Review: cheap 18650 mobile power bank

Is it because 1 of 5 is the 6316 and the other 4 are 9635?
What are you saying.

I’m saying that 4 of the 5 I got with the 6316’s are working fine…one of em smoked in a day

All of mine have the 6316’s (I ordered mine from a different company that buyincoins

Could be just crappy China QC or bad chips…who knows

I did put the .36” voltage meter on one of my black ones that was working after I drained it down to where it would quit charging…stopped at 3.3, charging stops at 4.2 [it is also one of the ones I put a RTV blob over the IC…but who knows if that is the reason…might just cause it to smoke too]

Will keep monitoring it

The reason I asked is because my 6316, the one that will discharge all the way to zero stops at about 2.8V when charging an iPhone 3GS. That is the iPhone refusing any more current. So I suppose one could get a false sense of security by testing the discharge characteristics with a device like the iPhone. That would be a “gotcha” just waiting to happen.
To test the discharge characteristics of the 6316 and 9635 I used an 82 ohm resistor spliced into a USB cable. The 9635 stops at about 2.75V, the 6316 goes right on down to 0.

hmmm…I haven’t tested mine like that yet

Not good to drain to 0, yikes

Spent too much time on this already but wanted to reiterate what myself, dchomak & others have found w/our FM6316CE implementations of this thing.

Although fabricated/resistor-based tests like dchomak’s are important & objective, our FM6316CE-based units fail to behave properly in common usage (e.g. my IPad 2 tests). Like dchomak related these devices must not rely on the thing being charged to limit draw or stop drawing when voltage falls.

So of the THREE tasks the device must perform to serve its function; charge battery safely, cease draw when battery gets low (e.g. 2.7v) and charge things attached to it while limiting current to component maximums (e.g. 1A per FM6316CE datasheet) - it may do ONE of them correctly (charge battery).

And based on dchomak’s parasitic drain test, lack of low voltage cut-off further punished the discharged cell - fully discharging it when left unused/idle. We know this irrevocably damages the cell & elevates risk in recharge - something the target users will know nothing about, they’ll just keep charging it.

Yes, better testing ala HKJ would be nice (stop pestering him WarHawk :wink: ), I don’t need this to know not to use my FM6316CE-based units. And who knows which implementation HKJ has & will eventually test.

I’ve placed my order for the “ENB USB Power Bank 2×18650”/battery box, that HKJ did review, for this task. But to each his own.

It is fine that people ask me to test specific equipment, that helps me decide what to buy & test.

Not me, I have one, but have not looked at it yet. It is in the todo pile.

Note: Not everything in my todo pile gets tested, sometimes I remove stuff because it is to old or to difficult to test.

Have you measured how far it drains the cells? I just ordered one of those $5 meters from dx, and would like to get a 2 or 4x pack like this one, if they don’t drain the cells too low (or charge too high).

That’s odd about the micro switch. If I found them for $4, I’d probably buy them for the shells, but at ~$10, I’ll wait, as I have several other USB chargers already, and for most devices, spare batteries.

Cfcubed wrote:

Spent too much time on this already but wanted to reiterate what myself, dchomak & others have found w/our FM6316CE implementations of this thing.

Please don't feel that way about taking our time. You provided very important safety information to us which I, and I'm sure many others, are very grateful.

Thank you also to the OP for for posting a notice at the top of the OP. This is a very good thread.

I'm going to convert my "bad units" to dedicated chargers.

… and it wasn’t too much time as I’ve spent some more:)
Reviewing the FM6316CE datasheet & PCB think I’ve isolated its problems root causes; no Output Current limit resistor from chip pin #2 to ground (UPDATE: useless as it doesn’t t lower draw to 1A rated max, so another chip flaw), and FM6316CE itself does not implement the battery undervoltage protection it says it does (no external components control this - it appears an unimplemented chip function).

So is it possible with some DIY, to improve this power bank, or is there a software problem also?

IF I’m correct, low battery cutoff function is absent/faulty in the chip itself & so is not fixable… Other than (the silly notion of) replacing it with a different/working pin-compatible one.

My/our FM6316CE versions look like reference implementations (less the useless OutputCurrent limit resistor - not respected by chip), so IF I’m right any/all devices using this (version of the) chip lack essential/advertised low battery cutoff / undervoltage protection. Hard to imagine but perhaps representative/indicative of lower-tier chip design/development/production houses.

You can maybe limit the output current more by increasing the resistor at the prog pin. If that schematic is correct: Review: cheap 18650 mobile power bank - But then it will probably still over-discharge the battery in terms of capacity (just not in terms of “too high current”)

When measuring my half-broken one (only still charges correctly, output is only at 4V at all times), I measure 36 Ohm from the Prog Pin to GND. Though there is no 36 Ohm Resistor on the board, so maybe it is burned/fused-together at some point inside the chip, but I can’t really say. Maybe someone with a still functioning board could try.

Overall though: I wouldn’t rely on these for critical applications, and would always expect it to start a fire… (I would do the same with the ETA-based ones though…). So don’t leave it on a pile of paper while leaving the house for more than 5 minutes….

Maybe using them with protected cells (short-cut and over-discharge protected) is a viable solution…

Hmmm,
by the way, did anyone search if these ICs are maybe renamed “fake”-versions of high end ICs from for example Texas Instruments, Linear Technology, … ? (catchword “Power Management ICs” I’d say) So maybe one of those chips could in theory be “dropped in” (though probably you wouldn’t waste one of their ICs, but also want to add those few remaining resistors and capacities for a complete high quality design then)

Thanks for the 411 on all of this.

On the ones in bound to me with a FM6313CE circuit I’m going to toss in a DIY protected 18500 cell pull I just slapped up (only made 1 so far as a test.) I made one for use with a 18650 battery carrier & thus did not use tin strips but just plain ole 24~26 gauge wire. Hopefully that will make it somewhat safe to use. I won’t rely on the protection PCB (dissected from an old AW CELL) to prevent overdischarging each time but hopefully it won’t fail on the times that I do need it to work.

Just to be safer -> I won’t gift these away and keep their limited usage for myself.

AZ

Don’t like the phrase but when something sticks in my craw I’m compelled to follow it to the end. So I tried adding the absent Output Current limit resistor from chip pin #2 to ground, to try to reduce the max current to 1A chip can safely handle. FM6316CE spec says it configures the output current limit from 2.5A - 0.5A using range 10k - 100k to ground. Found adding any such resistor raised the max current, the lower the resistance the higher the allowed draw (e.g. 10k resulted in about 2.2A draw into IPad). Tried 10k, 22k, 47k & 100k - all had bad effect of raising max current draw BUT also resulted in the BLUE LED lighting dimly during discharge - going off when batt dropped to about 2.6V (but not stopping drain/draw).

So I’m finally & fully done with this:) My research & testing suggests at least 2 faults in this FM6316CE chip itself, e.g. deviating from its published spec, neither battery low voltage cut-off OR current draw limiting (to <= its 1A rated max) are implemented properly or at all.

This seems to leave 2 possible uses for these FM6316CE-based units, with the only(?) cells that may fit - unprotected ones, charging batteries OR discharging them to proper storage charge (monitoring w/DMM to 3.6v –3.7v) then removing them. DIY hacker types could have fun adding a $2 Mini DC voltage meter + momentary switch for monitoring and just make sure they don’t charge things that draw too much.

Again an HKJ-reviewed “ENB USB Power Bank 2×18650”/battery box is what I’ll be using.

bluey - WRT your broken one measuring “36 Ohm from the Prog Pin to GND” must be toasted - my “working” ones measure more proper 1.4k.

Thanks for the update CFcubed. Your efforts to test and report on these units are much appreciated. Also your efforts to see if the issues could be fixed. Your tips on how to identify the bad units in Post 53 should be handy. I have come across 2 chargers (one black and one blue) that have the bad chip, but are configured differently.

Here are a few pictures. First on is like your to show how to identify from the top. Has 2 resistor perpendicular to tail wire and different led and other components configuration. Mouse over zooms in to the 2 resistors.

Bottom. Looks like dchomak's black unit.

Before I forget. Welcome to the forum Cfcubed!

So far, BIC is telling me they can not refund for the bad units.

I just used one of the 2 "bad chargers" that I mentioned in the above post to charge a couple phones to see where it would cut off. Halfway through discharging the second phone, it stopped charging. It flashed the blue led when it cut off. My wife quickly disconnected it before I got a chance to check the voltage. I did check the voltage within one minute of it cutting off. The cell was at 3.3x volts.

The charger was definitely running hotter than seems appropriate. I had the lid off. I think it would have been too warm with the lid on. I didn't check temperature or current while it charged. I will try another run sometime this week and report back.

I don't know if good units also get hot discharging. I think I will pot mine with the first recipe from here anyway.

Good additional data esp. the nice photos, so looks like there may be several variants of this thing. IMO most/all(?) of the research/testing WRT FM6316CE-based ones should still hold…. E.g. think any FM6316CE-based ones that cease charge prior to nearly or fully discharging cell (e.g. to below 2.2v) are only doing so because the device being charged is ceasing charge due to low USB voltage.

Good point. I don't know why this one stopped yesterday. It would be nice to know the cell voltage under load just prior to it stopping. I'll try to capture more info next time. Maybe tomorrow.

I dug around a bit, and didn’t find any flat 4 18650 usb chargers less than this, in qty 1, but I did get a flat 2 x 18650 usb charger for a few bucks, and is on the way. I think there were some triangular 3s, and square 4s between $4 and $10 but the shapes weren’t interesting to me.

This may be a decent power bank for the price?