Common Tool Pack Battery Idea: Need input on Mixed cells

Joined
Jun 15, 2019
Messages
809
So I've got a """fun""" idea for the braintrust.
I go to a warehouse reseller like a hoarder for electronic parts and returns. From this location I've obtained literally thousands of dollars of parts for hundreds, and some of those things are ~20V power tools typically missing their batteries. What started as a simple hand chainsaw with a brushed DC motor has grown to a hedge trimmer, my home cordless drill, a leaf blower, and now an impact gun. Most of these tools were bought for a handful of dollars, with the chainsaw bought for literally 50 cents. Most only have 2-wires for their packs, with some even marking the positive and negative terminals.

On top of bags and BAGS of XT60s, I've also nabbed nearly every lithium battery I could find and I've now salvaged tons of cells- Notably, I have ~111 A&S power cells and over 36 Samsung 18650-28A 2H52 cells. With the average price for a 6-cell laptop battery being 50 cents that means on average, I've spent a total of ~$0.08 cents per cell 8) Give or take a few. Problem however, is most of these cells are arguably pretty terrible with the A&S power ones having only a 1 Amp discharge. The Samsungs aren't much better with internal resistances averaging around 150 milliohm each. They all test as having capacities between 2300 to 2400mah. The Samsung cells seem to have discharge rates 2.5 times the A&S power cells.

Typically these cells would likely be useless- except for those tools, which gave me an idea for my first battery pack. Typical tool packs output 20-30 amps which a MASS of batteries can do, and I have enough (combined of both groups 147 cells) for a big 6S24P battery including a plastic enclosure with a weather seal I nabbed for 50 cents (are you seeing a theme lmao). Such a battery would only pump 22-24 amps depending on the number of parallel strings I have, could use a cheap BMS like One of these, and I have a large repository of 24v wall chargers for real slow, consistent recharging. I also plan to add quality of life and "Good sense" things, like a cell balancer or making it easy for me to test banks of cells. I'd even like to be able to plug it into a 24v inverter for whatever I need, as it would be over a kilowatt. I could put XT60s on my tool packs for connectors, and strap the big pack to a backpack for general use.

My question is- with ALL that nonsense, does that build seem to make sense? I imagine with such low drain the Samsung's high internal resistance wouldn't matter much, and the battery would be more for learning followed by consistent light use. I'd also try the copper-nickel sandwich for the cells as well. What do you all think? Pictured is one layout possible in the case that maxes at 22P, because there's another side about 3cm x 26cm where I should be able to sneak 2 more 6S strings.
 

Attachments

  • pack.jpg
    pack.jpg
    155 KB · Views: 309
Well I frocked up a little. For one, I keep getting my S'es and P's mixed up so this idea is really a 5S 22P battery, which is uh, not common.
And more uncommon is that I can only output ~20 amps total, give or take a few depending on how many cells I can shove in. I don't think there's a single option then for BMS or any other method of testing and maintaining cells, is there?
 
Not directly related to any of your questions, but here is a question for you:


Practical usage question (having used various batteries to power various tools, some of which were corded and not originally battery powered, I've found this is a matter worth consideration before doing):

Are you going to put this battery in a backpack or something, to carry it around, and then a cord to the tool? I dont' think it's going to be convenient to directly mount it to any of the tools.

I ask because a single 18650 is generally in the 70g range, and 5 x 22 of them is 110 times that, or around 7.7kg, not including any casing, wiring, interconnects, etc. So it's going to be at least 15 pounds, and let's just round that up to 20 for planning's sake (nice if it's lighter, but...).

It won't be huge, but it won't be that small, either: About (again, only the cells, not counting casing/etc) 18mm x 5 by 18mm x 22 by 65mm , (a little less with an offset-layer layout rather than grid layout, but not much). That's 90mm x 396mm x 65mm, which is 3.6 x 15.6 x 2.6 inches, assuming a literal 5 x 22 x 1 layout.

Now, some of the tools in some usages may pull more than that 20A (put a wattmeter on some existing power source you can use to test them with, if you like), so you may need more parallel cells, which will make it larger and heavier.


As an example, I have never enjoyed a seven thousand mile long extension cord to do the yard work with for the mower, string trimmer, hedge trimmer, chainsaw, etc. So I tried out my two spare trike traction packs, 14s 2p EIG NMC C020 cells...which are not small, or light, in series to give about the right voltage with as much current as any of these things would "ever" eat up. ;)

They work. But only the mower is capable of carrying the packs, since each one of them weighs nearly 40lbs. :( And that nearly 80lbs is so much it makes it very hard to push the mower around, and they stick up so high there are places I can't trim with them on there.

But having them on some separate cart with a cord on it to the device being used means...I still have a cord. And another big heavy thing to keep track of and pull along if I don't use a thirty thousand mile long cord and just leave the cart in a central location (and there *is* no central location when doing the sidewalk / alley stuff, since the cord won't slide along the top of the chainlink fence, it has to go entirely outside along the sidewalk/etc, and be much longer...or I have to keep going back to move the cart).

So...the mower is the only thing I keep using the packs on, and everythign else I use the fifty thousand mile long extension cord on. ;)

(if I had enough 1p busbars instead of 2p, I could build a 28s 1p battery and then it would be half the weight but still be able to drive the mower...but that's still effectively 40lbs of battery and not gonna carry that around for the toher tools :lol: And a 14s1p would probably drive the string trimmer since it is almost always used in "turtle mode" which has a diode in series with the AC input and thus is only half the voltage anyway...the "rabbit mode" breaks the string off much quicker and easier on any touch to something other than grass)


https://endless-sphere.com/forums/viewtopic.php?f=7&t=111386
 

Attachments

  • 20210415_184255%20-%20Copy[1].jpg
    20210415_184255%20-%20Copy[1].jpg
    47.5 KB · Views: 277
amberwolf said:
Are you going to put this battery in a backpack or something, to carry it around, and then a cord to the tool? I dont' think it's going to be convenient to directly mount it to any of the tools.
Yes! And I'll either solder XT60s to most/all the tool terminals, or build a common adapter from something like the Makita battery adapters I also have snagged for next to nothing. So far, only 1 of these tools has a 3rd wire connection to it's battery- the Worx trimmer- and I'm gonna pop it open to check what it is. For many it's a thermistor. The only one that I won't do that to, is the Impact gun because I found a battery for it- I'd try to make an adapter for that one, but it's only 2 terminals and most chinese power tools share common batteries with only changes in colors. It's case might genuinely be on alibaba, makitas are.

As for why I want to do it-
  • because I need to learn to build batteries and get more comfortable with them- ~25 volts at 22 Amps max is pretty safe.
  • The Cells are basically worthless otherwise.
Biggest problem is that the Samsungs claim they need a charging voltage of 4.3 volts, and that I know of no way to balance and maintain 22 Parallel groups at such a low amperage. Maybe just put an automotive blade fuse inline?

Now, some of the tools in some usages may pull more than that 20A (put a wattmeter on some existing power source you can use to test them with, if you like), so you may need more parallel cells, which will make it larger and heavier.
I would, but I'm not sure how to slap a wattmeter on them to see max draw with no battery for any of them to be powered off of :lol:. Other options are to research the packs and their internals to see what their discharge rates are, and see if this idea is good enough.

As an example...
Yep, that's a good concern and I thought of it too! Mine is also, that in doing this the tools will be imbalanced now that they don't have a counterweight, which is why I'm considering making some kind of weighted "Base" more than just the XT60 connector dangling' around.
I also am enamored with the idea of using it with a 24v inverter for emergency power, but I think i'd need a 7S form factor for that so the voltage doesn't drop too far, and then fix it to a lower voltage with some kind of DC/DC converter. I could do it- I have the spares, believe me- but that might be too big for my case (though that WOULD use all my cells in a neat 7P 21S config...).

Ideas aside, the fact remains- there's no easy or obvious way to keep these cells all in balance. There's simply no BMS for low amps that isn't for []P 7S or so, and while I've found active balancers up to 24S they're also hundreds of dollars for separate standalones for an idea made from literal secondhand parts. It'll still be far cheaper than buying new batteries, but the whole point was avoiding that :lol: Any ideas?
 
CONSIDERABLE SHOUTING said:
Biggest problem is that the Samsungs claim they need a charging voltage of 4.3 volts, and that I know of no way to balance and maintain 22 Parallel groups at such a low amperage.
Maybe just put an automotive blade fuse inline?

I don't really understand the interaction of the parts of your statements, so I'll just leave advice/info for each separate part:

A fuse won't balance or maintain any battery.

All a fuse does is blow for overcurrent above whatever it's limit is, in time and current (see manufacturer spec sheet for any specific fuse to know how it will work).

To balance a battery, you either have to use balancers (automatic or manual) or a BMS on the cells, or individual cell-group chargers, etc.

To maintain a battery, you can either use an automated process (BMS, etc) or manually monitor voltages during charge and discharge, and manually or automatically balance, etc.

I'm not sure what you mean by not knowing a way to balance and maintain the groups "at such a low amperage"; the current doesn't have anything to do with the process of balancing / maintaining.


I would, but I'm not sure how to slap a wattmeter on them to see max draw with no battery for any of them to be powered off of :lol:.
You'd have to interrupt the current path between them and the tool, either inside the battery or inside the tool, or with an adapter between the two.


I also am enamored with the idea of using it with a 24v inverter for emergency power, but I think i'd need a 7S form factor for that so the voltage doesn't drop too far, and then fix it to a lower voltage with some kind of DC/DC converter. I could do it- I have the spares, believe me- but that might be too big for my case (though that WOULD use all my cells in a neat 7P 21S config...)..
Any DC-DC must be able to handle all the surge current of the load as well as the continuous. For motorized devices, or other inductive loads, this can be very high during startup even without a load. So you may need to use a DC-DC that is several times the size and cost that your tools actually need, so it isnt' damaged or destroyed by the surge.

The problem can be the same for a BMS being used to control discharge via FETs on a motorized device that doesn't use a speed controller / ESC like ebikes do; many lawnmowers and other power tools are like that, using brushed motors that don't require controllers for applications that only have on/off control.


Ideas aside, the fact remains- there's no easy or obvious way to keep these cells all in balance. There's simply no BMS for low amps that isn't for []P 7S or so, and while I've found active balancers up to 24S they're also hundreds of dollars for separate standalones for an idea made from literal secondhand parts. It'll still be far cheaper than buying new batteries, but the whole point was avoiding that :lol: Any ideas?
I'm unsure of how many series cells you really need. At one point you've mentioned 22s (22 parallel groups, presumably in series?), 21s, and various powertools that probably run on different voltages but not clear on which ones you need.

Balancing cells can be done by a whole-pack BMS, or just using RC-type cell-checkers to check a bunch of voltages (6-8s at a time) and then manually draining or topping off the ones you need to with single-cell chargers. Or by the active balancers (either the resistive drain-type or the capacitive-charge-shuffle type).

A whole-pack BMS can be a single unit, or several, each controlling only a section of the pack. If all you need it for is to balance and monitor and limit *charge*, and depend on something else for LVC during discharge, this is fairly easy. Each BMS would be wired only to the cells it is monitoring, and you would not run discharge current thru it, only charge. You would use separate chargers for each section, so if you only can get 7s BMS for your purpose, you can have a 21s pack that uses 3 of them, each monitoring only 7 of the parallel groups of cells. Each charger would be the right voltage for only 7s, and would plug into the charge port of just the BMS for the section of the pack it is responsible for. The BMS would be able to shut off or restart charge as needed for each section this way. The only limitation is each charger must be electrically isolated, meaning it's output wires don't connect in any way to it's input wires.
 
Sorry Amber, I knew this whole "idea" was gonna come off like a word salad. It's a bit of a half-baked jumble :lol:

amberwolf said:
I'm not sure what you mean by not knowing a way to balance and maintain the groups "at such a low amperage"; the current doesn't have anything to do with the process of balancing / maintaining.
Okay so, my issue is really twofold- depending on personal choice, i'd either go 21S or 22S, which would be determined by if it's better to go 6P or 7P (7P for if I could use it with a 24v inverter). I have to go for huge numbers of series connections simply because each cell tops out at 1 amp of discharge, and some of these toolpacks might do 20-25 amps.
Sorry for the fuse comment, I should have said that i'd use one as a safety measure too.

You'd have to interrupt the current path between them and the tool, either inside the battery or inside the tool, or with an adapter between the two.
Yeah, like one of those little "Wattman" guys? I could probably snag one, I think they're premade to terminate connections in XT60s now.

Any DC-DC must be able to handle all the surge current of the load as well as the continuous. For motorized devices, or other inductive loads, this can be very high during startup even without a load. So you may need to use a DC-DC that is several times the size and cost that your tools actually need, so it isnt' damaged or destroyed by the surge.

The problem can be the same for a BMS being used to control discharge via FETs on a motorized device that doesn't use a speed controller / ESC like ebikes do; many lawnmowers and other power tools are like that, using brushed motors that don't require controllers for applications that only have on/off control.
Really? I've not even begun to learn about inductive loads. Can you teach me more? My guess is it's related to startup torque.

I'm unsure of how many series cells you really need. At one point you've mentioned 22s (22 parallel groups, presumably in series?), 21s, and various powertools that probably run on different voltages but not clear on which ones you need.
Uh, whichever gives me the best modularity :lol: I could do 6P22S which would give me 22 amps of discharge and a working voltage of ~22 volts, but I would be willing to try 7P21S if it allows me to use an inverter for goofy stuff. The latter might not be a good idea, so I'm fine nixing that idea.

Balancing cells can be done by a whole-pack BMS, or just using RC-type cell-checkers to check a bunch of voltages (6-8s at a time) and then manually draining or topping off the ones you need to with single-cell chargers. Or by the active balancers (either the resistive drain-type or the capacitive-charge-shuffle type).
Really? I have one of those little BattGO balance chargers (I think they can do up to 8S), but I never knew how, or that I could, add balance leads to individual cell banks. Wouldn't balance-charging them independently however, move power into the other batteries in an attempt to equalize the entire pack? Like spilling water on the floor, spreading out to achieve osmosis? Or are they incapable of doing that because nothing is connected to the battery negative and positive to allow electrons to move?
As for the active balancers- I'm interested and would love to hear more, but I know little about them versus common BMS. I know they constantly "trim" battery voltages and have less drain than common BMS, but don't have the same safety tools.

A whole-pack BMS can be a single unit, or several, each controlling only a section of the pack. If all you need it for is to balance and monitor and limit *charge*, and depend on something else for LVC during discharge, this is fairly easy. Each BMS would be wired only to the cells it is monitoring, and you would not run discharge current thru it, only charge. You would use separate chargers for each section, so if you only can get 7s BMS for your purpose, you can have a 21s pack that uses 3 of them, each monitoring only 7 of the parallel groups of cells. Each charger would be the right voltage for only 7s, and would plug into the charge port of just the BMS for the section of the pack it is responsible for. The BMS would be able to shut off or restart charge as needed for each section this way. The only limitation is each charger must be electrically isolated, meaning it's output wires don't connect in any way to it's input wires.
I didn't know I could "Stack" BMS like that! So if I understand you correctly, you would only connect them to the battery negative, and the balance wires? And none of them would be connected to the battery positive? Or am I mistaken?

By what you're telling me, it would likely be best to use multiple active balancers with attendant battery cell-checkers. I'm not sure if I can "set" the BattGO meter to a specific voltage, but that low of amps and volts would make fusing simple and I'd just watch my pack voltage and sag to make sure i'm using it properly. For "Babys first ebike battery" that would be pretty "safe"!

And thank you so much Amber!
 
Back
Top