Fullscreen HTML5Med Res (??MB)HTML4

Using Electronic Modules with Jim Rowe ZPB30A1 Module - 60W Programmable DC Load - Battery Capacity Tester This programmable constantcurrent DC load can be used for testing power supplies or checking the capacity of storage batteries. It is essentially self-contained and delivers good value for the money. T he ZPB30A1 module carries the brand name Zhiyu, but it seems to be made in China by a firm called AoSong ELE Co Ltd. As you can see from the photos, it has two PCBs, with the smaller one (69 × 36mm) mounted above the larger main PCB that measures 100 × 69mm. The 50 × 50 × 23mm heatsink is at the rear of the larger PCB, along with its associated cooling fan on the finned side, for cooling the main load transistor. The fan extends past the rear of the main PCB by about 11mm. Mounted on the flat front of the heatsink is the power Mosfet that acts as the controlled load (in the centre), with a thermistor to its left used for sensing its temperature. To its right is a dual schottky diode that protects the power transistor from damage due to reversed voltage polarity. The smaller PCB is the control and display board, or the main ‘user interface’. Its four-digit 7-segment LED displays the voltage, battery capacity or various control and error messages. In contrast, the three-digit LED display below it is mainly used to show the current flow. Six additional green 3mm LEDs indicate which parameter is being Features & Specifications ∎ Test modes: programmable constant-current DC load (“Fun1”) or battery capacity tester (“Fun2”) ∎ Maximum dissipation: 60W ∎ Operating voltage range: 1-30V (separate 12V 500mA supply required) ∎ Operating current range: 0.1-9.99A in steps of 0.01A (10mA) ∎ Rated current measurement accuracy: ±(0.7% + 10mA) ∎ Test termination voltage range: 1-25V ∎ Voltage measurement: directly at the P+ and P− terminals or remotely for four-terminal measurements ∎ Voltage measurement accuracy: ±(1% + 0.02V) ∎ Battery capacity maximum values: 999.9Ah, 9999Wh ∎ Battery capacity test accuracy: 2.5% <at> 0.5A, 1.5% <at> 2A or 1.2% <at> 5A+ ∎ Protection: over-temperature (“otP”), transient over-power (“oPP”), overvoltage (“ouP”), reverse polarity (“Err3”) and abnormal voltage (“Err6”) ∎ Fan control: automatic, temperature-controlled ∎ Size: 69 × 111 × 57mm ∎ Weight: 270g 62 Silicon Chip Australia's electronics magazine displayed or which 7-segment digit is being adjusted, while a red 3mm LED indicates when the module is running. The function of the 3mm yellow LED function is not explained; it is labelled “L-4” and seems to be a recent addition to the latest (V3.3) version of the module. On the right of the smaller PCB are the two controls. The first is a rotary encoder, which changes modes and adjusts current and voltage values. The second is a small pushbutton used to confirm the module’s current and termination voltage settings and as an on/off control. Currently, the ZPB30A1 module is available from several sources on the internet, including Banggood (www. banggood.com/search/1146280.html) and many suppliers on AliExpress and eBay, ranging from $14.91 plus $7.37 for delivery to $37.53 plus $4.48 for delivery. I ordered one from Banggood at a price near the high end, and after the usual wait, it arrived safely – even though it was only wrapped in bubble wrap inside a plastic bag. What it does The module has two modes of operation. One is to serve as a programmable constant-current DC load, while the other is to test the capacity of storage batteries like Li-ion, Nicad or lead-acid batteries. The basic specifications are shown in the sidebar. The module’s internal circuitry is siliconchip.com.au Fig.1: a simplified block diagram of the ZPB30A1 module. powered by a 12V DC supply that must be separate from the measurement current source. The supply voltage must be between 11V and 13V, delivering at least 500mA. Power is applied to the module via a standard barrel-type DC connector (5.5mm outer diameter, 2.1mm inner diameter) on the left side of the main PCB. When power is applied, the module powers up in whichever of its two operating modes was last used. This mode is displayed in the upper fourdigit 7-segment LED display as either “Fun1” for programmable load mode or “Fun2” for battery capacity mode. The module defaults initially to the Fun1 mode. If you want to switch it to the other mode, you need to switch off the power, wait a few seconds and then hold down the on/off pushbutton while re-applying power. The module then allows you to switch modes using the rotary encoder, after which you press the on/off button again to lock the module into that mode. The module’s testing inputs are on the right-hand side of the main PCB. The small two-way screw terminal block is the main test input connector, with its inputs labelled “P+” and “P−”. The smaller two-pin socket is siliconchip.com.au used for the optional remote voltage sensing, to avoid errors due to voltage drops in the connecting wires. Its pins are labelled “V+” and “V−”. How it works Fig.1 shows a simplified block diagram of the ZPB30A1. I would have liked to show a complete schematic, but all I could find online was a partial circuit (at www.voltlog.com) that had been ‘reverse engineered’ and didn’t cover everything on the main PCB, let alone any of the circuitry on the display/control PCB. Still, all the most important details are shown in Fig.1. The ‘brains’ of the device is an STM8S105K4 microcontroller unit (MCU), shown at lower left. This The rotary encoder on the display PCB is used for changing modes and adjusting the current & voltage values. Australia's electronics magazine responds to the controls on the display and control board shown at upper left and shows the parameter values and testing status on the same board. The load current control circuit is shown at upper right. This uses the W60N10 power Mosfet (Q2) to maintain the load current between the test terminals P+ and P−, under the control of the MCU via the I_SET line. The current is monitored using a 10mW shunt resistor in Q2’s source connection; op amp IC4b compares the voltage drop across the shunt with the control voltage from the MCU. You can see a simplified version of the remote differential voltage sensing input below the current control circuit, using op amp IC4a. Its output is taken to the AIN2 analog input of the MCU. The thermistor mounted next to the Mosfet on the heatsink is shown below the remote voltage sensing input in Fig.1. The TEMP SENSING line from the thermistor goes to the MCU’s AIN0 analog input. Note that the micro doesn’t have a way to monitor the actual load current – there is no connection from the 10mW shunt to the micro. The actual load current will equal the set current almost all the time; if the source cannot March 2023  63 The “main” PCB measures 100 x 69mm and has the heatsink mounted on it. It’s also where the majority of the components and power socket are located. supply enough current to meet the target, the voltage will drop to near-zero, triggering the under-voltage alarm. So it is a safe assumption. The cooling fan’s driver Mosfet, Q3, is controlled by a PWM (pulse-width modulated) signal from the PD0 digital output pin of the MCU. This allows the MCU to turn on the fan as soon as the thermistor reports that the heatsink temperature has risen significantly, and to increase the fan’s speed as necessary to keep the temperature under control. If the temperature keeps rising beyond a safe level, the MCU turns off the load current and stops the test. The piezo sounder is driven by the MCU’s PD4 digital output pin. This allows the MCU to attract your attention whenever it needs to do so; for example, when a test comes to an end, or it detects an error condition. The module I received did not have the 6-pin header fitted (shown above the piezo sounder); there was just a set of pads and holes labelled G, R, T, L, F and Vc. While there was no mention of these in the sketchy data provided on the Banggood website, when I searched the internet, I found a couple of suggestions that the G, R and T pins could be used for serial communication with the MCU, at a rate of 115,200 baud and with the standard 8N1 protocol. The information I found said that the module only transmitted serial data in programmable current mode (Fun1), containing three-byte messages with the first two bytes representing the voltage while the third byte indicated testing status (1 = OK, 0 = undervoltage alarm). Trying it out The information on using the module provided on just about all of the supplier websites is very vague and quite hard to follow. As a result, you are largely ‘on your own’ when it comes to using it. It’s a matter of trial and error, not made easy by the multiple functions of the module’s controls and LED displays. That is a pity, since it performs surprisingly well when you manage to get it doing what you want. The first thing I did was ensure that my module was set to constant-­ current load mode (Fun1). Then I used the rotary encoder to set the load current for the test. This can be any value between 0.1A (100mA) and 9.99A, in steps of 0.01A (10mA). After this, I connected the module’s P+ and P− terminals to a 0-30V/5A programmable power supply, with one high-resolution bench DMM (digital multimeter) monitoring the current and another monitoring the actual voltage at the P+ and P− terminals. Then I pressed the module’s on/off button to begin testing. I set the power supply to a range of voltage levels (3.30V, 5.00V, 9.00V, 12.00V, 15.00V, 20.00V, 25.00V and 30.00V), and at each voltage level, I set the module to draw a series of current levels. At every current level, I used the bench DMM to set the voltage to precisely the desired level and used the other DMM to check the exact current. The results of these tests are shown in Fig.2. As you can see, in each case, the applied voltage remained constant over a wide range of current levels. That remained true to a point where either the module stopped the test Fig.2: I tested the load on constant-current load mode (Fun1) at a range of different voltage levels. This figure shows the current drawn by the module at those voltages. 64 Silicon Chip Australia's electronics magazine siliconchip.com.au due to the temperature rising above the limit (plots ending in an “X”), or my programmable power supply could provide no more current at that voltage (plots ending in a dot). Just to make sure, I undertook one further test using a different power supply capable of supplying 13.8V at up to 12A. This resulted in the brown plot in Fig.2. This showed that the module could maintain a current just below 5A at this voltage, corresponding to around 68W dissipation – not bad considering that it is rated to handle a maximum of 60W. During these tests, I monitored the difference between the module’s voltage and current readings and those of the two reference DMMs, to get an idea of the module’s measurement accuracy. Its current readings turned out to be less than 0.3% low for currents of 2.0A and above, rising to 1.0% low at 0.5A and 4% down at the lowest current level of 100mA. These figures compare pretty well with the module’s rated accuracy of ±(0.7% + 0.01A). The voltage readings turned out to be less than 0.4% low over the entire range, which is significantly better than the rated accuracy of ±(1% + 0.02V). So the ZPB30A1 module performs well in programmable current load (Fun1) mode. I moved on to checking out its battery capacity/Fun2 mode. Battery capacity testing I fully charged an 18650 Li-ion cell, then set up the ZPB30A1 module in Fun2 mode with a discharge current of 1.0A and a minimum voltage of 3.00V. After connecting the Li-ion cell to the P+ and P− inputs, I pressed the module’s on/off button to begin testing. Since the module doesn’t seem to have any serial output in this mode, I had to record the time and battery voltage the old-fashioned way, using a pen and paper while reading a stopwatch. The results of this first test are shown in Fig.3 (red plot). As you can see, the cell didn’t last all that long at the 1A discharge rate, with its voltage dropping below 3V after only 41 minutes. The module then displayed its capacity as 0.679Ah, close to my calculated figure of 683mAh (1A × 41 minutes ÷ 60 minutes). So its measurement was only about 0.58% low. I recharged the same 18650 cell overnight and set the ZPB30A1 to perform a second test at 500mA. I then spent the next few hours recording the battery voltage every five minutes, again in the old-fashioned way. The results of this second test are shown in the blue plot in Fig.3. It lasted a lot longer this time, with its voltage only reaching just below the test cutoff voltage of 3V after 228 minutes, corresponding to a capacity of 1900mAh. So it’s pretty clear that this particular 18650 battery is only capable of delivering its rated capacity at load currents of 500mA or less. It’s also apparent that the ZPB30A1 is well suited to performing the battery capacity testing role, despite a few minor drawbacks. Summary The ZPB30A1 module performs both its main functions – a programmable constant current load and battery capacity tester – very well indeed, especially considering its modest price. But it does have a few failings, including the lack of good instructions. It’s also pretty disappointing that its serial communications are so limited. Having an adequately documented serial connector that worked in all modes and provided a complete set of information would make it much easier to monitor the load voltage and time for each measurement. Adding a serial port header and supporting MCU firmware should be straightforward and would make things a lot easier, especially when testing a battery’s capacity. Hopefully, the module makers will add this serial port feature to it soon, making it a really handy piece of test gear. Despite that, given its low cost, I still think it is worth getting if you think SC you will use it. Fig.3: battery capacity testing was performed with a fully-charged 18650 Li-ion cell and the module in Fun2 mode. The test was done with a discharge current of 1A (and later at 0.5A) and battery voltage above 3V. siliconchip.com.au Australia's electronics magazine March 2023  65