Li-ion / lipo battery pack building and soldering 123

[-LightOn-]

Hi,
I have been reading this great forum for a while, but now I just had to jump in to ask few things I couldn't find answers for :wave:

I have a multimode bike/ski/run light project which is based around one P7 dbin led and a r2 drop-in. I am building the driver around a LM3401 and d2flex, but this is just useless background information.

My real concern is the battery packs. I was planning on building one light weight 2s pack and one larger 2s3p pack for those long rides. I have ordered some trustfire 2500mah 18650 cells for the packs with protection circuits.

I haven't seen too many instructions or experiences on soldering li-ion cells, but this is how I planned to assemble the packs:
First glue the cells together with epoxy. Then balance the charge on individual cells by placing a 10Ohm resistor between the terminals. After this pre-solder the battery terminals and connection tabs with a powerful soldering iron at 380C for about 2-3 seconds using flux. Do the same for the main and balancing leads. Solder the wires, tabs and terminals together for 2-3 seconds. Then top-off the cell by shrink wrapping it and giving it a nice neoprene or cordura casing.

Is this just doomed to fail?

Additionally, I bought a bantam bc-5 balancing charger and the manual says I shouldn't charge protected cells. I have understood that you can charge protected cells just fine?

I would be grateful for any advice, tips, experiences before I go and blow my head off while burning down my house

 

[steve6690]

I would recommend that you buy 18650's with tabs. They didn't need pre-balancing because they are shipped in a "storage" state of charge. They are unprotected cells so you will need a pcb to control overcharge/discharge and cell balancing. The charger manual is, I believe, telling that you shouldn't balance charge protected cells in a pack.

 

[Fallingwater]

I haven't seen too many instructions or experiences on soldering li-ion cells, but this is how I planned to assemble the packs:
First glue the cells together with epoxy. Then balance the charge on individual cells by placing a 10Ohm resistor between the terminals. After this pre-solder the battery terminals and connection tabs with a powerful soldering iron at 380C for about 2-3 seconds using flux. Do the same for the main and balancing leads. Solder the wires, tabs and terminals together for 2-3 seconds. Then top-off the cell by shrink wrapping it and giving it a nice neoprene or cordura casing.

You might find my guide helpful.

Do note the disclaimer and safety reminder though.

 

[-LightOn-]

You might find my guide helpful.

Do note the disclaimer and safety reminder though.

Thanks for the instructions. Exactly the confirmation I was looking for. It seems that I was on the right track anyway. Have you tried soldering packs consisting of protected li-ions? Any problems?

 

[SilverFox]

Hello -LightOn-,

Welcome to CPF.

The main problem is that the high transient temperature during soldering may destroy the protection circuit. The protection circuit is a small chip that sits on the - end of the cell. You would be soldering directly to it.

You may be able to remove the protection circuits from the cells and approach the project as if you were using bare cells. Then you could simply add a global protection circuit for the whole pack.

Keep in mind that soldering Li-Ion cells without tabs can be very dangerous.

Tom

 

[Fallingwater]

Thanks for the instructions. Exactly the confirmation I was looking for. It seems that I was on the right track anyway. Have you tried soldering packs consisting of protected li-ions? Any problems?

I have as of yet not soldered directly to protected cells, so I can't tell you how the circuit takes this treatment.
I do have two bare protection boards from when I needed unprotected 18650s and only had protected ones; one I'll keep because I need it, but as soon as I get home I'll experiment on the other board by soldering a wire to the terminal. I'll then report on my findings.

 

[-LightOn-]

How would a fried protection circuit be verified? Is it an open or short (hope not) circuit? At which end (terminal) is the circuit usually located and is it easy to remove?

 

[SilverFox]

Hello -LightOn-,

I have had protection circuits fail both ways, but it is more common for them to fail open. You can get a voltage reading, but as soon as you apply a load it drops to 0.

The protection circuit is located at the - end of the cell. If you search you should be able to find a thread showing this. I seem to remember that member LuxLuthor was showing how the protection circuit was attached to the cell.

Removal is easy. Remove the outer layer of shrink wrap. This will expose the circuit and the ribbon wire running up the side of the cell to the + terminal. Now you have to disconnect the spot weld connections from the circuit board to the cell being careful not to short things out. Now you simply need to re-shrink wrap the cell and you end up with a bare cell.

Tom

 

[Fallingwater]

I want to stress the danger of shorting the metal strip connecting the positive terminal to the board. When I removed the protection from my 18650s I found the metal strip to be insulated from the body with a strip of paper-like stuff only marginally wider. In my experience the best way avoid shorts is not to remove the whole plastic sleeve around the cell and then remove the circuit; rather, to cut only the lower part of the sleeve, cut/rip off the negative tab, then grab the circuit firmly and tug. This way the strip itself will tear open the sleeve, and since you're pulling away from the cell the risk of it contacting the negative body is minimized.

 

[-LightOn-]

Hmm... this is getting more complicated than I thought. I had planned of just using protected cells in the pack to be safe, but this would mean I have parallel protection circuits. Now if one fails and the others don't I could be in for trouble. Would you rather just use individual protected cells or remove the protection circuits, solder the cells and then assemble 2 separate protection circuits to the balancing leads.

 

[Fallingwater]

I'm sorry, I cannot give you a decisive answer before I try soldering to a circuit, and this will have to wait a few days.

One thing that will work is solder the unprotected cells in parallel, then connect the parallel pack to a single protection circuit extracted from a previously protected cell, which can then be connected in series with circuits from other packs done in the same way. Once the circuit is removed you can trace the connections from the terminals, so you can solder to the board directly and avoid damaging components.

Note that this only works if the current draw is low enough that a single circuit won't give you trouble - if, in other words, your light would work from a single cell, and you're using a parallel pack just for runtime, not for getting more current out of them.

Also note that the cells need to have the same resting voltage before you put them in parallel, or damage can occur.

 

[-LightOn-]

Note that this only works if the current draw is low enough that a single circuit won't give you trouble - if, in other words, your light would work from a single cell, and you're using a parallel pack just for runtime, not for getting more current out of them.

Also note that the cells need to have the same resting voltage before you put them in parallel, or damage can occur.

Yup. Thanks for the advice. I'm not sure if the 2.8 Amps rated for the P7 is exactly low, but in the lower capacity pack the p circuits will have to handle 2.8A anyway so single protection circuit should do in the parallel pack also.

I was planning on evening the resting voltages with a resistor.

I'm sure this one has been answered earlier elsewhere but can the over discharge protection be left to the protection circuits? The driver circuit itself has no low input voltage protection or cut-off.

 

[LuxLuthor]

Here is my thread showing PTC's.

I don't think it would be reliable (as in damaging the PTC) to solder to the bottom of PTC. Note in these pics that the cell connections are made to the cell -/+ ends with welded nickel contact strips, which in turn are solder to contacts on PCB. The contact strips and PCB are insulated from cell. It won't take very much transferred heat from soldering PTC ends to melt thin insulator from + contact strip running down outside, and dead short against battery can.

For this reason, and the deleterious effect it MAY have on the actual PCB/components/solder points, I could never recommend soldering to a protected Lithium cell. I have made Lithium packs with individual AW protected cells using my battery pack welder, using thinner nickel strips and on reduced weld energy settings. There have been others who at their own peril, have soldered to unprotected Lithium Cobalt cells, but personally, I would never do that.

 

[Fallingwater]

I was planning on evening the resting voltages with a resistor.

No need.
Charge the pack fully, then set the charger to single cell (or use a single-cell charger) and hook it with small magnets on each cell on the pack, starting a new charge every time.
This will insure all cells are charged to full capacity, and shouldn't take long because the cells will be close to that point already.

I'm sure this one has been answered earlier elsewhere but can the over discharge protection be left to the protection circuits? The driver circuit itself has no low input voltage protection or cut-off.

I assumed that was why you wanted the protection circuits in the first place.

 

[-LightOn-]

Ok, I got most of the kit already so I can start this project.

I got the bantam bc-5 and converted an old atx-power to give 12V and enough amps for it. (Conversion was a fun little project itself )

First thing was to measure the resting voltages of all the cells:

1 3.985
2 3.797
3 3.796
4 3.981
5 3.796
6 3.983
7 3.802
8 3.976

So all the cells were over the nominal 3.7V. First I took the first cell and charged it alone. End result 4.20V and 610mAh into the cell. Then I took the cells 2 and 3 since they had equal resting voltages and charged them together. This time the result was 4.20V and 4.22V measured with a DMM. And the charge put to the two cells was 3100mAh which is 1550mAh per cell. There is a huge difference in the amount of charge put to the cells between 1, 2 and 3 even though the difference in resting voltage was only 200mV. Do the amounts put to the cells sound correct? Additionally I am concerned with the 4.22V of the third cell. The manual said that the li-ion will explode if the voltage of the cell reaches 4.25V which isn't too far away from the last figure.

 

[-LightOn-]

Update: I found that by doing a constant current charge and constant voltage charge for a pair to the point where the charge current has dropped to 1A work great. After this the cells are around 4.1V. Then top off the cells individually to 4.2V.

It seems that the first cell (from currents perspective) gets more charge because the minute resistance of the charge leads between the cells is enough to cause a difference in loading currents. (Kirchoff's laws)

 

[Packhorse]

If you get protected cells and remove the PCB you can then solder onto the Nickel strip that was used for the PCB. Then use a single PCB externally.

 

[waddup]

i connected 3 x rcr123s together with 'painters tape' (just rolled em up together like a cigarette) yesterday, to make 1 long 11.1v cell, charged it on my 'intelligent' trition charger, no problems, its works well.

why solder?

 

[-LightOn-]

why solder?

Well... I guess the electrical connection is a LITTLE bit more reliable than just bare tape.

 

[Fallingwater]

Tape will give you a very unreliable connection.