Power bank cases 18650/26650

[TheGrave]

Hi,

I'm looking for an alternative of the XTAR PB2S as it died after a few months of mild usage on me. Input and output currents on USB-C port were around 1.7A with an Anker cable which is OK but...bad bad design - magnets of the plastic cap are too weak to hold it and 18650s wobble inside and lose connection sometimes. Had to do stupid tricks and fill the internals with paper to keep them in place. Real piece of garbage.

Requirements:

1) No MicroUSB charging port, I'm sick of buying these faulty by design cables. Spent over $50 in the last year probably on MicroUSB cables (all brand names).
2) 2 output ports (could be USB-A or USB-C, doesn't matter)
3) QC3/QC4/USB PD of some sort, at least 18W.
4) Holds either 2, 3 or 4 cells.
5) Over/under voltate/charge and temperature protection.

Is there such a product on the market? I have a Soshine E3S but now one of its output ports is dead. They have an E3S-QC - nowhere to be found for sale.

Then there is the Tomo S4 which looks nice but no QC. Nitecore F4 also looks decent but no QC and a MicroUSB port, blah.

Alternative option - same specs but for 26650s, single- or dual-cell product. I found only cheap Chinese noname garbage power banks for these cells. Can't trust those.

 

[Lynx_Arc]

Years back I looked a lot for power banks that allowed cell swapping that supported QC and there were none at all available back then. I concluded that I would need prebuilt power banks or cob together a setup myself and proceeded to buy some 12/24 QC3 circuit modules and have a setup to charge my phone off a Black and Decker 20V battery pack.
It is not as elegant or compact a solution but if I didn't need to transport it around it is actually a better solution for me as
I have a half dozen 20V batteries which equates to 30+ 18650s and the higher voltage of the batteries is more efficient.

One problem with removable cells power banks is the contacts are a weak point and to get 18W out you have to boost the voltage about 3 times as high as the batteries meaning instead of 1.5A at 9v you need about 5A without losses and losses can be 30-50% leading to perhaps closer to 7-8A and using springs and contacts you could see even more losses involved. This can be a big problem as trying to source 4A each from a pair of batteries with possible contact resistance that could cause one battery to be forced to provide most of that power if the other had considerably more resistance you could see what problems could arise. You almost need to have welded/soldered etc batteries in higher output power banks to eliminate problems or designers need to have custom made parts for them to suffice. Because when you weld batteries in a power bank permanently you can closely match cells and even put them in series to instead of having to boost you can buck on lower voltages and boost less hard on the higher ones.

I've not seen or heard of 26650 power banks..... I seriously doubt you will see any substantial offering as the battery size is uncommon out there and the power density isn't optimal either likely 21700s would be considered before them.

 

[TheGrave]

I'm with you my friend, I was thinking about the very same issues today that might be the root cause of the lacking QC cases I'm looking for Seems like I'll need to migrate to 21700 Nitecore's solution if the batteries ain't soldered there.

 

[Lynx_Arc]

I'm with you my friend, I was thinking about the very same issues today that might be the root cause of the lacking QC cases I'm looking for Seems like I'll need to migrate to 21700 Nitecore's solution if the batteries ain't soldered there.

One new thing that is on the horizon is thunderbolt which is essentially USC C on steroids I believe.

 

[TheGrave]

One new thing that is on the horizon is thunderbolt which is essentially USC C on steroids I believe.

What does the interface have to do with the problem here? USB-C PD delivers plenty of power already.

By the way take a look at the internals of the XTAR PB2S - no springs used there. Didn't have a proper QC device to test the power delivery but 1.8A were coming on 5V for sure. I'm confident this design allows for 18W or even more.