diy solar

diy solar

4S No-Balance Challenge!

Substrate

Solar Addict
Joined
Apr 28, 2021
Messages
1,545
Location
SoCal
Who this is for:
Battery nerds that have top-balancing gear and maybe some spare similar cells. I'm one of them that has done this.

What this is NOT:
Not an anti-bms campaign. You are going to be using an external fuse and an LVD like a Victron or similar.
Not promoting unsafe operating conditions.
Not promoting a "cheap" alternative to a bms. These days, it actually costs more if you don't have stuff.
Not promoting the use of wiring up trash random cells with speaker wire.
Not a production-quality battery that you give to your unskilled neighbor.
Not designed to cause a furious forum uproar with any hidden campaign.

Objective:
After an initial top-balance, one does not dance around with a bleeder all over the cells to try and make them all perfectly equal in voltage. A 4S 12v configuration keeps it manageable. Your application could be as simple as an inexpensive msw inverter and a grocery-store led bulb in a house fixture as your bench light.

Operational goal with a practical twist:
Because each cell has slightly different amounts of capacity and internal resistance, after an initial top balance, in operation a wider delta-variance up to 100mv or so is allowed because they are going to dance around a bit. If your top balance is good, and your cells aren't totally different, then a CV is set to get them no wider than between say 3.45 and 3.55. Or say 3.5 and 3.6v.

A low CV is recommended, like 14.0 to maybe 14.2v

Unlike other top-balance demonstrations which show ops dancing around with a manual bleeder resistor, you are going to simply let this 100mv delta variance be. The battery nerd in you will be watching this perhaps every cycle with your multimeter collecting dust. Later, after many cycles, you may be confident enough to not do this each and every time.

If you can't obtain a near 100mv variance or less, that means your top-balance was not good in the first place, or you have too much of a variance of capacity or internal resistance in your cells, or your wiring structure has high-resistance elements. (you cleaned your terminals and bus-bars right? Ring terminals instead of clip leads!)

HERESY!
Not really. This is just an interesting experiment for us battery nerds. Will your battery operated in this manner degrade so fast that you won't get past a week's use? Or will it last 8 years? You tell me! :)

What it will do is make use of some of the gear you bought to do a top balance on your 48v bank only once, and maybe make use of some spare cells getting lonely in the corner. And you aren't dancing around with a manual bleeder, which from a propeller-head standpoint, sometimes making cells exactly equal in voltage is actually UNbalancing a bank. It's a fine detail, but don't tell anyone.
 
Can you state this with a lot less words?

Under "Objective:" I don't actually see an objective.

"Operational goal" - is it to maintain operation with no more than 100mV variation?

If so, would this satisfy your challenge?

I take my 4S Navitas 25Ah cells with a JBD BMS and disable all balancing after an initial balance. Then just cycle the shit out of it between 10.8V and 14.2V? Check it every X cycles at min/max 4S voltage to identify dV?
 
Can you state this with a lot less words?
Not until I stop drinking espresso. :) I do appreciate the feedback!

Under "Objective:" I don't actually see an objective.
Yeah, kind of unclear. Make the initially top-balanced system practical, by not having to dance around with a manual bleeder of some sort after the cells have been put into series for use. For many, but not all, this method is imprecise and involves a lot of dancing. I've done that dance more than I liked and never got it quite right. :)

"Operational goal" - is it to maintain operation with no more than 100mV variation?
Right. Less is even better, but that means all your cells are quite well matched to begin with.

The *worst* case would be a low of 3.45, and a high of 3.65v for the delta, but that's a bit too extreme. If your initial top balance was good, then this is what I'd call an unstable situation which calls for a LOT of babysitting, swapping out cells to get a closer match etc.

So only a 100mv delta or less seems to make it manageable.

If so, would this satisfy your challenge?

I take my 4S Navitas 25Ah cells with a JBD BMS and disable all balancing after an initial balance. Then just cycle the shit out of it between 10.8V and 14.2V? Check it every X cycles at min/max 4S voltage to identify dV?

You got it, and brings up an interesting point for those who may be able to actually disable balancing (after the initial sanity-check top balancing has been done of course).

Personally, when I run like this, my LVD is a bit higher, so I prefer to stay out of even the start of the knee, so usually set a bit more conservatively to 12.2 - 12.5v instead of 10.8. Ops choice!

On, paper, assuming perfection with cells, and you desired to CV of 3.5v, that would mean an overall 4S pack-level charge of 14.0v cv. Of course, this rarely works out in the real world, so a slight delta of 100mv or less, provided it doesn't shoot a cell above 3.6v might be quite acceptable for this. Tweak your CV on your charger or SCC to accomodate.

I prefer to avoid the bottom and top. CV is tweaked until cells are delta at like 3.48 to 3.58v but who's counting? :)
 
Last edited:
I've done this, ran a crypto rig for a month. Needed to charge at 3.42 per cell max to keep one (or two) from running away.

Worked fine. But I fully expected it to eventually fail in a bad way if ran for too long, so I shut her down.

EDIT Found it:
 
You kids! :) I'm talking reliving the days of trying to do the dance on a 96 cell EV and the bleeder consists of 100 feet of insulated fence-wire coiled into a bucket of water.

Here with this experiment, we don't want to dance and live with no more than 100mv delta.
 
Perfection and production quality? NO.

Drive all your cells to zero-current? Don't do that for too many hours. Set a timer on your phone. Use an ac-timer attached to your charger as a safety backup.

Do you ALWAYS charge to full? Not if you don't want to. Pull the plug.

Lower your CV? Sure, after demonstrating that the system is stable within 100mv delta at higher cv's like 14.0 / 14.2 or so, feel free to back it down if you like.

The options here are numerous, and to your liking. Much discussed in the other thread categories.
 
It is an interesting subject. I mainly run with my balancing disabled. Every few months i check the voltages when the battery is full, and when the dV is over 0.1V when the highest cell is at 3.5V i will enable the balancing for a while.

The objective here is to minimise the time cells are held above 3.4V in a charging condition.

With good cells and staying above 3.0V i think you’ll go many years before you require balancing.
 
Sure enough. A related experiment, for those who CAN turn off their BMS balancing, is to see what happens on the very next charge. If the cells are reasonably similar, you should have no runners.

If you do, then that is an indication that the bms balancer is hiding the poor cell quality, and is now a nanny band-aid with the bleeders active all the time.

Don't get me started Tom. Otherwise you'll have me filling the bed of a pickup with agm's, and whipping up a bunch of Lee-Hart batt-bridges ! :)

(which might still be useful today for those running lead-acid 24 to 48v banks)
 
Last edited:
Personally, when I run like this, my LVD is a bit higher, so I prefer to stay out of even the start of the knee, so usually set a bit more conservatively to 12.2 - 12.5v instead of 10.8. Ops choice!

On, paper, assuming perfection with cells, and you desired to CV of 3.5v, that would mean an overall 4S pack-level charge of 14.0v cv. Of course, this rarely works out in the real world, so a slight delta of 100mv or less, provided it doesn't shoot a cell above 3.6v might be quite acceptable for this. Tweak your CV on your charger or SCC to accomodate.

I prefer to avoid the bottom and top. CV is tweaked until cells are delta at like 3.48 to 3.58v but who's counting? :)

I have no use for these cells. They're just cute. Had to have them. I've cycled them individually, and they come in right at or just above 25Ah.

Every once in a while I pull it off the shelf after sitting for a few months and run it up to 14.4V to check the balance. It's out of balance by much with the runner at 3.65 and the other three at 3.4X.

My thinking was to make it a stress test. My equipment can cycle and log everything. 10.8-14.2V would be at the extremes of real-world use. I should be able to do it at 1C even though these are 3C rated.
 
Yup, something similar here.

I’ve done a 16s without a top balance using the Chargery BMS with it’s minuscule passive balance current on a 60 Ah battery.

Just started another 16s with 100Ah CALB cells that I will use the Chargery on until my JK BMS gets here on the slow boat from china. No top balance for these either. These will serve in parallel with my EG4 that is over 2 years old.

I’ve never confessed doing this but since we’re in the Danger Zone I can ‘fess up. I’m not recommending anyone follow my lead here.

I think since I use relatively small cells and live well between the knees (usually 3.1 to 3.4 V daily cycle) it works for me.
 
Last edited:
I have no use for these cells. They're just cute. Had to have them. I've cycled them individually, and they come in right at or just above 25Ah.

If anyone else wants to play they can get 4 cute cells like those for $90 (22.50 each) here :

 
I saw those too and it was my first thought! Keeping it relatively inexpensive, a CZH-Labs 30A LVD with dip switches makes that part easy. At that small ah, even a hobby-charger if one knows how to use it could be employed.

Heck, what used to be the only way to do it, has become danger-zone time. :)

What makes it a tad easier is that we are using it for sub-c application. Not driving an EV with hundreds of amps flowing going up your driveway with the lvd turned off. :)

I guess I'm so wordy about it is that it is just FUN if one stays safe and sane about it.
 
If anyone else wants to play they can get 4 cute cells like those for $90 (22.50 each) here :

Hmmm I've been eyeing these which seem like a better deal but twice the price.
 
Or a 4-pack of 700ah Winstons.

I'll fess up though. I had an application Joe will like - an Elecraft K2 qrp rig and loop preamp.

Used the cells that Andy was tripping on here (yeah, 5300mah), and he later put a 35A hurt on them. My K2 draws far less. Notice the fun factor? Source: Amazon Cityork w/holders.


Sure I could just go off-the shelf, but it wasn't a good match for the philosophy of the Elecraft K2. :)
 
Hmmm I've been eyeing these which seem like a better deal but twice the price.

You might as well jump up to 100 Ah for only a little more :


That’s what I’m using for the new build.
 
I power my KX3 with these on SOTA activations:


I'm glad you brought that up, since it is a clear demonstration between a "power" cell and an "energy" cell.

The glut of 32650/32700's you see out there are high-c "power" cells. When Andy did his discharge test at 35A continuous, it looks very similar to what you would get out of an A123 cylindrical. The typical application would be motorcycle LFP starter batteries, where 4 of these would be in a little box. In earlier days, there was NO bms or balancing (except done once at the factory). Perhaps this glut of cells came from a motorcycle battery manufacturer that went under.

For an "energy" cell, like a solar-powered garden light LFP battery, it is designed for long term low-current draw.

If you were to parallel these 14430 sized energy cell LFP's to be the equivalent to the capacity of the power-cell, it still wouldn't start a bike. Or at least not many times. But perfect for your application. Not all solar-power garden lights are junk either. Westinghouse made some nice ones.

(Heh, KX3, my man!)

So now, one can play the pricing game. You can use a power-cell in a slow-draw application of course. But you can't use an energy-cell to start a bike. So when you look at it, the solar-garden cells should be about $1 a piece. :)

The flip side, is one would have to be totally crazy to pay $300 bucks for a powersports 5Ah starter battery to power the low-current KX3.
 
Last edited:
Re those 100ah Calbs you're looking at ...

If you are going to do this "top balance once, and no balance afterwards unless it exceeds 100mv delta" game, then buy 5 just in case.

Here is where a reputable dealer is important, because you *might* get away with only buying 4 and not have a dud. This is why I caution those that think this is a cheap option by not using a bms balancer, might not be! :)
 
Yup, KX3 working Japan and France from an Arizona mountain top back to back on 20m CW at 5W with a quarter wave vertical was a hoot !
 
Back
Top