Go Back  DoItYourself.com Community Forums > Electrical, AC & DC. Electronic Equipment and Computers > Electrical - AC & DC
Reload this Page >

Creating an Array from 18650 cells to work with Solar/Wind/UPS

Creating an Array from 18650 cells to work with Solar/Wind/UPS

Reply

  #1  
Old 10-12-15, 05:27 PM
Member
Thread Starter
Join Date: Oct 2015
Location: USA
Posts: 7
Creating an Array from 18650 cells to work with Solar/Wind/UPS

What I'm thinking of doing is using a 48V 30A/25A/15A Balance Charger. There are 3 different types on eBay I'm curious about:

* tinyurl .com/q2ezbp9 (48V 25A)
* tinyurl .com/pdedpql (48V 30A) (in reviewing, I notice this one is 16S, which would be nominal 59.2V [3.7/cell]; is this safe for a 48V system?)
* tinyurl .com/q4vx79s] (48V 15A)

I wonder about these designs though. Especially the 2nd one since it has one section that has 3 wires going across the same area (bottom right) instead of bridging the gap. I don't see any reason for the wires if it was just all one solder point.

I like the 1st one since it has a better layout and looks like heatsinked top. And the 3rd one has a nice layout as well, just different components.

Any thoughts on these BMS's?

The 18650's I'm going to be getting from laptop/power-tool battery packs, and go through testing each one to make sure they are good. That way I can get each 18650 for less than a dollar.

I will be using a 1600Watt, 48V/18AH (4 * 12V-18AH batteries standard) Battery Smart-BackUPS that will serve as a Grid Charger when no solar/wind is available, or at night (I will be grid tied, but will use primarily alternative power unless battery packs get too low).

So, with 13 * 3.7V = 48.1V/Pack; figuring 2700 - 3000mAh/cell, I would need 18 / 2.7 = 6.6 Packs to match the equivalent of the UPS standard batteries.

17 (6.6 rounded) * 13 = 221 Total Cells. Then, to add more capacity, I could just stack more packs in parallel to increase my AH capacity, right?

This is first part of this project. I'll add more later (monitoring and such abilities).

Am I on the right track? Did I do my math correctly? Any help on this will be greatly appreciated.

Thanks
 
Sponsored Links
  #2  
Old 10-12-15, 05:51 PM
PJmax's Avatar
Group Moderator
Join Date: Oct 2012
Location: NJ - USA
Posts: 43,483
Welcome to the forums.

You did your math correctly but I'm not sure using that many cells in series/parallel packs is particularly reliable. Lithium ion cells are touchy to charge. You need to monitor current, voltage and heat generated by the cells.

You're talking about choosing and connecting "hand tested" cells. When you build series packs using LiIon cells you want all the cells to be of the same lot and age. You'd want to build packs of matched cells. One lazy cell could be a potential explosion nightmare.

My opinion...... stick with lead acid.

Others will stop by and I'll be back. Just wanted to offer a quick opinion.
Also... just FYI.... the DIY board won't allow hotlinking to tinypics...which you may have already found out.
 
  #3  
Old 10-12-15, 08:05 PM
Member
Thread Starter
Join Date: Oct 2015
Location: USA
Posts: 7
Thanks for the info and quick reply.

The way I'm going to test each cell is fully charge using a Li-Ion charger (such as an Imax) and do a a discharge/charge test to determine the proper mAh of each cell. That way I can make sure I match up like cells for each pack. I guess I should mentioned that in the post.

The main reason I would like to use Li-Ion, or even LiPo's, is because of the length of life over lead-acid batteries.

The tiny urls were to ebay links. Ohwell. Learn as ya go
 
  #4  
Old 10-13-15, 06:09 AM
Group Moderator
Join Date: Mar 2003
Location: NC, USA
Posts: 17,123
Your links don't work so I can't see what chargers you are considering. Without seeing them I'm very curious how they balance charge for 13 cells in series. A 16s balancing charger must be something. I'm sure it's got quite a rat's nest of wires

What's a BMS?

I think "testing" each cell's capacity will have minimal benefit. Knowing each cells capacity will help created matched sets but with repeated charge/discharge cycles the cells will get out of balance unless you have some means of maintaining that balance. This is most important if you choose to go with LiPo's as they contain the most energy for their mass and tend to be more "fussy" and I can confirm that they can make a spectacular fire.

Also consider if you will be getting bare 18650 cells or protected ones.

Do you have a method of monitoring the batteries voltage during discharge? Or a way of preventing them from being over discharged?
 
  #5  
Old 10-13-15, 06:52 AM
Member
Thread Starter
Join Date: Oct 2015
Location: USA
Posts: 7
Lightbulb

I had to add a "space" to put the links in since the forum was ******'ing them out. (New to the forum and still figuring out the nuances)
The following are in the same order as the others.

Battery-Protection-BMS-PCB-Board-w-Balance-for-13-Packs-48V-Li-ion-Cell-Max-25A
48V-16-cells-30A-LiFePo4-Lithium-iron-Battery-BMS-Protection-Board-balancing-16S
Battery-Protection-BMS-PCB-Board-w-Balance-for-13-Packs-48V-Li-ion-Cell-max-15A

A BMS is a Battery Management System. From Wiki:

------------
A BMS may monitor the state of the battery as represented by various items, such as:
* Voltage: total voltage, voltages of individual cells, minimum and maximum cell voltage or voltage of periodic taps
* Temperature: average temperature, coolant intake temperature, coolant output temperature, or temperatures of individual cells
* State of charge (SOC) or depth of discharge (DOD): to indicate the charge level of the battery
* State of health (SOH), a variously-defined measurement of the overall condition of the battery
* Coolant flow: for air or fluid cooled batteries
* Current: current in or out of the battery
* Maximum charge current as a charge current limit (CCL)
* Maximum discharge current as a discharge current limit (DCL)
* Energy [kWh] delivered since last charge or charge cycle
* Internal impedance of a cell (to determine open circuit voltage)
* Charge [Ah] delivered or stored (sometimes this feature is called Coulomb counter)
* Total energy delivered since first use
* Total operating time since first use
* Total number of cycles
------------

Although, these BMS's only perform Overcurrent protection (CCL), Undervoltage protection (DCL), and keeps each cell balanced in voltage charge.

And yes, I plan to use the protected 18650's. That's partly why for using the laptop/power-tool battery packs. Each cell is already protected and they are brand names (usually). I have read that Panasonic and Sony (or Sanyo (?)) are the top ones to use.
 
  #6  
Old 10-13-15, 11:34 AM
Group Moderator
Join Date: Mar 2003
Location: NC, USA
Posts: 17,123
Ah, so those boards are just for balancing. How do you plan on doing the charging?

I would read up on stacking a balancing circuit on top of a cell's protection circuit. I have only seen setups that have one or the other but not both. Though I don't know how older computers that used 18650's managed their charging and balancing.
 
  #7  
Old 10-13-15, 02:37 PM
Member
Thread Starter
Join Date: Oct 2015
Location: USA
Posts: 7
Well, that's where the UPS comes in, partially.

To do initial testing with the UPS, it will be keeping the packs charged via the connections on the BMS controllers. They have a charge connection, and a load connection (share a common positive).

Once I know that the packs are working, and as I build the solar panels, they will be tied into the packs as well. The whole DC side of the system will be running 48V.

I'll be using the UPS as a grid tie unit. When/if I need more power, I can turn on the mains breaker and switch over to Grid power. But mainly I'll be using the Solar/Wind-Turbine to supply my power.

Maybe I'll find/build a DC-DC charger I can put between the UPS/Solar that would give me more detail stats of each of the packs' charge status and ratings.

I found someone who has built thier own BMS controller with a variety of options:
diy-open-source-lithium-bms-battery-management-system-20445

Currently reading that post. I might forget about the other ones and build this one myself, or at least start with the others and upgrade to this one later on.
 
  #8  
Old 10-13-15, 08:54 PM
PJmax's Avatar
Group Moderator
Join Date: Oct 2012
Location: NJ - USA
Posts: 43,483
In the link he's using celllog devices to monitor each cell. One module only monitors 8 cells @14.50 each. That's going to get expensive. Interesting design.
 
  #9  
Old 10-14-15, 03:27 AM
Member
Join Date: Aug 2015
Location: USA
Posts: 227
still don't see why you wouldn't use large 12 volt deep cycle lead acid batteries would think they should have a decent life span and certainly probably longer than packs made out of used 18650 batteries that may not have much life left in them.
 
  #10  
Old 10-14-15, 06:07 AM
Group Moderator
Join Date: Mar 2003
Location: NC, USA
Posts: 17,123
One of the main reasons laptops and their batteries are replaced are because they don't last very long anymore (have lost their ability to hold a charge). So, taking used batteries from a battery pack that's mostly shot is probably a lot of work for little storage capacity. New 18650 batteries are available but might be more than you want to spend.

Consider that lithium batteries are used where space and weight are a concern. They are expensive for the amount of energy they store. It's a big reason why UPS power supplies and cars still use old fashioned lead batteries. When weight and space are not a major concern lead is still the cheapest way.
 
  #11  
Old 10-14-15, 06:24 AM
Member
Thread Starter
Join Date: Oct 2015
Location: USA
Posts: 7
To primarily alan73,
The laptop/power-tool packs dont have to be old/used ones. I can purchase new packs off eBay for relative cheap cost. Most cells come to a cost of <$2/cell for legit mAH/V ratings. So, for my above configuration of having approximately 221 cells, it would come to ~$400 or less. That's 48Volt with 18AH. To get that same configuration in Lead-Acid batteries, that comes to at least $400 for just 48Volt. And to use actual dedicated batteries specifically for solar/etc systems, that would be at least $100 or more for a "2V" cells!
True, the AH goes up with that, but it also means that when 1 Lead-Acid battery goes down, the whole system is down. With LiIon/LiPo's, I can loose a pack and still be operation (though at diminished run time obviously), the system is still running. I can also easily expand the capacity.

Yes, I understand it might cost a bit for the parts for the BMS and time consuming putting all the pieces together. But, that's part of the learning curve.
As for the potential of burning down my house; I don't plan on having the charging/distribution station in the house. At most, it would all be in the garage (which, in my case, will be built away from the house with a connecting portico). I'm actually going to have a separate building for all this equipment because the alternative energy producers will not be next to the house. At worse, I might torch a small clump of trees a few hundred feet away.

So, I am playing it safely.

I do appreciate your concern. It does give room for thought, and I am open to suggestions and ideas. Afterall, that's what these forums are for

- Pilot Dane, the batteries would most likely be from new replacement packs, as I mentioned above. I will be using older packs as well, at least initially to do the testing and configuring stages.
A lot of times, a relatively new pack can become non-working or not holding/taking a charge because 1 cell is bad, but the other 5 or 7 (depending on pack size) would be just fine, if not almost brand new. After all, the laptop packs do have regulators built in to protect the packs from over-heating/charging/discharging/etc.
 
  #12  
Old 10-14-15, 06:59 AM
Member
Join Date: Oct 2005
Location: USA
Posts: 1,147
It sounds like you want to play with Li-Ion no matter what. So, go ahead. This is a one off, so not a big deal.

But, the use of many, many Li-Ion cells might not be optimum for a UPS.
First off, Li-Ion doesn't want to sit with a full charge. Optimum standby life is 30 to 50% charge.
This is counter to a UPS design, which calls for full charge for 99% of it's life.
Also, load sharing during discharge is problematic for many small cells, no matter the chemistry.
Charging, as noted past, for many, many cells connected together takes many monitor points to insure charge balance; doable, but getting complex.

Last but not least, you don't need the light weight, small, high energy density that Li-Ion offers.

I'd look for surplus flooded lead acid, or flooded ni-cad of the correct mA-hours, and stack to provide the correct voltage.
 
  #13  
Old 10-14-15, 07:15 AM
Forum Topic Moderator
Join Date: Feb 2005
Location: Near Lansing, Michigan
Posts: 10,295
On the other side, how do you intend to carry out a grid tie using a UPS? Interconnecting an alternatively derived source of power (generator, inverter, etc) to the mains without an interlock is illegal and dangerous. In some jurisdictions it must be a physical switch, some allow a solid state interlock such as what you find on a commercial solar inverter. A UPS does not have the capability to either grid sync or do the automatic interlock.
 
  #14  
Old 10-14-15, 07:48 AM
Member
Thread Starter
Join Date: Oct 2015
Location: USA
Posts: 7
ibpooks:
I thought I already mentioned it, but for the "grid tie", I plan on just throwing the mains breaker. I don't mean the grid tie that feeds "back" into the grid. Yes, that would be illegal without the proper equipment.
It is perfectly legal to use a UPS that's connected, keep the batteries charged via a secondary source, and just have the breaker turned off to the UPS until it's needed to use the mains power.

telecom guy:
I did think about using "returned/used" Lead-Acid batteries from the local auto-parts stores. Usually the cost is about $45/battery, and many of them are practically brand new. I saw a couple of deep-cycle and AGM's the other day that were on their shelf. I may still go that route, if I can get enough of the batteries.
 
  #15  
Old 10-14-15, 07:58 AM
Forum Topic Moderator
Join Date: Feb 2005
Location: Near Lansing, Michigan
Posts: 10,295
Yes it's okay as long as the UPS doesn't feed back into the house wiring. Usually that's what "grid tie" means, so I wanted to make sure you weren't on that path.
 
Reply

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Thread Tools
Search this Thread
Display Modes
'