Packs pulled from solar post light - are they LiFePO4 or Lithium-Ion?

[chaosdsm]

I pulled 4 battery packs (2x 18650 in parallel) out of my new "commercial grade" solar post light because the light output is about 1/2 what it claims to be - so I opened it up to check for bad batteries or under-charged batteries. Battery packs are labeled as IFR-18650 3000mAh 3.2V 9.6Wh.... but I tested all 4 battery packs twice - 1 through 4 then 1 through 4 again - with my Klein mm720 multi-meter at:
1> 3.743V / 3.743V
2> 3.740V / 3.741V
3> 3.742V / 3.742V
4> 3.737V / 3.737V

So I'm thinking, they have wrong protection circuit on them, bad protection circuit, or they are regular Lithium-Ion batteries... I went back after an hour and they're still showing the exact same voltage readings, which tells me that these are very likely to be regular Lithium-Ion & NOT LiFePO4 batteries at all since all my cylindrical LiFePO4 cells drop to 3.3V in less than 1 hour after charging whether new or used. My LiFePO4 prismatic cell is a bit different as it takes hours to drop back down to 3.3V.

Putting regular lithium Ion batteries into the erratic charging cycle of a solar charging system seems like a REALLY bad idea.... Unfortunately, I cannot just open up the battery packs without voiding the 5-year warranty to see if the base batteries are labeled differently... and the company wants to charge $35 + shipping for one battery pack so not really a good option to buy a separate pack just to cut open & see if the base batteries are labeled by the original manufacturer

But once the included batteries give out (assuming they don't catch fire & burn up the light), I plan on making my own battery packs with 3000mAh 21700 LiFePO4 cells that I recently found (if they're any good...) & trying them in 2S instead of 2P for improved output.

 

[vicv]

Yeah, the voltages is sure telling me that they're normal lithium ion. I also agree that it's a bad choice to use that chemistry in solar lights but my buddy gave me a couple 14500 that came from those. Labeled as 1400 mah, but I tested them at under 300 and told him to toss them.

What do you mean by protection circuit, though? Adding a protection circuit would not increase the voltage.

 

[DRW]

... I cannot just open up the battery packs without voiding the 5-year warranty to see if the base batteries are labeled differently... and the company wants to charge $35 + shipping for one battery pack ...

It's unlikely you will need the warranty, go for it. Sounds like you plan to service them yourself anyway.

 

[scalpel_ninja]

…& trying them in 2S instead of 2P for improved output.

Would the system be able to handle the increased voltage? 2P with Li-ion would be 3.7V, but 2S with LiFePO4 would be over 6V?

 

[chaosdsm]

It's unlikely you will need the warranty, go for it. Sounds like you plan to service them yourself anyway.

I usually like to keep the warranty intact for at least 60 days on new products I'm not familiar with, especially when they're 3x the price of other similar products... Most problems surface within the first 60 days. Only the light & solar panels are covered for 5 years.

Would the system be able to handle the increased voltage? 2P with Li-ion would be 3.7V, but 2S with LiFePO4 would be over 6V?

Not sure, no data on what LED's are in the light, so all I can do is try & see when the time comes. I don't see anything on the PCB that indicates what voltage can be used, possibly on the underside, but I won't mess with that till it's time to swap batteries.

What do you mean by protection circuit, though? Adding a protection circuit would not increase the voltage.

In solar charging, if there's not an over charge protection built in (or if protection circuit is not working correctly), the solar cells will continue to feed, whatever voltage & amperage the sunlight generates, to the batteries even after they've reached full charge. Potentially allowing them to go beyond their maximum voltage specification. Manufacturers specs indicate 3-5 hours of good sunlight to reach full charge. These cells had about 2.5 ideal days of charging before I pulled them out & tested.

 

[alpg88]

they are fully charged and yet at nominal voltage, it tells me there is a charge protection circuit there, it seems to be overprotective, lol

Looks as if the circuit was made for 3v cells, but the manufacturer put 3,7v cells instead

 

[chaosdsm]

I had the brilliant idea last night about an hour after sunset to see if the High/Low switch had been reverse wired.... The head unit comes off with just a slight twist, so it's a really simple check.

High is supposed to supply up to 36 hours of light if it's to heavily cloudy / rainy to charge in the day. Low is supposed to supply up to 72 hours of light, so, about half the current draw... which would make for a very noticeable difference in output... Zero visible difference in light output between low & high.

Talking with the manufacturer about getting a replacement head unit sent out, & I still have a little over 2 weeks to send it back through Amazon if needed.

 

[chaosdsm]

I ordered a replacement lamp through Amazon, & tested all 8 battery packs out of both lamps once the replacement arrived.... The 4 packs pulled from the original lamp about 1.5 hours before sunset all read between 3.95V & 4.0V, 1 hour later, they all read the exact same voltages. The 4 packs I pulled out of the new lamp all read about 3.65V to 3.7V & no telling how long those have been stored in the box....

On the positive side, the new lamp does cast light out to about a 16ft radius, where the original only cast out to about a 13ft radius.

 

[lumen aeternum]

I'd like to find some quality solar walkway lamps, but advertisements never tell you the details. Having a replaceable battery pack sounds like better quality than the usual hair-thin wires with crummy solder & gobs of flux.

 

[chaosdsm]

My biggest problem is al the BS lumen ratings.... this post light was supposed to be 300 lumens....


The bright spot on the right is my Convoy L6 (original XHP-70, not the newer 70.2) on medium setting which is about 180 - 200 lumens.

 

[Hooked on Fenix]

If the batteries are single or in wired in parallel, I don't think you'll have the same safety concerns with regular lithium ion batteries as if they are wired in series.

 

[chaosdsm]

If the batteries are single or in wired in parallel, I don't think you'll have the same safety concerns with regular lithium ion batteries as if they are wired in series.

2P packs, but lithium Ion is not a solar friendly chemistry, and probably why they say that the battery packs need to be replaced every 2.5 - 3 years... at $35 per 3000mAh pack x 4. Not going to pay that, I'll build my own battery packs instead.... batteryspace has 1800mAh 18650 LiFePO4 cells with tabs for under $5 each which would give me 3600mAh packs.

 

[chaosdsm]

The new light does seem to be brighter now. I was able to see more definition in the bushes nearby last night than I could the first night I got the replacement light. I also applied "Frosted Glass" vinyl to the glass, looks so much better now that you cannot see individual LED's!

I was also able to get one of the GamaSonic brand replacement battery packs for just $24 shipped.

> It shipped with an adapter which I was able to hook up to my T180 charger. I charged it on Lithium Ion setting & the pack charged to 3.99V before high voltage cutoff kicked in. Within 30 minutes it was back down to 3.4V.
> Put the charger in LiFe mode & 2A discharge to 2.8V with the charger showing about 3200mAh pack capacity. Voltage bounced back up to 3.22V after 30 minutes.
> Charged back up to 3.6V with the charger showing about 3100mAh pack capacity, and after 30 minutes pack was back down to 3.39V. So the replacement pack is definitely LiFePO4.

Now I'm wondering if the packs in the post light have the same protection circuit in them... and why on Earth do they make LiFePO4 protection circuits that cut off at 4V? It's about 100mAh gained at the cost of loosing charging cycles... The five solar panels on the light are very capable of keeping these charged to 4V even with just 8 hours of clear strong sunlight.

 

[Dave_H]

> It shipped with an adapter which I was able to hook up to my T180 charger. I charged it on Lithium Ion setting & the pack charged to 3.99V before high voltage cutoff kicked in. Within 30 minutes it was back down to 3.4V.
> Put the charger in LiFe mode & 2A discharge to 2.8V with the charger showing about 3200mAh pack capacity. Voltage bounced back up to 3.22V after 30 minutes.
> Charged back up to 3.6V with the charger showing about 3100mAh pack capacity, and after 30 minutes pack was back down to 3.39V. So the replacement pack is definitely LiFePO4.

Now I'm wondering if the packs in the post light have the same protection circuit in them... and why on Earth do they make LiFePO4 protection circuits that cut off at 4V? It's about 100mAh gained at the cost of loosing charging cycles... The five solar panels on the light are very capable of keeping these charged to 4V even with just 8 hours of clear strong sunlight.

LiFePO4 charging is supposed to terminate at around 3.63v, so 4v high cutoff sounds reasonable. If the light is charging them to 4v it is higher than should be. Li-ion charged above 4.2v is bad, not sure about LiFePO4 above 3.6v but would rather not push it (some debate on this earlier, don't recall the outcome).

New Li-ion may be shipped partly discharged i.e. 3.8v which I believe may be a requirement. It would limit amount of available energy in fault condition, and for longer storage life, 40% charge is recommended.

Agreed that Li-ion/LiFePO4 are not the best chemistries for solar lights, especially in areas which drop below freezing (like here). Home Depot sells some nice little solar spots cheap but they use LiFePO4 18500 and come inside in winter. Not clear to me how well the charging circuits cope with varying current availability during charge cycle, or not running to completion, without getting confused. I also wonder about under-voltage shutdown protection on some of these designs, or do they just allow the cell(s) to run down until operation stops.

I open just about everything to check what's inside, including specific chips such as Li-ion charger. Are you able to identify anything inside your lights? Some chips have oddball or no marking, some are identifiable e.g. TP4056. Some have dual capability (Li-ion or LiFePO4).
Are you able to see charge circuits, maybe get an image?

Dave

 

[Dave_H]

I opened up a solar spotlight which uses Li-ion cell. The 8-pin DIP control IC YX8183 is interesting, combines linear constant-current driver with charge circuit. Data on this chip is hard to find and sketchy, but a hobbyist reported some useful info. Current is good only to 250mA or so.

A pin on IC apparently selects Li-ion or LiFePO4. In theory some designs might be able to work with either, but not necessarily. If the cell were not marked (although it was, Li-ion 3.7v 2200mA -> overkill), cell type might have been determined from the design.

Dave