TP4056 current overhead?

[Grum]

Hi all,
I am ordering a bunch of TP4056 and single cell 18650 holders, with a view to building a charging station for up to 5 individual 18650 cells. The TP4056 modules will all be run from a common regulated 5v supply, but I am wondering what kind of current will be drawn from it if all 5 cells are charging at (for argument's sake) 1A. In that case, does each module draw 1A from the supply; is there an overhead to operate the chip and LEDs, etc. (or maybe that is catered for by the 0.8v difference between source and charge voltages); or is there no direct correllation between the charging current and the current drawn from source? I have an 18650 charger module I salvaged from a cheap single cell powerbank, but it is not a TP4056 (no marking on chip, but pinout does not match) and it is only charging at about 290mA, but draws around 400mA. Not a huge difference in the grand scheme of things, and my 5v bench supply can happily provide 13A... just wondered if anyone knows - I can't see it listed on the datasheets.

TIA
Grum

 

[ajw22]

I haven't used TP4056 modules before, but from what I can gather...

There is only one type of "TP4056" IC chip, but there are many PCB/modules from many manufacturers, likely using different components and configurations.

You can set the max charge current of the module by replacing a resistor. It's _probably_ an SMD resistor in a 0603 package size, but may vary between manufacturers. If you're planning to get into electronics, it's quite handy to have an assortment of resistors/capacitors. Search on aliexpress/amazon for "sample book 0603 resistor", as well as "0805" and "1206" sizes.

If the max output is set to 1A, the module should draw 1A + perhaps a few~dozen mA "overhead", depending on what else is mounted on the module. So 5 modules in parallel would likely draw max 5.1A or so.
The relevant line in the TP4056 datasheet is "Input Supply Current", stating at worst 500uA (0.5mA). But that's just what the IC chip consumes, and does not include anything else on the module.

Looks like the chip "burns" off any surplus voltage to keep within the set current limit, so the chip will get quite hot under certain circumstances.
Eg. suppose supplying 5V, set to max 1A, but the battery cell is at only 3.0V. Then the TP4056 has to get rid of 1A * 2.0V = 2.0W of heat... that's a LOT for such a small chip to dissipate.
So you'll probably want to...
1) lower the supply voltage as much as possible (4.5V?), and/or
2) add a heatsink/fan, and/or
3) significantly lower the max charge current.

Also no experience with it, but looks like the TP5100 is a much more efficient module that does not generate as much heat.

 

[Grum]

Many thanks for the insight ajw22. I hadn't thought too much about the excess energy, but was intending mount a small laptop fan alongside the units. I'll maybe have a rethink and put some sort of heatsink across the chips too. Thanks also for the suggestion of the TP5100 - I hadn't even heard of it... I'll take a look at the specs.
Oh, and I have plenty of resistors and so on, though not SMD (my eyesight isn't good enough to habitually work with them) - i've been tinkering with electronics since my early teens, and I've now just retired - LOL

 

[Oleksii]

If you will take a look on a typical schema for TP/TC 4056, you will see that they recommend to power the IC by 0.4 Ohm resistor in series.
Its purpose is to make a voltage drop by dissipating some energy, so the IC then need to dissipate less, than if it would be powered by pure 5v.
This way the IC is not overheating and not throttling charge current.


So, with 1A current, the voltage drop on resistor would be 1*0.4=0.4v, so IC itself would need to drop additional 4.6-4.2=0.4v. So the resistor dissipates ~50% of heat to drop 5.0v to max battery charge voltage 4.2v.

So, it's possible to remove the resistor and power TP4056 from lower voltage., like 4.3v to spend less energy to heat. Also, there no sense provide input voltage higher than 5.0v, as it will just cause circuit overheating.

Also, when a discharged cell just starts charging, the TP4056 is heating more, because voltage drop on it is higher. During charge, as voltage of the cell is rising (so voltage drop on IC is decreasing), heat of the IC is decreasing.

 

[Grum]

Many thanks for your detailed explanation, Oleksii. I’m not sure if you misread my title, or were responding to the previous post in which overheating was discussed. My original question was actually about how much of an overhead (with a D) the TP4056 circuitry caused, like how much current is drawn from the source supply compared to the amount being delivered to the battery.
However, one can never have too much knowledge, no matter how it comes about - so your info is very welcome either way
Grum

 

[Oleksii]

Yead, overheaD != overheaT
As was noted earlier, overheaD by IC itself is very tiny. If LEDs are connected, then one of them will consume way more than IC itself (a few mA vs the 0.5mA).

 

[Korishan]

The TP4056 is a linear buck converter. So basically any energy that doesn't go into the cell, gets wasted as heat. There's also the current limiting factor as well.
So if you plug in a 2.8V cell and the V+ is 5V, then there's a difference of 2.2V. If the current limit is set to 500mA, that's 1.1W of heat being delivered to a very small surface.

The other issue with the TP4056 modules is that there are a lot of clones out there. So you might not have the legitimate version of the IC. This in turn leads to even higher heat issues and worse charging.

The better unit to go with is the TP5100. There aren't many clones, if any, right now. And it also is a high frequency buck converter, so it wastes a lot less energy. It also has a much wider range of voltage input, up to 19V. Just don't connect 2 cells in series to the board, even though they are supposed to be capable of doing it. Off the shelf, they have the wrong sized capacitor and resistor to do properly do 2s. In 1s charging, they are great.

 

[Grum]

Many thanks Korishan. Your comparison between the two options is very interesting. Unfortunately, my TP4056 modules (sourced from China ebay seller, so yes, entirely possible they will be clones or QC rejects) are already in transit, but they didn't cost much, so I may swap them out and use TP5100 units instead. I'll be monitoring the heat and charging efficiency of those TP4056 modules closely ;-)
Thanks also for the heads-up on the changes required for 2S operation on the TP5100.

Out of interest, I have read (about the TP4056) that as well as a single 18650 cell, it is possible to charge a 1S2P 18650 pack through a TP4056, but no more than 2 cells, so a 1S3P pack should not be charged through a TP4056. Does anyone know why this is?

 

[Korishan]

What you could do to narrow down the issues with the 4056 modules is to go through them and match as closely as possible all the top voltages.
Test each one to see what voltage it cuts out at. Then you can match them together.
Another is to use them for top charge voltage only. If cells are below 3.5V, use a different charger, for instance. This will minimize the heat generated and wasted energy, plus increase the life span of the 4056 module as well.

Out of interest, I have read (about the TP4056) that as well as a single 18650 cell, it is possible to charge a 1S2P 18650 pack through a TP4056, but no more than 2 cells, so a 1S3P pack should not be charged through a TP4056. Does anyone know why this is?

You can charge any number of cells in parallel. The only thing to be aware of is that the charge current will be divided, mostly, equally between them. So if the charger, regardless of which one, can output 1A, and you have 5 cells in parallel, then each cell will get about 200mA of current. If you have 10P, then each one will get 100mA, and if they are 20P, then 50mA per cell.
The only issue with this is that you don't want to do this kind of charging on mostly unknown condition cells. Because if one of them happens to be a heater, it'll keep the whole batch from charging as it'll take most or all of the current and convert to heat.
The best way to bulk charge batches of unknown cells is to have something smart enough to monitor the power being consumed and report if the voltages don't rise by a certain amount in a period of time. For instance if the voltage doesn't go up by 0.2V after 10mins, there's probably something wrong.

 

[Grum]

Many thanks once again Korishan, for sharing your wisdom and advice
I’m not likely to ever want to charge more than 1S2P in my little chargers (but it’s nice to know I could), and the 2P comes from pairs of cells that have been connected in parallel their whole lives - I harvest them from laptop batteries and keep them in their pairs as assembled. I can live with the charging taking twice as long (that’s simply logical), as long as I’m not doing any damage. Though I’ll be sure to check that everything stays cool during the charge testing.

 

[Korishan]

Yeah, I can see how those life long partners of 2P cells would be nice to just recharge them together. As they generally "should" have the same specs. There's lots of ppl who leave them connected and test them together. Just a little harder to detect when one is a heater/self-discharger in this case as they wouldn't get as hot during normal charge operation. However it is clear that there would be something wrong if the voltage doesn't go up as expected.

 

[paddy72]

I use TP4056 for at least 2,5 or 3 years to charge my 18650 cells - for each cell one TP4056 - 4 or more in parallel on the same voltage source. I used them with the original resistor that limitates the charge current (its little less than 1 Amp with 5 V input). Yes, the IC can get quite hot when charging and i mounted them on an alu-bar for cooling. I also choose a variable power supply and mostly use 4.5 to 4.8 V as input which reduces the charge current. The current from the input source is nearly the same as the sum of the charge currents - i guess - so the overhead voltage is burnt into heat For efficient usage its best to regulate the input voltage - so the ICs stay rather cool. At 5 V and 3V cells you cannot touch the chips - they get real hot.

And be aware: the fake 4056 you can get from china tend to switch off too late. I had ICs that charged up to 4.35V and even didn't stop but the current was very low. The better ones switch off just before reaching 4.2 V or a few mV more.
And be careful when inserting the cells: the chips dont like reverse polarity! I burnt about 10 of the chips easily - you will see a little smoke cloud dissapearing from the IC and afterwards a little hole in it

P.S. I also use the IC to charge 2 or even 3 cells in parallel - no problem with that - the current is shared by the cells but still limited. But its definetly better to start with lower input voltage if the cells are really low. I leave many cell pairs connected when they tested well. I still could test them on higher charge currents with several diy power supplies (literally buck converters on a DC source - like a laptop PSU) to see if one is a heater. You would also recognize higher self discharge if one cell would be defective.

 

[vicriz]

My pararel TP4056 can draw more than 5a lol

 

[Korishan]

My pararel TP4056 can draw more than 5a lol

Of course, when you parallel the TP boards themselves

 

[ajw22]

My pararel TP4056 can draw more than 5a lol

current of 5.26 A for 3 of the TP4056 in parallel is much too high without any cooling.

EDIT: I didn't look closely at the picture - I assumed you're using 5 modules. If 3 modules are really taking 5.26A, then expect release of magic smoke sooner rather than later.
Also, not sure, but combining the outputs of multiple TP4056 modules may not be the best idea. I would not be surprised if they start doing weird stuff.


IF you WERE using 5 modules, THEN...
5.26A is a bit high, but reasonably close to the expected. There could be several reasons:

1.) I believe the Current/Voltage display you're using has 2 calibration pots in the back to adjust the readings. It's likely off by a bit. Try matching the setting to a _calibrated_ multimeter. There is at least one online report stating it has poor "linearity", meaning that adjusting to accurately measure 5A leads to inaccurate reading at lower currents, and vice versa.

2.) All basic components have some variability to enable cheap mass production. The TP4056 datasheet for example states that the charge current could be anywhere between 950mA ~ 1050mA. That's not even including the inaccuracy of the resistor to set the current, usually up to +/-1%.

3.) The LEDs on the module are regulated with just a resistor (1k Ohm?), so increasing the supply voltage from 5.0V -> 5.8V will increase the current from about 3mA -> 4mA, so with 5 modules a 5mA increase in current... not much, but it adds up.

 

[paddy72]

My pararel TP4056 can draw more than 5a lol

Input voltage of 5.8 V is too high and the current of 5.26 A for 3 of the TP4056 in parallel is much too high without any cooling. I guess they wouldn't last long this way... You should definetly drop input voltage to 5V max., better less (4.5... 4.8V is best imho)

 

[italianuser]

1.) I believe the Current/Voltage display you're using has 2 calibration pots in the back to adjust the readings. It's likely off by a bit. Try matching the setting to a _calibrated_ multimeter. There is at least one online report stating it has poor "linearity", meaning that adjusting to accurately measure 5A leads to inaccurate reading at lower currents, and vice versa.

True, I use those displays and spend some time calibrating them before usage. With a 0.5% accuracy DMM for current, and with my lovely 1980 calibrated Fluke 8500, I'm in love...!

 

[100kwh-hunter]

I tested approx a 10.000 18650's with the tp's, i just connected them to a 6v psu 30A i could load 30 18650 in each round
I put them away for 2 weeks to find sd and then i put them in a bank for discharging.
All i would like to say i have had never one problem, sorry

dis/charger/tester

I am going to build a dis/charger/tester, does this count as a project? for this section of the forum? With the knowledge i accumulated around here, ofcourse. If not, please delete this post, thanks in advance. With best regards, Igor

secondlifestorage.com

I regred that i never posted pictures when the setup was completely finished.

Igor