Teardown: Tomo V8-4 / Soshine E3 USB Power Bank

38 posts / 0 new
Last post
bella-headlight
Offline
Last seen: 5 months 4 weeks ago
Joined: 01/10/2016 - 16:17
Posts: 1180
Location: United Kingdom

Yes we are talking about extremely small percentages here.
I use an Amprobe DMM which is supposed to be pretty accurate but who knows ?
I will probably get slaughtered for saying this but as your highest cell voltage reading is only 4.23v if it were me I would not worry & be happy charging & using normal 4.20v Li-Ions in the power bank.
At 4.23v, if that is really what the cells are at after charging, I would say that the worst thing that would happen is that you would lose a few life cycles from the cell.
That is only my opinion though & I am not suggesting that that is what you do as only you can decide what you feel is safe.

Ian

Ansarogu
Ansarogu's picture
Offline
Last seen: 5 years 2 weeks ago
Joined: 03/20/2016 - 03:52
Posts: 25
Location: United States

Hey no worries! I understand what you mean, but for all intents and purposes i’m gonna stick with the 4.3v cells, just to be on the safe side.

bella-headlight
Offline
Last seen: 5 months 4 weeks ago
Joined: 01/10/2016 - 16:17
Posts: 1180
Location: United Kingdom
Ansarogu wrote:
Hey no worries! I understand what you mean, but for all intents and purposes i’m gonna stick with the 4.3v cells, just to be on the safe side.

That is absolutely your decision to make, & the right decision for you Smile
I suppose a plus of doing that is that you will get good life cycles out of the 4.30v cells as they will always be slightly under charged Wink

Ian

Ansarogu
Ansarogu's picture
Offline
Last seen: 5 years 2 weeks ago
Joined: 03/20/2016 - 03:52
Posts: 25
Location: United States

That’s right, at 4.2v my Xtar vc2 is showing a mAh capacity of 2,4xx as opposed to the full 2,800mAh when charged at 4.3v.
But that’s fine since my smartphone has a 3,140mAh battery i could charge it at least 2.80~3 times. Or charge 2 18650 cells (2,900mAh) with my Xtar vc2 since it only uses 500mAh per channel. So in my case at least, it works as i need it to work.
Talk to you later mate, i’m of to bed! Big Smile

bella-headlight
Offline
Last seen: 5 months 4 weeks ago
Joined: 01/10/2016 - 16:17
Posts: 1180
Location: United Kingdom

“Talk to you later mate, i’m of to bed! Big Smile”
No worries, I forget the time differences with where members on here live.
It is 12.30 noon here Smile

Ian

jcatena
Offline
Last seen: 5 years 6 months ago
Joined: 06/30/2016 - 16:35
Posts: 1

I got the 2 slots version (V8-2).

The charger voltage limits in my unit are 4.217 and 4.223 V, luckily very well inside the 1% margin. The curves are very good. The charging current is 400mA, and I will modify it to be 700mA (a resistor changed for each channel). So the charging section is pretty good.

When unused there is some drain from the batteries totaling 1.6 mA with peaks of 2.8 mA every second, the average is about 2 mA. This excessive drain can be eliminated inserting a switch to suspend power to the uC and converter. Anyway, it should only be a problem if batteries are left inside for weeks without charge.

The output is common to both ports, and can deliver in theory up to 2.1A total, it can not deliver 2+1A. Both ports are connected in parallel as one 2.1A, only the current sensing is separate for each port.

The uC senses the current of each port and shut downs the converter (always both ports) when any exceeds its firmware set arbitrary rating, as both are actually the same. I did not check the current limits for each port nor if the uC checks the sum to be below 2.1A for the maximum combined output. I verified that the 2A port can deliver 1.5A, good for most or all standard BC devices, and the 1A good only for devices known to draw 1A or less.

The charger port identification resistors are set for the proprietary, non standard, Apple approach, and are the same for both ports as 1A ports (which does tot match the 2A port label). I think the reason to identify both as 1A for Apple devices is to achieve compatibility with standard BC1.2 devices, although it is not the standard way to do it (it should be a short between D+ and D-, but the chosen Apple 1A setting should work with most or all BC devices too). For BC devices, it would mean that they can draw up to 1.5A, and that’s why we see up to about 1.4A with most devices, and why these devices (those that can draw more than 1A) do not work well with the port marked as 1A, as they do not know about the 1A limit. In theory, it could deliver 2A total but only Apple or uncommon devices would draw 2A, and Apples will not because the port identifies as 1A per the Apple spec.

The drop in d4…d7 diodes causes the converter to stop working with battery voltage below about 3.5V, and to have an overall efficiency below 75%. With better shotky diodes it could go down to 3.25V and about 80% efficiency, but only with smart bypass diodes will go down to 3.0V to exploit the full battery capacity and increase efficiency up to 85%. There are not smart bypass diodes similar packaging as the SS32, but i think they could be mounted anyway.
The converter is designed to work with single cell without bypass diode, but here it is used with bypass diodes, that’s the reason for the abnormally high shutdown voltage and efficiency losses.

Overall, as charger the product is very decent. As power bank, not so much, but could be improved easily to be pretty good only changing the diodes, and placing a switch to avoid the excessive drain when unused.

Hope that helps.

benwilliam
Offline
Last seen: 5 years 5 months ago
Joined: 08/11/2016 - 10:46
Posts: 1

I had the same issue with the v8-4.
It powers off already at <3.5V

Exchanging the U5 IC, a LDO which provides the PIC Mcu a constant 2,5V Voltage out of the 18650 Cells, helped. (see schematics here: http://shrani.si/f/G/13l/41eFUNyj/soshine-e3.jpg)

The original one had a far to big dropout of 1V. It already shuts down at <3.5V
The new LDO MCP 1700T-2502E has just a drop out voltage of 178mV. It cost me 60 Cents.

In theory it could down to 2,7V but since there is a shotkey diode in front of the LDO it breaks down at around 3.0V, exactly what I wanted.

But after all I’m very disappointed of this powerbank.

Pages