Comparison of power Bank for IoT ESP8266, ESP32, Arduino, GPS, Bluetooth
Table of Contents
Power banks are a convenient means of choice for supplying small projects such as IoT, streaming or sensors. But what's with low quiescent current, efficiency and maximum performance? Ideally the cells should be exchangeable. Pre-equipped (cheap) power banks from China are mostly of questionable quality and rarely afford what the seller promises. Below is a test of various models which will be offered without cells.
For comparison, two ready-made models were added to the test. A "RealPower PB-260" (1 x 18650) and, on the other hand, an "ECTechnology B30224" (8 x 18650).
Overview
Power Bank comparison
Model | 1 x silver | 1 x black | 3 x Gold | 4 x black | 5 x black / red | 8 x Green |
---|---|---|---|---|---|---|
IC | MP3401A | TP4333E | HT4936S | HT4936S | - | - |
Cells | 1 | 1 | 3 | 4 | 5 | 8 |
Always on | x | - | - | - | - | - |
Vout idle | 5,05V | 3,31V | 3,03V | 3,03V | 2,88V | 0, 31V |
IQ at 3, 6V | 94uA | 44, 7uA | 14, 4uA | 14, 26uA | 35, 5uA | 32, 92uA |
Ioutmax at 3.6 v (minimum Vout 4,7V) | 1.1A | 1.1A | 1,2A | 1,1A | VER 1.9A | 2,3A |
USB out | 1 | 1 | 1 | 1 | 3 | 2 |
USB in | 1 | 1 | 1 | 1 | 1 | 2 |
Download power base data (Libre Office. ods)
AGML battery charger
Testobjekte
1 x Silver (MP3401A)
1 x Black (TP4333E)
3 x Gold (HT4936S)
4 x Black (HT4936S)
5 x Black / Red (unknown)
8 x Green (unknown)
1x RealPower PB-260
8x ECTechnology B30224
Test
Devices
Supply: Korad KA3005D
Current measurement (input side): Voltcraft VC820
Voltage measurement (input side): UNI-T UT71C
Performance measurement (output side): 150W programmable load with Kelvin connection
Implementation
The input current of the converter is measured for different output currents. The input-side measurement is performed in such a way that wire/contact resistance and the associated voltage dip are manually compensated. (Voltage-correct measurement) The load is connected via a USB cable of 14cm length in 4-wire technology (Kelvin connection).
Efficiency for different load currents
1 x Silver (MP3401A)
1 x Black (TP4333E)
3 x Gold (HT4936S)
4 x Black (HT4936S)
5 x Black / Red (unknown)
8 x Green (unknown)
1x RealPower PB-260 (unknown)
8x ECTechnology B30224 (unknown)
Comparison of efficiency for fixed load currents
Load current 0.1A
Load current 0.5A
Load current 0.75A
Load current 1.00A
AGML battery charger
Comparison of the efficiency for different load currents at 3,6V
Quiescent current requirements - comparison of candidates
Evaluation
Quiescent current requirements
In the tabular comparison of the demand for retirement, it becomes clear that the "RealPower PB-260" model has by far the highest quiescent current demand, followed up by "1x Silver (MP3401A)", which can score with "Always on" (Durable 5.00V at the output). Compared to "1x Black (TP4333E)" it is shown that this function requires 40-45% of the quiescent current in total. The "3x Gold" and "4x Black" models, based on a HT4936S IC, show similar values as the ready-made product of "ECTechnology". In the midfield, there are "5x Black / Red" and "8x Green".
Efficiency
It is clear from the table that the (ready-made) product of "ECTechnology" (B30224) is characterized by a high efficiency over the entire load range. Similar values are provided by "8x Green", the difference is 3-4% for currents > 0.1A. The standard deviation for load currents of 0.1A is ~ 6.7% and at an average of 3.63% for load currents between 0.5-1.0A. It can be seen from the diagrams as well as from the table that the maximum efficiency is reached at 0.5A.
Not shown is the maximum possible output current with the lowest possible/allowable (by spec) USB voltage of 4.70V. The advertised 2.1A maximum load current (for modells with 3 or more cells) are only available with the models "8x Green" and "B30224" from "ECTechnology". The latter does even more than 3.4A. For all others, the maximum load current is averaged at 1.1A over the entire cell voltage utilization range.
AGML battery charger