Would you charge a 4s LiFePo with a "15V" wall wart (intended for charging a 12V SLA)?

Buk___

10 kW
Joined
Jul 28, 2017
Messages
750
The title says it all.

I'm looking to replace a 12V SLA (portable vacuum) with Li***, and would like to reuse the existing charger, because it is a wall mount&charger with bespoke connectors, and to save money.

4 x 3.65 = 14.6; but the 15V is the OCV; once loaded it will inevitably drop.

Would you do it?
 
I recently soldered up a 3s-3P battery (4.2V cells) and bought a 3S charger on ebay for $11 shipped.

For that kind of money, I didn't have to worry about whether which one of the many wall warts I have in a box might have worked.

On the other hand, I still charge this battery outside, so I could have used a wall wart.
 
I wouldn't charge LiFePO4 up to 4.2V per cell, but...I have read that doing so will not cause the electrolyte to off-gas. This is touted as one of the reasons LiFePO4 is considered to be very safe, and less likely to catch fire.

I have recently become interested in 4S LiFePO4 as a 12V SLA replacement in several applications. The standard car battery charger puts out 13.8V, as does a car alternator. Divided by 4 cells that is 3.45V per cell, well below the recommended charge of 3.65V

It could be argued that using 13.8V for a 4S pack would never come close to overcharging, but also that...the pack would end up not carrying a full charge (less range). This is something that I will be exploring because the job I have for this 12V pack will not require long range.

If the wall wart is truly putting out 15V, then a 4S pack would be getting 15/4= 3.75V per cell. This is a hair more than the recommended 3.65V, but...it is also far away from the 4.2V danger zone. I wouldn't do this to an expensive cell, but I must admit I am curious as to how this would turn out...

I recall a long time ago, there was a build that used small and individual 4.1V chargers for each parallel string in the pack. I seem to remember they were turned down as far as they would go to about 4.0V?...
 
3.5v is full for lifepo4 there is very little capacity above that. 3.8v is something it could tolerate. But the SLA charger doesn't disconnect if goes into trickle charge best to charge it then unplug it.
 
What does the waveform out of the wallwart look like? Is it actually good DC? Or lots of AC ripple in it?

I've run into lots of cheap SLA-powered stuff that lets the battery do most of the AC smoothing, and I wouldn't do that to a non-SLA battery. (Wouldn't do ti to SLA either, but it handles it better if you had to).

If the wallwart is heavy, then with a high OCV it's probalby just a transformer and diodes (maybe not even a full-bridge) and caps, no regulators.

So it also means it has no real current control. If its' a really tiny wallwart then it's probably incapable of providing much current without enough voltage sag to self-limit current (by teh difference in voltage vs the cell resistance at that voltage), but if it's big enough it could provide more charge current than the pack you're installing should be charged with.


As for voltage, if the new pack has a BMS it should prevent overcharge as long as the BMS works; if it fails or is not designed correctly then it could overcharge the cells a bit every time it'scharged, which may over time shorten the lifespan of the pack.
 
Because the charger trickles, no. Unless it does have a reliable bms.

Good excuse to buy a 50w RC charger IMO, like a b6
 
dogman dan said:
Because the charger trickles, no. Unless it does have a reliable bms.

Good excuse to buy a 50w RC charger IMO, like a b6

The device has some sort of cut-off; there's an led that goes from red to green once the (SLA) battery is charged. And it doesn't go back to red until it has been used for a good amount of time, so it isn't a simple voltage threshold detection.

I spent yesterday trying to reverse engineer the circuit so I could work out what is being detected, but it's an infuriatingly complex circuit for doing some thing essentially simple and getting accurate specs on the parts is proving harder than I expected :(

circuits.jpg
 
That's a much more complex regulator system than I've seen on any of the SLA-powered devices I've worked on so far (except UPS units)--most of them depend on the battery smoothing stuff out.

It makes it more likely it's got some safety cutoffs, but whether they are appropriate for anything but SLA might be another story.
 
amberwolf said:
It makes it more likely it's got some safety cutoffs, but whether they are appropriate for anything but SLA might be another story.

Indeed. That's why I spent yesterday trying to simulate it. So I can try to figure out what circumstances control the "charged" light and associated circuits.

BTW: I"deglued" the wall wart and it has a full bridge and a smoothing capacitor.
 
Huge fan of the LiFePO4 chemistry here!

I have used one but monitored voltage as it reaches top and disconect
 
amberwolf said:
That's a much more complex regulator system than I've seen on any of the SLA-powered devices I've worked on so far (except UPS units)--most of them depend on the battery smoothing stuff out.

It makes it more likely it's got some safety cutoffs, but whether they are appropriate for anything but SLA might be another story.

I finally got my simulation working with a few guestimations based on stuff I can measure on the real circuit, it looks to enable charging below 12.4V, and cut-off above 13.5V; (thought the trigger appears to be the rapid rise to 15V when the SLA is charged, but that could be just my rapidly dying SLA) which is probably too restrictive a range to be useful for 4S LiFePO4 :(

Worse, it seems like it also does trickle charge, though I can't work out what controls or triggers it.

The (imperfect) sim should anyone be interested.
 
mark963 said:
Isn't Q1 (the S9012) a PNP transistor, not an NPN?

Um. You may well be correct?

I downloaded the spec. sheet from mouser, and nowhere does it actually say whether its NPN or PNP, nor (to my limited knowledge) provide any other indication of that?

I also have the datasheet for the SS9013 which spells it out "NPN Epitaxial Silicon Transistor"; so I assumed they were variations on a theme per the ids.

[strike]If you know definitively different, that could make a big difference to things?[/strike]

And then I do a search and there right in the first match is "SS9012: PNP Epitaxial Silicon Transistor - ON Semiconductor". Thanks for the heads-up! :oops:
 
yes
I would but with some kind of reliable LVC.
Since such "wallwart" doesnt have any CC/CV charging profile it might be hard on LiFePo cells.
 
If the wall wart was made for charging lead-acid, then it should have some kind of current limiting and should work with LiFePO4.

I would worry a little about balancing though. Some LiFePO4 are prone to getting out of balance. Check individual voltages after charging.
 
In reading more carefully, it seems the circuit board might do the limiting and the wall wart is just a raw supply. It will need some kind of current limiting to avoid overloading the transformer.
 
999zip999 said:
So can we see the brand the manufacturer of the name of the Lifepo4 you're going to use

I haven't got to even considering that part yet. The SLA needs replacing again -- they seem to last about 5 years -- and is going to cost around £12.

I wondered whether I could replace it with some form of Li pack, and looking around LiFePO4 seems to be the closest chemistry to being "fit & forget".

I'd very much have liked to convert it such that is can be used with the existing wall mount and wall-wart and in the same way -- use it and drop it back into the holster and it'll be charged for next time.

I don't want to have to half disassemble the thing, transfer the cells/pack to a different charger unit, monitor them closely for an hour just-in-case, reassemble. I'll throw it away and buy a new one (or just buy another SLA) rather than have that faff.

The SLA is 48 x 70 x 110, so I'd need to fit 4 x 3.2 /3.8 LiFePO4 cells, or 3 x 3.7/4.2 LiCo cells into that space along with any BMS etc.

I naively hoped that the existing SLA charging board internal to the unit might work, or be adapted to work directly with the pack I make, thus avoiding the need to purchase a separate unit. The wall wart is spec'd at 15V@200mA, is full-wave rectified and smoothed. The charge/ready cycle appears to be controlled by 12V zener diodes. Our habit is to switch the wall-wart off once the green ready LED shows -- my wife is very dilligent about such things.

My thought was that if I replaced the 12V zeners with 14.4v ones, that would give me close enough to the 3.8V/15.2V max for the LiFePO4 and undercharge would be good for longevity. Provided we religiously turned it off once green shows, we would avoid any trickle charging.

4S LiFePO4 balance boards are common and cheap, which should take care of balancing for such a low demand (~100-120W) pack.

The whole idea has taken on a life of its own as I've spent an inordinate amount time trying to model and understand the exist circuit board, just cos its interesting.
 
Just because it's printed on the side of the charger doesn't mean what the charger does or doesn't output do you have a multimeter and can you check it with the pros and find the exact voltage please
 
999zip999 said:
Just because it's printed on the side of the charger doesn't mean what the charger does or doesn't output do you have a multimeter and can you check it with the pros and find the exact voltage please

The open circuit voltage floats at around 25V. Once connected to the charging circuit it drops to 15V at the input plug, but ~13.5V is supplied to the SLA.
 
Back
Top