Power Bank Charge Fee

I welcome everyone who looked at the light. The review will focus, as you probably already guessed, about replacing the guts of one of the popular single-bank PBs with more successful ones and testing the resulting device. Who cares, please, under hood cat. At first, I planned to do a comparative test of 3 popular single-bank PBs (for a company and a small PB on an AA element) and several specialized boards that can give from 1 to 2 amperes. But as a result, the article was very large, because there were a lot of measurements, therefore, in this article I will limit myself to just a small theory on circuitry and operation of the safety switch, testing a specialized SC-0241 1.3A DIY board and assembling a new safety switch. Here are the same subjects:

Power Bank Charge Fee

Testing the electronic fillings of various PBs, perhaps, will be in a separate article, if the desire to test all this economy does not disappear.
Of the available tools, all the same budgetary methods: a DT-832 multimeter with good probes, a 2-digit ammeter, a 3-digit voltmeter and a homemade 4.15 Ohm load resistor.

According to tradition, a little theory, where without it:

I already talked about what PB is, how it connects to devices and works, in an article about Miller ML-102, I don’t like to repeat and will not. I only note that Power Bank (“energy storage”, “power bank”) is a mobile charge, a portable additional battery that can transfer its energy to other devices.

The bulk of industrially produced PBs differ in several ways:
– dimensions / appearance (in the form of a cylinder / parallelepiped / square / for which there is enough imagination)
– material (aluminum / plastic) and body type (collapsible / non-collapsible)
– on / off (always on in standby mode / on-off button / auto off)
– the presence of batteries (pre-installed / self-installation of cans)
– type of batteries used (NiMH, Li-Ion, Li-Pol)
– the number of batteries (single-bank / multi-bank) and their connection (1, 1S2P, 1S3P, 1S4P, 2S2P, 1S5P, 1S6P, 2S3P, 3S2P, etc.)
– capacity (from 1Ah to 20Ah)
– voltage converter (step-up / step-down)
– built-in protection (against short circuit / overcharge / overdischarge / reverse polarity / overheating)
– versatility in banks (protected and / or unprotected) and the ability to simply charge
– input parameters (DC port, miniUSB or microUSB connectors, charge current 0.5A, 1A or 2A)
– output parameters: fixed / adjustable voltage (from 5V USB to 13V DC port) and fixed / adjustable current (0.5A, 1A for smartphones, 2.1A for tablets or 3.5A for “gluttonous” devices)
– number of output ports (1.2 or 3 outputs, sometimes USB 5V and DC port 12V are also present at the same time)
– independence of charge / discharge channels (at the same time charge / return, or only charge or return)
– independence of output channels (several dependent / independent outputs)
– indication of modes (charge / charged / return) and remaining capacity (screen / LED (s))
– additional “lotions” (screen, flashlight, charge indicator, solar panel, etc.)
– additional accessories (power supply / adapter, charging cable, adapters / connectors, etc.)
Most devices have an efficiency in the region of 80-90%. Multibank PBs, as a rule, have higher efficiency in comparison with one / two-bank ones. This is allowed, first of all, with a more “sophisticated” electronic filling (a full choke, a high-quality high-frequency pulse DC-DC converter, good capacitive conders, etc.), which is simply not included in small single-bank cases, and also, often, more successful connection of cans (2S2P, 2S3P, 3S2P). As they say, it’s easier to lower than raise.

General circuitry PB:

The vast majority of PBs have the following main components:
1) battery
2) DC-DC (Direct Current) converter
– PWM controller / pulse generator (microcircuit)
– accumulator choke
– smoothing capacitors
– diode
– control transistors
– resistors
– indicator LEDs
3) charge / discharge controller
– microcircuit
– transistor assembly
– resistors
– capacitors
– indicator LEDs
4) input / output connectors
– DC port, miniUSB, microUSB
– DC port, USB

How the step-up DC-DC converter works in a very simple language:

All this economy, like most modern devices, operate on the basis of pulse-width modulation (PWM). This means that the current does not flow continuously, but in small intervals (pulses) with a certain frequency. In simpler terms, suppose that current flows in the first 5 microseconds, there is no current in the next 5 microseconds, current flows again in the next 5 microseconds, and so on, everything alternates. On the graph, it looks like this (photo from somewhere on the Internet):
From the PWM controller (pulse generator) pulses are fed with a certain frequency to the base of the control transistor. For simplicity, we assume that the first 5 µs is a signal, the next 5 µs is missing, and so on. Therefore, the transistor closes and passes a current of 5 μs, then closes at 5 μs, then opens again and so on. In the interval when the transistor is closed (a pulse is applied to the base, the collector-emitter junction is opened), the current goes from the power source (battery) through the inductor, the latter storing energy. But, even though the diode is turned on in the forward direction, the voltage is not enough to fully open the transition. As soon as the pulse at the base of the transistor disappears (the next 5 μs), the transistor closes. The energy accumulated in the inductor is summed with the battery, it opens the P-N junction of the diode completely and the current rushes to the capacitor and the load. The capacitor is being charged (it accumulates energy). In the next 5 µs, the transistor opens again (the collector-emitter junction is “closed”) and the current flows through the inductor, while the diode practically does not pass current in the forward direction and prevents the current from flowing back from the capacitor (there it is connected in the opposite direction). The load at this time (5 µs) is fed from the capacitor (it is discharged). In the next 5 μs, the transistor closes again and the current again flows through the diode to the capacitor and load. The cycle repeats. In this case, the increased voltage is obtained from the sum of the voltages from the battery and the inductor, minus the losses, which is what we need (the battery and the inductor are connected in series, the total voltage is summed). By controlling the frequency of the pulses, they achieve the desired output parameters and stable operation of the entire system. With poorly selected components, during long pauses, the load may not be enough for the energy stored in the capacitor (there will be surges / dips in the output voltage), and for short ones, the inductor may not have enough time to accumulate enough energy (the output voltage will be low). Everything must be balanced. That is why small handkerchief converters do not hold parameters
This is how the Step-Up converter works with a storage choke. There are several more species, but this is another topic.

So, enough theory, back to our lamb

Actually, this PB did not suit me: Appearance, of course, is good, but the electronic filling is terrible. How it works and what it can, in comparison with others, as well as detailed TTX, may be in another article. From myself I’ll say that it’s not suitable for serious use because of a bunch of jambs, in addition, the native battery is bad, even very bad

From this PB only a stylish reliable housing is required. The electronic board and battery will be different.
So, a specialized PB board with FastTech SC-0241 1.3A DIY. Brief TTX from the description and (according to the test results):
– charge current. 0.4-0.52A (0.5A)
– input voltage. constant 4.5-5.8V, microUSB connector
– charge termination voltage. 4.2V (4.16V)
– output current. up to 1.3A (1.1A maximum at fully operational parameters)
– output voltage. 5-5,2V, USB connector (4,2-5,21V, depending on the load 0,5-1,2А)
– discharge end voltage. 2.5V (2.4V)

General view of the board:
A little on the device and the principle of operation of this board:
The charge / discharge controller is assembled on a DW01 chip. Found only for DW01-A

I originally planned to put this board in a cylindrical PB, because her performance characteristics from the site were very good. But here I was in for an unpleasant surprise. The fact is that this board is slightly wider and without doping it simply does not enter the PB case: But there is a way out of this situation. As you can see, the layout of the printed circuit board does not go all the way to the edges, which means you can cut the sides of the board. For other PBs, this operation may not be useful. If you decide to install in this PB, then you need to cut it to the very conductors, otherwise the board will not fit, but cut without fanaticism! The board, it seems, is made of fiberglass, it is easy to cut with a sharp knife. Unfortunately, there is no “cropping” photo, but if you cut it as I wrote, then everything fits in nicely, even there is room on the sides for heat shrinkage.
A high-current capacious Panasonic NCR18650PF 2900mah battery with a low discharge threshold up to 2.5V was purchased for this board. I already mentioned a little about him in the article on castration of protected batteries.

Here are the brief specifications:
– form factor. 18650 (18.5mm65.3mm)
– rated capacity. 2900mah (minimum 2750mah)
– internal resistance. 25mOhm
– rated voltage. 3,7V
– maximum discharge current. 10A (short-term, up to 5 seconds. 18A)
– full charge. 4.2V
– full discharge. 2.5V
– recommended charge current. 1.35A
– estimated charge time. 4 hours
– charge method. CV / CC
– weight. 48 g

Schedule of discharge 3A current from the forum ELECTRIC TRANSPORT (I have not found a discharge anywhere up to 2.5V, everywhere up to 3V): I’ll add from myself that it’s nothing special, but it should be better than the praised NCR18650B 3400mah, because the load on the jar at the end of the discharge is large and the last can not cope with it. This battery is somehow, but still holds the load, and the price is not too expensive, so a pretty good choice. And given the low discharge threshold of the board, suitable candidates can be counted on the fingers.

In idle mode with a freshly charged NCR18650PF 2900mah porn elephant, the board produces 5.18-5.21V. Under a load of 4.15 Ohm, the voltage sags strongly, but not critically, some power-hungry devices may not start up (1.1A and 4.66V): The board copes well with average load (it holds the output voltage well):
On a depleted battery and a heavy load, the parameters sag even more (output 1.08A and 4.56-4.6V), at an average load. everything is “bundled”:
With the further discharge of the built-in battery, the output parameters also decrease:
On a sagging battery, the output parameters are not ice at all. Voltage drops to 4.17-4.2V. Many devices simply will not start. As they say, for a high current you have to pay with a low voltage. The converter circuit is too simple, currents above 1A do not “pull”, although up to 1A the parameters are still good:
With a further discharge of the battery, the output voltage also sinks at a small load:
On a completely sagging battery with a heavy load, the output parameters are very poor. Further testing with a large load does not make sense: And on a completely sagging battery, the board keeps the light load very well (0.6A load): Load shedding occurs around 2.4V. A bit underestimated, of course, but that’s okay. As they say, although this panos is high-current, it holds the load at the very end poorly. In simpler words, with a constant load, the voltage on the battery very quickly sinks from 2.7V to 2.4V. Voltage below the specified 2.5V lasts about 15-20 seconds: The best thing, unlike the 3 power supplies tested by me, when the built-in battery is completely discharged, even if you reconnect the load, this power supply will no longer give energy. It is as if “blocked.” To reactivate, he needs a “push” charge. She kind of “unlocks” him. This is a very good protection that will not allow to discharge the battery built into the PB.

Example of charging from a 4-can PB: An example of the end of a charge. At the level of 4.174V, the charge ends, the voltage at the bank is 4.16V: As the saying goes, the same trouble as in Miller ML-102 v7.1. And here it should be ideally:

To summarize the performance characteristics:
ideal output parameters for currents up to 1A
It does not heat up too much at recoil currents up to 1A (it is warm tolerably)
± low discharge threshold (only NCR porn elephants can be discharged without fear)
– price (almost like a whole PB)
– not too good output parameters at a high load, although it is stated up to 1.3A (voltage greatly sags after 1A)
– charging with a low current (500mA) and a long last phase
– slight undercharging (end of charge at 4.16V)
– lack of an indicator in the “current release” mode (only indication of charging / end of charge)

Now the final build of the PB:

What we will use:
– housing cylindrical PB
– Purchased custom board SC-0241 DIY
– Panasonic NCR18650PF 2900mah battery
– heat shrink tube for 18650 batteries

Now relative figures for comparison. This test, in principle, does not mean anything sensible, but it can provide an overall picture. The PB was charged before the green indicator turned on, but as it was written above, it was not completely (up to 4.16V, 90-95% of the battery capacity). I connected to the test after 20 minutes of inactivity. The native cable that came with the smartphone always acted as a charging cable. The load was the smartphone Samsung Galaxy S3 with a capacity of 2100mAh Acc. Charging PB Miller ML102 with a Sanyo UR18650ZY 2600mah battery, discharged to 3V, I will quote later, too lazy to discharge. The smartphone was charged up to 70% to exclude the last phase of the charge and up to 100%. If you do not confuse anything, then the picture is as follows:
– Charging with the native adapter of the SGS3 smartphone from 10% to 70%. 75 minutes
– Charging with the native adapter of the SGS3 smartphone from 10% to 100%. 120 minutes
– Charging the adapter with the smartphone SGS3 of this PB. about 6 hours (charging current is small)
– charging a 100% charged PB of the SGS3 smartphone with a 10% charge level to 100%. 105 minutes, plus the remaining balance in the PB was enough to charge the “next round” from 10% to 21% (in 15 minutes).
– Charging with a 100% charged PB of the SGS3 smartphone with a 10% charge level to 70%. 65 minutes, plus the remaining balance in the PB was enough to charge the “next round” from 10% to 56% (in 52 minutes).

Note: it is important to understand that the efficiency of the converters (PB / smartphone) is not ideal, and the charging cable is not so good, so 2900 vs 2100 does not mean an unconditional victory of the first. Why this happens, mentioned above.

To summarize, I will say the following. I have not seen the ideal board yet. Because I rarely use this device, then I closed my eyes to the low charge current. In fact, I do not use BP every day, so I don’t need to charge it every day. But high return current is important to me. In general, you need to understand that with such dimensions of the board, you can not wait for high parameters from it. Therefore, the return current of 0.9A, I think is just excellent. The collected PB suits me. First of all, I needed a stylish look and high current efficiency, other parameters are not critical. The assembled PB just fits into the conditions.