18650 With PC UPS

kimbo

New member
Joined
Jun 22, 2019
Messages
9
HI

I apologize for my spelling, English is not my native language

I want to make a backup for my small aquaponics system. I collected 18650 cells and I am testing them at the moment. I have a question please and would like some help, please

I have been using a computer UPS with a 35ah SLA for backup, its been working great but I would like to use solar and 18650 cells with the ups and trough the SLA out. I drew a simple picture of how I think it might work


image_zoduxy.jpg


The ups will be on power, and it will charge the battery up to 13.6 volt, too much for the lithium pack hence the dc to dc set to 12.3volt. The diode in the middle is to protect the dc to dc, I am not sure it will like power flowing through the wrong way, the diode on the side is to protect the battery being charged to 13.6volt but it will open the flow from the battery to the ups once the power goes out

Must include a solar charger?

Or how else can I have the ups function without using the ups?
 
with diodes in place, the UPS won't charge the batteries as it can't push current back into them. So they'd only run on solar.

Not sure why you'd only want to charge the battery up to 12.3V. That's quite low for a lead acid and the UPS will charge up to about 14V. 12V is considered almost drained. Most units will disconnect when voltage drops to about 11.5V.

If you go with:
1) 4s, you'll have 12.8V - 16.4V (charging up to 4.1V/cell). However, using the UPS to charge while on grid this causes an issue, as it's a little higher than what it normally would charge to. Mine charges up to 13.7V. So at 3s that'd be 4.5V, way to high for lithiums. At 4s that'd be 3.425V, way too low for top voltage lithiums. Unless you forgo using the UPS to charge all together and then use a buck converter between the battery pack and the UPS

2) Go with LiFePO4 cells instead, as a 4s string would directly replace the Lead Acid battery. No need for special wiring, configuring, or such. The only down side is if using the UPS to charge, it won't do proper charge curve. But, LiFePO4's are lot more tolerant of this than the other Lithium chemistries found in laptop/powertool/etc packs.

I would recommend using Solar to charge the battery only, and use a proper charge controller to do that job. As long as the batteries are charge high enough, the UPS won't try to charge it. If you can communicate with the UPS (usually through a serial coms) then you might be able to turn off charging completely, or at least set it's limits so high that it would basically never turn on.

What make/model UPS do you have?
 
Korishan said:
with diodes in place, the UPS won't charge the batteries as it can't push current back into them. So they'd only run on solar.

Not sure why you'd only want to charge the battery up to 12.3V. That's quite low for a lead acid and the UPS will charge up to about 14V. 12V is considered almost drained. Most units will disconnect when voltage drops to about 11.5V.

If you go with:
1) 4s, you'll have 12.8V - 16.4V (charging up to 4.1V/cell). However, using the UPS to charge while on grid this causes an issue, as it's a little higher than what it normally would charge to. Mine charges up to 13.7V. So at 3s that'd be 4.5V, way to high for lithiums. At 4s that'd be 3.425V, way too low for top voltage lithiums. Unless you forgo using the UPS to charge all together and then use a buck converter between the battery pack and the UPS

2) Go with LiFePO4 cells instead, as a 4s string would directly replace the Lead Acid battery. No need for special wiring, configuring, or such. The only down side is if using the UPS to charge, it won't do proper charge curve. But, LiFePO4's are lot more tolerant of this than the other Lithium chemistries found in laptop/powertool/etc packs.

I would recommend using Solar to charge the battery only, and use a proper charge controller to do that job. As long as the batteries are charge high enough, the UPS won't try to charge it. If you can communicate with the UPS (usually through a serial coms) then you might be able to turn off charging completely, or at least set it's limits so high that it would basically never turn on.

What make/model UPS do you have?

Thank you

I thought he diodes will block the ups from charging them (the 18650's)

Its a bit late here i will take some pics tomorrow, but it is a cheepy

i have some 18650 cells i am in the process of checking, wile i am doing that i am just gattering some knolage so i can take the next step

Can i maybe use something like this: https://www.aliexpress.com/item/AC-...048441.html?spm=a2g0o.cart.0.0.757e3c00POOFOA

Or is there a better way
 
Please respond "after" the [/quote] block. Makes it easier to read replies

Not sure what the Switching Controller Module would be used for. That's only for AC loads from what I can tell.
If you are trying to switch between charger solutions, there's not really a need. As long as the UPS sees the upper voltage, it won't try to charge in the first place

The diodes could keep the UPS from charging, but it could also keep it from properly seeing/sensing the batteries in the first place. Also, diodes have about 1V drop each, so you'd loose up to 2V across those two diodes. In this situation, that's really a big deal.
 
Korishan said:
Please respond "after" the [ /quote]

Morning

I am sorry, i am new to this forum software and did not notice.

Last night i as looking at this


image_qeavmp.jpg


Apologies i am visual so i drew this out for me, i was just thinking the UPS might be in the way, you said if it sees the upper limit of the battery it will not charge, i had a normal charger on the SLA and it was at 13.8 and the UPS freaked, i thought it might because the battery is over charged

Back to the drawing, can that setup work, must i rather put a BMS on as well?
 
Here is the pic of the UPS, like i said it is a cheepy, but i see it have a serial port


image_istmsl.jpg
 
Maybe you can measure the voltages on the SLA when charged to confirm.
Like Korishan said, 4s LiFeP04 cells are a direct match for SLA voltages and you should be able to just install them.

You can use an inexpensive PWM solar controller between the panels & the battery instead of a buck converter.
 
Redpacket said:
Maybe you can measure the voltages on the SLA when charged to confirm.
Like Korishan said, 4s LiFeP04 cells are a direct match for SLA voltages and you should be able to just install them.

You can use an inexpensive PWM solar controller between the panels & the battery instead of a buck converter.

Thank you for the reply

Here in south africa cells are not readily available and i was just lucky to get some 18650 cells, also i am in a wheelchair and disability grant so money is very scares, so i must use what i have.
 
kimbo said:
Here is the pic of the UPS, like i said it is a cheepy, but i see it have a serial port


image_istmsl.jpg

That UPS will not be very suitable for long term use as an inverter. Most consumer grade UPS are not designed to run any longer than the original battery would have powered them. Even with additional cooling, it is the transformer itself that will be underrated. I wouldn't use much more than about 25% of its rated output long term. It also is very unlikely to be pure sine output either.

If you can, your best bet is to look for a server grade UPS, it will most likely be 24v, be capable longer run times, and have a pure sine output.

Your second diagram looks more promising. However, bear in mind, the switch over time could be a problem, depending how sensitive your load is.
 
Geek said:
kimbo said:
Here is the pic of the UPS, like i said it is a cheepy, but i see it have a serial port


image_istmsl.jpg
Hi

Thank you for your answer, that is the reason i want to change the setup, because i know the ups is not good for an extended time of use

regarding the second diagram: Thank you, do you think i must run a BMS ons the battery, also must i include a solar charge controler or will that be enough?

My solar panel dont have a blocking diode on it, how do i size a diode for there?

Just a note: Looks like the forum sofware dont like quote in quote, or i am doing something wrong
 
kimbo said:
Just a note: Looks like the forum sofware dont like quote in quote, or i am doing something wrong

I'm guessing you are using the "Reply" button to post. If so, just move your cursor further down outside the quote box. Press the down arrow till you are outside the box. I just tested and it works.
The other method is to use the "Quote" button and then insert in the quick reply.
 
Korishan how do i size a diode, my solar panel has no diode on and i want to add one but are not sure how big, can you please help me
 
kimbo said:
Korishan how do i size a diode, my solar panel has no diode on and i want to add one but are not sure how big, can you please help me

That one's actually a little beyond my knowledge. I'm not sure of the part numbers like others are.
But I believe if you just went with a schottky diode that is rated for your voltage and amps, you'll be fine. The diode needs to have the least amount of forward drop as possible (in this case, schottky's are usually in the .6-.8V range).
If you do a search for "schottky solar panel diode" you'll probably get a bunch of the right ones.

There's a few on here who are solar wizards and I'm sure they'll chime in
 
Back
Top