Components | Comments Off on The PGA2311

Keysight 33522A Function Generator

We have there are no issues of importance with the reference device. This signal generator is what it sets out to be and it functions in the same way too. However, for professional equipment of this rank you should expect to pay a good €3,000 / £2,775 / $3,580.

Sine-wave signals are generated up to 30 MHz with a sample rate of 250 MHz. The ratio of 12 % allows more than eight grid points per period. The D-to-A converter provides 16-bit resolution. The maximum amplitude amounts to 20 Vpp (Hi-Z) up to 30 MHz without any restriction.

Squarewave signals and pulses are possible up to 30 MHz. Rise and fall times are fixed for square-wave at 8.4 ns; for pulses you can also select larger values.

Ramps are limited to a maximum frequency of 200 kHz here.

Using the DC function you can produce an adjustable DC voltage in the range ±10 V.

The PRBS (pseudo-random bit stream) function is available exclusively on the Keysight product. A noise signal of this kind is very handy for determining bandwidth.

The Keysight generator is the only device that has the sync output on the front panel. There is no frequency restriction, nor any visible jitter.

As to modulation, using the external input signals can be modulated with frequencies up to 100 kHz.

Frequency counter

All three products under review have an integrated frequency counter with a bandwidth of up to 200 MHz. I measured
the sensitivity of each at three different frequencies. Table 1 shows the results. The levels in dBm are valid for Hi-Z (= high-impedance). The last line shows additionally the maximum frequency up to which each counter still works.

All three products can switch between AC and DC coupling. It is also feasible to activate a low-pass filter and set the trigger level. On the PeakTech device you can set the sensitivity in three steps, although this requires a higher level than on the other two examples. Furthermore, only six settings were given — a drawback versus the Rigol product with seven and the generator from Siglent with yet eight positions. There’s another restriction with the frequency counter on the PeakTech device: only the direct frequency is displayed, whereas both of the others can indicate the center, minimum and maximum values. The Rigol function generator shines out as very practical,
with its counter input on the front panel.

For me what is missing on all three devices is a switchable 50-Ω terminating resistor. To avoid signal reflections you will need to use a through-pass termination where appropriate.

Signal quality
For comparing signal quality on each product I set up sinewave signals at 1 and 10 kHz with a level of 1 V RMS and measured them with an audio analyzer. Table 2 indicates the harmonic distortion and Table 3 the headroom between the wanted signal and the combination of distortion and noise. Although the harmonics differed by more than 10 dB, from my point of view not one of the generators is fit to serve as a low-distortion signal source for audio measurements without additional
filtering.

Spectrum test
The spectra of a 10 MHz signal with a level of 0 dBm were measured using a spectrum analyzer. When you make direct measurements the bulk of the first harmonic is attributable directly to the spectrum analyzer account, so a 20 MHz high-pass filter was connected in series. The latter attenuates the 10 MHz signal, making it possible to raise the sensitivity of the spectrum analyzer.

www.elektormagazine.com

10. June 2019 by sam
Categories: Products | Comments Off on Keysight 33522A Function Generator

Rigol DG1032Z Function Generator

This comes with a ‘Quick Guide’ but no CD. The ‘proper’ operating instructions are of course downloadable from the Rigol website and there is even a German version. Plus point: the fan noise is moderate. At around €560 / £520 / $670 the function generator is a good 20% dearer than the other two devices.

Sine-wave signals are generated up to 30 MHz with a sample rate of 200 MHz. With a ratio of only 16.7 % an extra grid point per period is produced. If the impedance is set to 50 Ω (not as Hi-Z) the amplitude can be defined only in dBm. With open-circuit (unterminated) outputs the maximum amplitude is once again 20 Vpp up to 10 MHz, dropping to 10 Vpp above this frequency. Here again there is a ‘Harmonics’ function for setting harmonics, albeit restricted to eight harmonics.

Squarewave signals are limited to 15 MHz. You can adjust the duty cycle within the range 19 to 81 %, with an even broader range at lower frequencies. The rise and fall times are given as <10 ns; my oscilloscope confirmed this as around 7 ns. There is an error in the data sheet [3], where 25 MHz is given as the maximum frequency. Pulses are again feasible up to 15 MHz. You can additionally make the rise and fall times larger than the minimum values of 10 ns. Ramps can be generated with a frequency up to 500 kHz maximum. On the rear panel of the Rigol product two independent Sync signals are provided for Channels 1 and 2. No jitter is detectable, even at 15 MHz. This works for sine-wave signals up to the maximum frequency of 30 MHz. Here again both channels are apparently identical — I could not detect any restrictions. Modulation: at 1 MHz, the maximum modulation frequency is extraordinarily high. The reference output provides a 10 MHZ square-wave signal with 1.5 VPP into 50 Ω with steep flanks. Here too differential signals can be generated very easily. Beside the Copy command there is a Tracking Menu Next to Copy command there is a Tracking Menu, in which you can set the parameters to be linked, even by using an offset value. pic1 4 Rigol DG1032Z Function Generator

10. June 2019 by sam
Categories: Products | Comments Off on Rigol DG1032Z Function Generator

Siglent SDG1032X Function Generator

This time the user instructions do match up with the actual device, thank goodness. Once again the fan noise is definitely audible, even if not nearly as intrusive as on the PeakTech. In price terms, at around €450 / £420 / $540, this generator is in the same ballpark.

Sine-wave signals are generated up to 30 MHz with a sample rate of 150 MHz, which results in a ratio of 20 %. Here again the maximum amplitude is 20 Vpp up to 10 MHz and 10 Vpp above this — both measured into Hi-Z. A remarkable setting option, which I had not observed previously, is the ‘Harmonic’ function. If you enable this you can add defined harmonic components with adjustable amplitude and phase to a sinewave signal. This not only lets you check how you can avert defined harmonics in circuits but also even eliminate a harmonic produced in an amplifier by adding a signal of equal amplitude with 180° phase.

For squarewave signals you have virtually no restrictions. Even at 30 MHz you can still adjust the duty cycle in the range 41 to 59%. At lower frequencies the adjustment range of the duty cycle is extended significantly.
Rise and fall times of 4.2 ns are indicated.

When it comes to pulses, if you select the ‘Pulse’ curve shape you can even preset the rise and fall times — an extraordinarily useful feature. Compared to the square-wave function, the minimum times are slightly increased by 16.8 ns. Using ‘Delay’ you can shift the signal of the second channel in a relative manner, similar to adjusting the phase of the sine-wave signal.

Ramps up to 500 kHz are possible.

Using the DC function you can generate an adjustable DC voltage in the range of ±10 V.

On the rear panel of the case there is a Sync output that you can enable in the Sync menu and link this with one of the channels. However, in comparison with the signal channel this displays jitter at higher frequencies and is therefore of limited use. On the data sheet [3] its maximum frequency is given as 1 MHz, although the function continues to work up to 10 MHz.

On this Siglent device both channels appear to work identically — at least I could not detect any restrictions. You can even add the signals from the two channels.

Regarding modulation, the maximum internal modulation frequency is 20 kHz, as against 50 kHz for the external input.

The reference output provides a square-wave signal (filtered by an R-C low-pass) of 10 MHz at an amplitude of 1.5 VPP into 50 Ω.

Differential signals can be generated very simply. Next to the Copy command there is a Tracking Menu, in which you can set the parameters to be equiplinked, even by using an offset value.

pic1 3 Siglent SDG1032X Function Generator

10. June 2019 by sam
Categories: Products | Tags: | Comments Off on Siglent SDG1032X Function Generator

PeakTech 4125 Function Generator

The operating manual supplied on CD clearly relates to an older version of the device (Figure 1). The front panel of the newer version in fact has a USB connector for an external hard drive. On the other hand the manual still shows the connector on the rear panel, which looks completely different. The reference inputs and outputs are labeled ‘20 MHz’ but customarily these are 10 MHz. It’s a shame that the manufacturer does not (yet) offer a newer version of the user manual.

Irritation begins at switch-on: the noise of the fan is loud and intrusive, arising not from any defect but more probably from excessive rotational speed, providing abundant ventilation at the price of peace and quiet. For around 35€455 / £420 / $540 one might expect something different.

Operation, as on all such devices, uses soft-keys surrounding the LCD together with a rotary encoder switch and a keypad. BNC connectors are used for the signal inputs and outputs. The description that follows mentions the means of performing individual functions only when they are particularly good or terrible solutions.

Sinewave signals can be adjusted up to a frequency of 25 MHz, which accounts for an acceptable 20 % of the sampling frequency. There’s an informative discussion on sample rates and ratios at [1]. This provides sufficient reserve for the use of a simple output filter. The maximum amplitude amounts to 20 Vpp up to 10 MHz and 10 Vpp above this. The offset possible is ±10 V. All specifications relate to Hi-Z (= high impedance, unterminated) outputs.

Squarewave signals are limited to 5 MHz and additionally to a duty cycle of 50%. The maximum frequency is too low to clock modern microcontrollers or digital circuitry for test purposes.

Pulses are also limited to 5 MHz. Unlike the duty cycle, the <10 ns rise and fall times indicated are not adjustable. Ramp signals can be generated up to 1 MHz. There is no Sync-Output for triggering oscilloscopes externally if you alter the amplitude and offset. The handbook mentions this function but I was unable to find it. The PeakTech product does indeed provide two channels but they do not have the same characteristics. You can use all types of modulation and the frequency sweep facility on Channel 1 only. In reality this is not a serious restriction. Differential outputs for audio or rapid digital signals are certainly possible but are somewhat complex to set up. You can copy the characteristics of one channel over to the other but you cannot link them. For instance, as soon as you change the frequency, you have to transfer this across. With modulation the maximum internal modulation frequency amounts to 20 kHz, which makes it suitable for the audio range. Reference frequency: the reference output (10 MHz in fact) provides a trapezium-shaped signal of 1.6 Vpp into 50 Ω. pic1 2 PeakTech 4125 Function Generator

10. June 2019 by sam
Categories: Products | Comments Off on PeakTech 4125 Function Generator

Failed Fuse Indicator

If the device went bang, fizzed or simply did nothing when turned on, you could assume that it has cost you a fuse (or several). If, with it still plugged in, the main circuit breaker has not cut out and there is no overall power failure, the other items connected to the same wiring circuit will at least still work.

Some instrument fuses are filled with silica sand, however, in order to extinguish any arcs arising when the fuse element (a length of fine wire) blows and cuts the current flow of any short circuit as rapidly as possible. This is good for the appliance we want to be protected but has the disadvantage that it’s hard to see whether the fuse has actually blown or not. An optical indication would be no bad thing, particularly for devices with which this happens more often or not. This is easy at mains (line) voltage: a small bulb can be wired in parallel with the fuse. However, the failure indicator described here uses an LED to light up when the fuse blows and will also work at lower voltages.

pic1 1 Failed Fuse Indicator

A failed fuse indicator needs to be simple, cheap, reliable and small. Ergo it comshould also require not many components. The circuit in Figure 1 fulfills all of these criteria. Its operation is supremely simple: an LED is wired in parallel with the fuse along with a little electronic wizardry. Under normal conditions the LED is shunted by the intact fuse and remains unlit. If the fuse blows, the operating voltage is available via the load connected to the defective fuse. The parallel LED can avail itself of this and light up.

The electronics needed can remain straightforward. The first consideration is that the failure indication should operate regardless of DC polarity and on AC as well. This means having the LED plus current limiting circuitry driven by a bridge rectifier. To avoid too much voltage drop and maintain operation at low supply voltages, Schottky diodes are employed for D1 to D4.

The only other thing we need to ensure now is that the current through the LED is not only limited but also maintained unaltered across a broad voltage range. This works best using a constant current source. The simplest option is an N-channel barrier layer FET with a resistor between the Gate and Source. The current flowing is dependent mainly on the slope of the FET or its DC voltage at a particular current across the resistor. However, one has to keep in mind the maximum voltage between Drain and Source and the maximum power dissipation. With the BF245 FET you can reach 30 V. The failure indication can therefore be used with fused voltages of between 3 and 30 V.

Performance considerations now: according to the datasheet [1] a BF245 can tolerate ambient temperatures up to 75 ºC 300 mW, getting significantly hot and bothered nevertheless. Better to keep it below half of that. The sub-type BF245A produces a current of 4 mA when R1 = 0Ω. This way any LED will light up clearly and total power dissipation
for T1 remains in the region of 100 mW even at 30 V. With the BF245B the figure is around 10 mA and finally around 18 mA with the BF245C. This makes things brighter. With R1 you can reduce the current through the LED at will. For the BF245B, a current of approximately 7 mA is obtained with a value of 68Ω for R1.

www.elektormagazine.com

10. June 2019 by sam
Categories: Labs | Comments Off on Failed Fuse Indicator

Magnetic Memory

Today we have lots of options for storing information electronically like SRAM, DRAM, Flash and EEPROM. Random access memories (RAM) like SRAM and DRAM store information using capacitance which makes them very fast without a lot of extra complication. But their downside is that all of their information is lost when the memory isn’t powered. Nonvolatile memory like Flash or EEPROM keep their contents while powered off at the expense of speed and requiring more complex access methods like memory paging. Enter magnetic memory which combines the qualities of both kinds of memories.

One of the first incarnations was magnetic core memory that was used in the 50s and 60s because it was so much faster than other storage methods like Williams tubes which were based on CRTs. The name comes from the fact that the memory was constructed using magnetic toroids called cores with wires threaded through them for read and write operations. Each toroid would be set to a 1 or 0 by controlling its magnetic direction although some extra circuitry was required because every read operation would clear the core’s magnetization. This storage method proved to be very reliable and could even withstand an EMP pulse. But the downside of the toroids was that it was very difficult to manufacture and that made it relatively expensive so it was eventually superseded by SRAM which became available in the 60s. The Computer History Museum in Mountain View, CA, has a fine collection of magnetic-core memory devices on display, also check out their coasters with an MM image — buy one from the Gifts Shop (Ed.) A modern version of magnetic memory is magnetoresistive RAM (MRAM) which has been developed over the last 30 years. Early versions used ferromagnetic plates with an insulating layer between them.

One plate has a permanent magnetic field and the other plate’s magnetization is writeable to store a bit. The logic state of the cell is then determined by measuring its resistance which changes depending on the magnetic
orientation of the writeable plate to the permanent plate. Newer versions of MRAM use the spin transfer torque of electrons in their memory cells to lower their power consumption.

Another variation is ferroelectric RAM (FRAM) which has been developed in parallel with MRAM. Conventional DRAM memory uses one transistor and one capacitor per memory cell which can be made into a FRAM cell by adding a ferroelectric material in place of the regular dielectric. The ferroelectric material Comchanges the normally linear behavior of the cell to one that has magnetic hysteresis which gives FRAM its nonvolatile properties. Writing data is pretty straightforward but reading a FRAM cell requires that a transistor put the cell into a known state and then the cell is monitored to see if the ferroelectric material causes the current to flow. This will also clear the cell so it needs to be rewritten. Both MRAM and FRAM are being produced today by various companies. MRAM devices tend to focus on density and speed whereas FRAM concentrates on low power. Either way, they are very interesting parts!

pic1 Magnetic Memory

www.elektormagazine.com

10. June 2019 by sam
Categories: Labs | Comments Off on Magnetic Memory

Types of rechargeable battery

First some terminology. The word ‘battery’ comes to us via a French word meaning an array of artillery weapons, to which Benjamin Franklin compared his experimental array of Leyden jars. So strictly speaking a ‘battery’ should comprise more than one cell, but this distinction is only rarely made in everyday usage. The English term ‘battery’ covers both rechargeable and non-rechargeable types, the more precise terms ‘primary cell’, ‘secondary cell’ and ‘accumulator’ no longer being in common use. Now, with that out of the way, we can look at some of the more common battery types used today.

At over 160 years old, the lead-acid battery (see Figure 1) is certainly the oldest type of rechargeable battery
that still finds practical use. The cells, which typically have a nominal voltage of 2 V, have been used in automobiles since around 1900. In that year the celebrated Lohner-Porsche Semper Vivus [1] was unveiled, the first ever hybrid car (are you listening, Toyota?). Another early application was in telephone and telegraph offices. The lead-acid battery is still difficult to beat in terms of price and robustness, and so, despite its heavy weight and other disadvantages it is still used to power the starter motor for internal combustion engines.

The NiCd battery also has a long history, going back over 100 years. In particular, in the 1980s, small NiCd cells were the most popular rechargeables for powering consumer electronic devices despite their nominal voltage of only 1.2 V. Their ability to deliver large currents made them the battery of choice for cordless tools and radio-controlled models. Also, Toyota Prius models up to and including the Prius III use this battery technology for their hybrid drive. The biggest disadvantages are poor environmental credentials (cadmium is toxic) and the notorious ‘memory effect’.

pic1 1 Types of rechargeable battery

The better is the enemy of the good: low-cost NiMH batteries (see Figure 3) came to replace NiCd batteries in consumer devices, having first been manufactured on an industrial scale some 35 years ago. Offering the same nominal voltage of 1.2 V they were an ideal substitute. They are also by and large more environmentally friendly, do not exhibit the memory effect, have reasonable energy density and are economical. On the other hand, they are a bit trickier to charge. Since about ten years ago NiMH batteries with very low self-discharge have been available.

More recent rechargeable battery technologies are based around the element lithium. These chemistries allow the construction of lightweight batteries with a very high energy density, which are important factors
in mobile applications such as laptops, tablets and smartphones as well as in electric vehicles. The three main types of practical importance are as follows: LiPo (‘lithium polymer’: see Figure 4) whose compactness makes them ideal for mobile applications; and LiCoO2 (see Figure 5) and LiFePO4 (see Figure 6) for electrical drives. Their nominal voltages, at 3.7 V for LiPo, 3.6 V for LiCoO2, and 3.2 V for LiFePO4, are significantly different from those of other chemistries. They do not exhibit the memory effect and have very low self-discharge. On the other hand they are sensitive to environmental conditions and there are strict rules to obey when charging and when discharging them. As well as a range of cylindrical formats lithium batteries are available in customized packs for mobile devices and in prismatic housings with higher capacity. It is interesting to note that the current range of Tesla electric vehicles use ‘batteries’ comprising many thousands of type 18650 cylindrical cells. The RAM (rechargeable alkaline manganese) battery (AA size shown in Figure 7) is certainly the newest kid on the block in the world of rechargeable batteries. These alkaline secondary cells have a nominal voltage of 1.5 V and so are ideal as direct replacements for zinc-carbon or alkaline manganese primary cells. However, they are only suitable for use in applications with low discharge current, must not under any circumstances be subjected to deep discharge, and require a special charger: they are decidedly not compatible with NiCd or NiMH chargers!

pic1 2 Types of rechargeable battery

05. June 2019 by sam
Categories: Updates | Comments Off on Types of rechargeable battery

← Older posts