Thanks to the anonymous donor, here’s another power bank subjected to the “Gough” treatment. This one, like many others, comes in a very “generic” colour cardboard box packaging with a clear plastic window.
The power bank and its included pouch are visible through the window, along with the “moko” branding. Many power banks on the market seem to have the phrase “wonderful life because of you”, including other non-sensical phrases. In this case, “shortcut high speed” is yet another confusing statement.
The rear of the box seems only to list a single output of 5.3v (higher than USB limits of 5.25v) at 1000mA. The unit itself has the real specifications, and it seems the packaging may have been recycled from another similar unit.
Included in the package is a nice pouch, a basic microUSB charging only cable, and the power bank unit itself. No manuals or leaflets were included. The front of the power bank has a clear window, where four blue LEDs are used to indicate the charge level of the power bank.
The top of the power bank has two outputs, which are unlabelled, as well as the microUSB B charging input and a single 5mm LED as an “emergency torch”. It’s pretty feeble, and using the power switch, can be toggled (via long-press) between on, SOS and off.
The power button is situated to the side, and has to be short pressed to activate the power bank’s operation (i.e. not automatic).
The model number of the unit is MK555, and it offers two outputs, up to 1A and 2.1A. It seems to imply a high charge current input of 2A which would mean faster charging, although I did not verify this. The moko brand appears to belong to Shenzhen Moko Technology, a producer of PCBs.
The power bank itself was fairly well closed, and opening it was a destructive process of prying at the seams which are on the rear. The rear is a plastic cover which attaches by clips, most of which were snapped in the process of opening.
With the cover off, it is already apparent that this sample of the power bank appears to be built with three genuine Samsung ICR18650-26F 2600mAh cells. Astute readers will see that this is only a total of 7800mAh, and not the 8000mAh claimed on the outside, however, this lie is “relatively small” compared to what other vendors have done.
A closer look at the top of the PCB seems to show it is coded with SSJN-MP-104. The circuitry reveals an SS54A 5A Schottky freewheeling diode, an open-inductor (rather than a more efficient enclosed type), the power switch and two 8205A MOSFETs on this side of the board.
A little concerning is the quality of the soldering of the battery wires – there is significant loss of insulation towards the ends of the wires that were soldered, and they are soldered at a precarious angle. There is a probability of short circuits should the battery negative wire make contact with more pins on the MOSFETs than expected.
As is common for many inductors used on switching power converters, this one is bifilar wound to try and improve the performance. The LED is also seen mounted “bent” into the air with no support, a common method of facing LEDs into cutouts in cases.
The underside is marked week 34 of 2013, and shows a microcontroller with ground off markings (U1) that runs the whole unit, and an assortment of resistors, transistors and MOSFETs. Another 8205A MOSFET is visible on this side. As many of the parts are Chinese locally sourced parts with cryptic markings, identification of U2 and U3 cannot be positively made, but likely they may have battery charging roles, or that may be the responsibility of the microcontroller itself. Four surface mount blue diodes are also seen and indicate the charge state.
Concerningly, there is a discolouration around the lettering of U3, Q7 and the USB port itself, which may have happened during manufacturing (i.e. discoloured flux remnants), or could be a result of prolonged operation in testing causing overheating of the PCB itself.
Notably, there are no electrolytic capacitors for smoothing, which could result in poor quality power output.
Testing of this power bank was performed on the new test rig with a similar methodology to the other power bank tests done to date. Of note was the tendency of this power bank to emit audible noise which varies as the batteries discharge. This noise is similar to a squeal, which is load and battery voltage dependent, and is often the byproduct of a poorly designed switching converter with low switching frequencies. It can even happen to properly designed switching converters when operating outside their design specifications – e.g. overloaded. Where the switching frequencies are low, a process known as magnetostriction causes the inductor windings to “vibrate” at the same frequency as the switched current thus emitting audible noise.
The capacity results are as follows:
|Load (mA)||Run||Capacity (mAh)|
|Load (mA)||Run||Capacity (mAh)|
|Load (mA)||Run||Capacity (mAh)|
From the capacity results, it is very likely that the batteries are genuine and are of the 2600mAh per cell as stated. The capacity is 7182mAh at 500mA, 6904mAh at 1A and 6297mAh at 2A which is an efficiency of 92%, 86% and 81% respectively. These figures are fairly good at low loads, however the reduction in efficiency as the load increases suggests the switching converter hasn’t been optimally designed for high load operation. This is amazing, especially as the power bank leaves the LEDs on during discharge.
The voltage versus sample (time) graphs seems to show relatively stable regulation at low loads, however, at 2A, the voltage is below the 4.75v USB minimum requirement and may mean that devices expecting high currents could suffer slower charge as the devices back-off to prevent overloading the charger.
There is also an anomaly with the regulation at all loads which manifests itself as a slow small “ramp up” of the voltage before returning back to a stable regulation regime. This is consistent from run to run and is probably a specific consequence of the program being executed in the microcontroller. This is, however, of no specific consequence to the end user.
At 500mA load, the ripple is high at 734.2mV (compared to a charger normal output of 150mV at most). The USB worst-case scenario is 4.75v to 5.25v, a span of just 500mV. The switching frequency initially is very low at 26.08khz, which is just above audible for a 500mA load, but may be audible by pets and animals (and could repel or cause them distress). The switching frequency varies throughout discharge.
At a load of 1A, the ripple now measures 1066mV, which is yet even higher, with a slightly lower switching frequency of 21.45khz at the beginning of discharge. This is on the borderline of audibility by the “youngest” ears, but eventually falls to something which can be slightly annoying.
At a load of 2A, the ripple is now 1291mV, which is almost 26%!!! The waveform has a lot of strange components, but a periodic component seems to exist at 4.28khz and a frequency about four-times this (17.12khz). These are audible as a fairly annoying whistle/whine.
While this supplied unit featured quality genuine Samsung cells, with a very modest “round-up” of 200mAh to 8000mAh, and a very commendable conversion efficiency, I cannot recommend this unit because of the low quality of the output power which is out of specification, problems with build quality and audible noise which can be annoying to end users. There is a real potential for harm or stress to devices connected to such poor quality sources of power.