Short-Circuit Behavior, Take-II
For this second take, I upgraded to a 0.005Ω shunt resistor to eliminate it as a significant current limiter. I also tracked voltages from the 18650 cells’ negative to cell positive, PACK_N and PACK_P to see how far the cell voltage drops, what the absolute current limit actually is, how much quicker the current limit kicks in, and what the current limit is acting on.
Upon contact, the cell voltage drops from 4.2V to 2.5V while PACK_N rises to 2V, telling us that over-current protection is low-side switched. From the ~400mV difference from PACK_P to PACK_N, we can also deduce that about 80A are flowing through the resistor for the first 800µs, going down from there until it cuts off near the 1.4ms mark. This was enough energy to create a weak spot-weld between my resistor and the Q-Cell contacts.
MORE: SilverStone PB06 USB Power Bank Tear-Down
MORE: Best Deals
Hunting For The Q-Cell’s Limit
After charging the packs overnight, I used a simple adjustable shunt to slowly ramp current up from 0A to the cut-off current while writing down terminal voltages along the way. On the unassembled pack, I also recorded raw cell voltages to determine how much resistance the Q-Cell’s wiring and electronics added. The gap between cell and contact voltage increased by approximately 80mV per ampere, translating to an effective circuit resistance of 80mΩ, or about twice as much as a typical 5A e-fuse. That’s enough to rob you of about 3% of the cell’s capacity.
My unassembled Q-Cell cut off a hair above 5A, while my two factory-assembled units cut off at 5.6A and 5.8A.
MORE: SilverStone PB06 USB Power Bank Tear-Down
MORE: Best Deals
Q-Boost Output Waveform
What does the Q-Boost output look like? It operates at a switching frequency of about 1 MHz while putting out 2A at 5.03V (it looks like 4.2V here due to wiring losses) and about 30mVpp worth of ripples. Needless to say, I would have been very disappointed if a $120 power bank fell short of its 2.1A rating like most units from my cheap 12V USB adapter round-up last year.
Although 30mVpp is fine, seeing only two out of six populated capacitor footprints on the Q-Boost’s board tells me that the engineers behind it originally aimed for much lower ripples. Let’s see what happens if I outfit my unassembled Q-Boost with some extra caps.
MORE: SilverStone PB06 USB Power Bank Tear-Down
MORE: Best Deals
Modded Output Waveform
I meant to add some 1µF 0603 capacitors like those already in place, but then realized that I had already plucked two capacitors from a strip of 0805 parts instead of 0603s. How much difference does 2µF of extra output filtering capacitance make? Ripples went down from ~32mVpp to ~25mVpp, which is only about 20% better. To get a more significant improvement, I would need to wack a 10µF capacitor in there. Unfortunately, the only ones I have are standard 1210 size and aren’t going to fit.
Practically speaking, 30mVpp is well within “don’t care” territory for USB power, and there is little point in obsessing over it beyond the nagging feeling that it could have been better with $0.02 worth of parts.
MORE: SilverStone PB06 USB Power Bank Tear-Down
MORE: Best Deals
Q-Boost Output Regulation
While the Q-Boost’s 120mVpp noise under no-load condition may not look pretty due to the regulator operating in pulse-skipping mode for low-load efficiency, there is no point in worrying about it when there is nothing plugged in. This pulse-skipping operation continues until about 400mA, albeit with less frequent skips as load increases. From 500mA all the way to its 2.3A cut-off, where you are more likely to use it, the noise remains at a much more respectable 30-33mVpp while the output voltage holds up quite well from 5.21V open-circuit down to 5.04V right before the cut-off. Once tripped, the over-current protection limits the output to 400mA.
I have two minor complaints here: an always-on design slowly drains the cells and poses a potential short-circuit hazard.
MORE: SilverStone PB06 USB Power Bank Tear-Down
MORE: Best Deals
Mystery Data Line
Using aluminum foil strips to access the pads with the Q-Cell installed in the Q-Boost and a strip of paper on the positive pack terminal to keep the Q-Boost frame powered down until I was done setting up the measurement, I recorded the first second of activity after pulling out the paper strip. What did I find? Not much. The “data” pad rises to 2.25V and nothing happens. Leaving my oscilloscope hooked up with a window trigger and faster time base for a few hours didn’t catch anything either.
Instead of a One-Wire chip at the end of the line in the Q-Boost, the steady oddball voltage may point towards a digital potentiometer.
MORE: SilverStone PB06 USB Power Bank Tear-Down
MORE: Best Deals
Charge-Discharge Cycle
How much energy do you get out of a fully charged Q-Cell after the Q-Boost frame? Discharging a fully charged pack at 1A and 5.14V coming out of the Q-Boost yielded 3.049Ah, translating to 15.7Wh. On the charging side of things, it took 4.637Ah at 4.96V (23Wh) to charge it from dead. Pitting energy in against energy out, you get a 69% overall cycle efficiency, which could easily be improved by a few points if QQC simply beefed up the charging circuit’s inductors a little and smoothed out its switching behavior.
While there were no signs of irregular switching behavior in the Q-Boost, it did get warm enough under heavy load to make me think it'd also benefit from a heavier-duty inductor. It has plenty of height clearance to accommodate one, too.
MORE: SilverStone PB06 USB Power Bank Tear-Down
MORE: Best Deals
Natural Progression
At such a jaw-dropping price, there had to be more to the Q-Swap than just its Q-Boost frame. Picking at QQC’s PR representative revealed that these are meant to be pitched at companies who want to have a fleet of interchangeable batteries for their mobile workforce. That still didn’t sound compelling enough to justify the hefty premium over a pair of stand-alone 5Ah power banks. Further digging confirmed that QQC does plan to offer other Q-Cell-powered accessories at some point in the future.
I guessed a few, but QQC asked me not to tell, so I’ll only share what seemed like the most obvious one to me: a multi-cell charging frame to keep your spare cells organized and topped off without a mess of cables.
MORE: SilverStone PB06 USB Power Bank Tear-Down
MORE: Best Deals
To Swap Or Not To Swap
While physical build quality looks and feels quite good, the DC-DC converters’ efficiency and the Q-Cell’s over-current protection could use some work, as 80A shorts across exposed terminals can mar your shiny devices’ metallic surfaces. I consider that a surprising oversight in such an expensive product.
My greatest point of contention with modular battery systems is the availability of accessories powered by them. It is difficult to justify premium pricing for modularity when the only accessory currently available makes the product functionally comparable to stand-alone power banks costing a fraction as much.
Ignoring the Q-Swap’s price, as it clearly isn’t intended for budget-conscious buyers, what sort of must-have accessory would make you consider buying into such a system?
MORE: SilverStone PB06 USB Power Bank Tear-Down
MORE: Best Deals