Testing 21700 Li-Ion battery


For years I have been using LiPo batteries to power my rigs in the field.

They are cheap, it is easy to find R/C LiPo cells on stock, a 3S pack works fine and I have used several of them without troubles. They last for years if you are careful with its charging and discharging process, but still have some issues:

  • You better keep them on a fireproof safe bag just in case of self ignition (rare).
  • Supervision advised during charging.
  • The cells tend to swell with time (years), indicating ageing and progressive degradation.

Searching for alternatives, some people have migrated towards LiFePO4 chemistry, a safer option, although sometimes it is hard to find them on stock. They are also a bit pricey.

Another technology at hand is Li-Ion. This technology was introduced many years ago in consumers, like cameras and electrical appliance, and they are developed widely.

An easy source are 18650 cells, that some people get from salvaged laptop packs, or buy them spare. They are a bit bigger than a standard AA size battery, and have more capacity.

It is easy to build a 3 cell pack with such 18650 cells and power a small rig, like the MTR. Myself, I have one of such packs:

The only concern I found is they are a bit limited when powering bigger rigs, with higher RX & TX currents, and therefore, I can’t run a long activation with them, although it’s still possible if you do a short one. Is it because I bought a cheap brand with low capacity, maybe?

Thinking that I would need a bit more of capacity I found the 21700 cell, a big brother of the 18650. Let’s see if this one can also be used to power my portable rigs for SOTA activation.


Both the 18650 and the 21700 can be found as the hearth of several power packs in the market.

Why this extrange name, you could think? It is not that complex, they indicate its diameter and length:

  • 18650 is 18 mm diameter and 65 mm length. Weight: 39 gram (1,37 oz).
  • 21700 is 21 mm diameter and 70 mm length. Weight: 70 gram (2,47 oz).

If you compare both Li-Ion cells, the 21700 has 23% bigger volume compared to 18650.

Maximum capacity rated for a 18650 is about 3500 mAmp. You could say this is well enough to power any of our small radios in the wild, but the truth is you can never get all that current from them. Why? Because during the supply, the voltage will drop to a value that is degrading the performance of the rig, and therefore you have to stop using them even when there is a residual amount of current that you can’t get.

What about 21700 then?

Due to its increased volumen, they are 33% more capacity, thus they could supply 4800 mAmp (a good brand), but remember you could never get all that current as you should stop early to avoid a deep discharge that could damage the battery.

Preparing the pack

With all that data at hand I was motivated to give them an opportunity.

How could I prepare a pack? First calculate how many cells to join in series.

  • Nominal cell voltage: 3,6 volts

  • Fully charged voltage: 4,2 volts (hey, that’s exactly the same as for a LiPo!)

I decided to prepare a 3S pack, as they would supply 12,6 volts when fully charged. Having a 4S pack would provide 16,8 V which is too much. Yes, you can still add diodes in series to reduce voltage but that is wasting your power. 12,6 V is okay with current small rigs.

How could I connect them in series? There are techniques to solder them, but for convenience, I find more useful putting them in a battery holder.

To be honest, it is easier to find such holders for 18650 than for 21700, but it wasn’t that hard to find an offer for a 3D printer holder for 21700. They come with Keystone contacts, and you just need to solder some wires to put the cell in series.

I bought 3 of them and prepared a pack, adding a power switch to ensure they are off when I walk with them on my backpack, and put a XT60 connector in the outlet:

Then I bought three 21700 Samsung 40T cells. They are rated 4000 mAmp. Some people say there are fake cells that look very similar to the real ones. I don’t know which one I bought. Price is about 9 euros each; reasonable.

They came loaded to 40% that is the custom if you intend to store them for a long time, as the shops do.

I also bought a specific charger for them, that I could easily transport with me and not beinig bulky. It indicates the charged current during process and monitors each cell individually:

When I first charged them they stopped fully loaded with about 2600 mAmp. Remember they were partly loaded as a storage charge.
The pack was ready for the field.

Field test

The pack was fully charged, so I went and activate the summit EA2/NV-119 monitoring the data. To make things harder, it had snowed receintly, so I would face a real hard test on them. They do not recommend to use lithium batteries at low temperatures, but that would be a good test to know how they work in hard condition.

Let’s see some data I annotated during my trial:

  • Temperature: -2 C

  • Start of the activation voltage: 12,3 V (they show 12,6 V on the multimeter, but 12,3 when the rig is just powered on).

  • Activation duration: the radio was on for 1 hour 40 minutes, 65 minutes from which were used running the pile up or working S2S . I never stopped the radio when QSY.

  • RX current: 0,17 Amp.

  • TX current: 1,8 Amps (radio set to 7w out).

  • QSO: 73 contacts on 7, 10 and 14 MHz CW and SSB.

  • End of activation voltage: 11,3 V (1 volt drop).

Cheching the current consumption when back at home

In order to know what was the current used during the activation, I put them in the charger to fully load them again.

  • Pack voltage after one day storage: 11,7 volts. They were too cold when back at home so I better put them to charge after allowing to warm and stabilize.

  • Time to fully charge them after the activation: 2h 15 minutes

  • Current charged: 804 / 906 / 876 per cell (862 mAmps average).

This means I used less than 1 amp in 1h and a half activation time.

By looking at the results, I could use these cells for a longer time on an activation.This is a good result taking into account I used them on a very cold day.

Test is passed satisfactorily; this is going to be my main pack for SOTA activation from now on.

Any beta tester interested on building a pack and checking it? With 210 gram it is a light pack that does worth to bear in mind.

73 de Ignacio


Hello Ignacio, thanks for that report on your battery research. Very informative.

73 de Geoff vk3sq


Thanks Ignacio for the nice and informative report.
73 de Franz on9cbq/dl3rbf

1 Like

Thanks for sharing Ignacio.

I always find reports on batteries interesting.

I usually use a 4200mAh 4S LiFePO4 battery for my KX3, purchased from HobbyKing.

As you say these LiFePO4 batteries are not always in stock, so I have occasionally used a 4S Li-Po battery with a diode reducer.


1 Like

Hi Ignacio,
Interestingly there are posts on the SOTA-Australia groups-io list at the moment asking how to dispose of LifeOP4 batteries which have swelled up and are no longer good to use - so going to LifePO4 doesn’t seem to have solved that problem.

I use “Hardcase LIPO” batteries - so far all OK.

I haven’t tried Li-Ion chemistry as yet.
73 Ed.

1 Like

Thanks for the info Ed, I thought that this issue didn’t happen in LiFePO4.

I will be using my Li-Ion and will see if they are durable and have the feeling they last at least the same as I had with my LiPOs in the past… Hope it works.
73 de Ignacio

Same here - I was surprised to see the reports, but it appears to be the same problem with LiFePO4 and LIPO.

Should not be a problem with Li-Ion

73 Ed.

I can confirm that my first LIFEPO4 4.2AH pack did swell up after about 5 years of usage. I think, but am not sure, that the swelling commenced after I mistakenly set the charger to CHARGE instead of BALANCED CHARGE when I first owned the charger. I had assumed that connecting the balancing lead would trigger the balancing circuit. It doesn’t, well not with that charger!

I did continue to use the pack for several years. Initially the swelling was minor but eventually all cells had expanded so much that the pack was twice the normal thickness. As it would only accept 2000 mah of charge by then, its usefulness was at the turning point and i handed it in to a battery retailer who said they could recycle it safely. The replacement pack, a 4.2 AH LIFEPO4 is still its original size and I take care to use the balanced charge option.

73 Andrew VK1DA/VK2UH


I am using homemade pack from 3 pcs of 18650 elements for about 3 years. I charge each element individually from the bench power supply at 4.2 V and 1.5 A. No any issues so far. It is my power source for HB1B QRP 5w CW only radio, and the fully charged pack can provide power for 3 activations. I expect that 21700 elements will provide higher discharge current and more capacity.

1 Like

Nov-19, I built 2 batteries.
One was packed with BMS and one not, as I don’t need a BMS on the field, my charger will take all necessary precautions to charge and balance each cell individually, once I return home…

One day I was stranded, flat battery, not a problem, as I have a 16W solar panel with 16V output. Straight to the sun, I left it out and forgotten, after 8 hrs under the Oz sky, my battery was cooked, overcharged and boiling hot. I thought that 4S LiFe can handle 16V safe charging. I was wrong.

Maximum Static Voltage of LiFe Battery

The upper battery has traces of whitish substance left on the battery holder, one cell is marked with “F”, a dead cell. It was overcharged and leaked.
The lower pack is working fine. Current capacity is 5.5A not 7 as labelled.
From now on, all my pack will be BMSed before leaving home.

Happy 2021, 73, Pascal


I had one swell when I over discharged it (some idiot left it connected to his FT817!) - LIFEPO4 hardpack, almost as rare as hen’s teeth these days

Hello Pascal,
thanks for your info about using 32700 Li-Ion cells (powerful batteries!)

BMS (Battery management system) is a quite useful add on.

  • They are necessary if you intend to charge the cells while connected in series with an external source of Voltage, moreover if the supply doesn’t regulate the current during the charge. The BMS will be in charge of modulating the charging current individually for every cell iin the pack.

  • When you use the pack in the activation, the BMS will control the discharge current at each cell and will make sure you don’t over discharge the pack below the limit level.

So far I have not added a BMS. I will use a charger that will load every cell individually, so not a risk there.

Concerning using the pack without the BMS, it is not a big problem when you use new cells from the same batch and you look after the output voltage during the activation and makes sure you are not overdischarging (say below 10 V for a 3S).

I know every cell degradates with time, so I must be careful and monitor from time to time if there is any battery that is deviating from the rest. If that was the case I should remove and replace that cell or even to replace all of them by new ones.

Anyway, adding a BMS is a good feature.

That’s it; it all depends on what is your current requirements both for RX and TX.
If you need more amps either you sum 18650 in parallel (3S2P, in example) or move to 21700 that is what I did.

73 de Ignacio

I will describe my li-ion configuration.

I am using this charger.

It works perfectly for me. Power is from 12V source and can charge 2S, 3S and 4S bateries though I never use it for 4S, only 2S and 3S. For charging use computer motherboards contacts for speaker, hdd led etc. Wires are going directly to cell contacts, not via bms.

I will make some photos and post them in new post.

NEVER, NEVER, NEVER use li-ion without BMS

1 Like

You were lucky! My LiFePo4 did swell after 3 years of usage. It still works but…

1 Like

thanks for your input. Using the balance charger as in your picture is a fine way to charge them safely. I also have one of these chargers here.

Concerning BMS, yes, using a quality one is safe, but as I already explained, as long as you do a constant monitor of your voltage during discharge (in radio operation), and as long as you charge them individually and monitors if there is any deviation in a cell, you can go without a BMS.

If there is no way for you to monitor the voltage during your activation, then I agree you better use a BMS to avoid destroying your batts in case of an unaware deep discharge.
In my radios (KX-3 & LNR LD-5 I see the voltage on the display all time).

I suggest to watch this as it is quite well explained:

73 de Ignacio


This is my 3S3P battery. Box is little bit big but I have it in house.

Charger is connected via 9pin female connector. Cell contacts are directly connected so female contacts are must. Balance charger is connected via that port.

Cells are connected to bms. Banana sockets are output and connected to P+ and P- on BMS via fuse. You can charge cells also using banana sockets. I am using old current source 250 mA for charging. Also tried photovolt panel via shotky diode. Both ways work. BMS stops charging battery when firs cell reaches 4,2V. That makes little non balance but it works. Just then and now use ballance charger.

I am using this BMS because he stops discharging at 2,7 V while most of others stops at 2,4V. And, you can use it for 3S, 4S and 5S batteries. I don t have link but little browsing on e bay looking for “red” BMS and you will find them.