diy solar

diy solar

Controlling chargers with a multi-BMS system

scott harris

New Member
Joined
Jan 1, 2020
Messages
88
If one is building a large battery, one with 4 parallel pacts and thus 4 bms units, but only one charger, how do you connect all 4 BMS units to the charger to lower the current when a pack is nearing the end of its charge?
 
The BMS should never be seen unless it's intervening to avoid a problem. You generally want the equipment responsible for the actions (charging/discharging) to operate inside the BMS protection limits, so it's never tripped.

It's better to program the charger to never hit the point.

16 * 3.65 = 58.4V

If it's not important that you capture 100% of your capacity and only about 90% is fine, then you could pull back to 3.55/cell or 56.8V. The charger you use should be configurable to work with the LFP chemistry and programmable to whatever you want.

As far as the BMS function goes for protection, you essentially need a relay for each BMS such that if any ONE battery has a problem there's a way to open the circuit. Yes, that means you have 4 relays in series. All 4 must be closed to operate. You'll need 4 on the motor supply line as well to cut the load if the battery gets dangerously low.
 
The BMS should never be seen unless it's intervening to avoid a problem. You generally want the equipment responsible for the actions (charging/discharging) to operate inside the BMS protection limits, so it's never tripped.

It's better to program the charger to never hit the point.

16 * 3.65 = 58.4V

If it's not important that you capture 100% of your capacity and only about 90% is fine, then you could pull back to 3.55/cell or 56.8V. The charger you use should be configurable to work with the LFP chemistry and programmable to whatever you want.

As far as the BMS function goes for protection, you essentially need a relay for each BMS such that if any ONE battery has a problem there's a way to open the circuit. Yes, that means you have 4 relays in series. All 4 must be closed to operate. You'll need 4 on the motor supply line as well to cut the load if the battery gets dangerously low.
Thanks for the response

My plan was not for the BMS to open the contactor when the pack reaches max voltage, but for it to reduce the current from the charger and eventually turn the charger off. My application is traction batteries for an electric boat. So the charger is not an SCC but a rack mount charger using shore power. My problem is that for each battery there are 4 packs (each with a BMS) but only one charger. I guess I could have the first pack to reach top just attenuate the charger (which is connected to all 4 packs) and eventually turn it off. This may work ok if the packs stay balanced.

Can I expect my packs to stay balanced or is this too much to count on?
 
I haven't thought very hard about it, and am not familiar with that BMS (beyond knowing what it is, and knowing its a pretty nice and pretty pricey unit), but my initial impression is that you can't achieve what you hope to achieve with your current design.

If you have 4 packs but 1 charger and you want the charger to be the charge control device. Any action taken by 1 BMS will affect all 4 packs, so cutting charging to 1 pack would cut charging to all packs, wouldn't it?

Do you know if the Orion has the ability to throttle current as you are envisioning. I'm not familiar with the capabilities of higher priced BMSes but the ones I am familiar with generally have the ability to cut charging but not to reduce current gradually.

Most people I see using multiple packs in parallel do charge / load control locally at the pack, not for the whole system overall.

One of the reasons many of the high priced BMSes cost what they do is the cost of good documentation and support. It may be worth reaching out to the makers of the Orion Jr and explaining what you want to accomplish and see if they can suggest a solution.
 
Most people I see using multiple packs in parallel do charge / load control locally at the pack, not for the whole system overall.
So are you saying that if someone had multiple parallel packs on a solar installation that they would also have multiple SCCs, one for each pack?
 
So are you saying that if someone had multiple parallel packs on a solar installation that they would also have multiple SCCs, one for each pack?
No, the opposite, they would have one BMS per pack (as I believe you intend to have) and each BMS would control its own individual pack (so that each pack can be managed semi-independently.

There may be ways to accomplish what you want, but it is not something I have seen done personally (if I am correctly understanding your goals). If any BMS can cut charging or discharging to the whole system, it seems to me your system will be constrained by the weakest or smallest pack.

(do bear in mind that I am not very familiar with more advanced high end BMSes or more complex topologies--so there may be a way to accomplish what you seek to accomplish that I have not imagined or understood).
 
I'll chime in on this but want to clarify terms as it's a wee off. Battery "Pack" is a complete battery with X # of cells & it's own BMS. Battery Bank is multiple packs set up as a bank. As I understand your description, you have 4 Battery Packs, each with BMS set in Parallel to make up one bank of batteries. Orion BMS is Top Quality stuff but it is not a Distributed BMS (Think Batrium $$). Even with a distributed BMS, it would have to allow the packs to disconnect independently for Hi/Low Volt or Out of Temp Range conditions while keeping the whole bank running down to the "Last Pack Standing" at which point it could tell Charger to shut down or cut loads out, depending on the situation.

Ideally, each pack should be set up so it should be capable of being the "Last Man Standing" by handling the full Discharge or Charge expected from the system. You can never to certain that all packs will remain connected and don't want to be left in a lurch. The packs should be programmed identically for cutoff points, limits, capable of handling the amps in/out etc.

Chargers are Constant Current / Constant Voltage, it should be capable of reading the amperage, as the packs fill-up the resistance will change to the point they will no longer accept voltage. By that point in time eh BMS should be ready to cutoff. The tricky part is that not all packs will reach that point at the same time because every cells is slightly different and so the "pack" as a whole may vary a bit between each other. If you are using Premium Matched & Batched / Binned cells this is less likely to occur but with the usual Commodity Cells we are using like the EVE-280's (even the Bulk Calbs) will deviate and diverge a bit, on average up to 1mv per AH of capacity, so a 100AH cell can diverge by 100mv and that is not unusual. But the end result is that can make the pack's thresholds vary a bit. The ideal is that the Charger should push as many amps it can or up to 0.5C as that is what makes these LFP's happiest, even though they can take 1C (that heats them up and can cause expansion). The Bank Collectively will absorb all of that until resistance starts to lower the amps being taken in, then as the packs hit Full and the BMS' cuts off, the remaining charge receivers get more and top off and cut, the last pack standing will drop amps to the point of a trickle by which time the charger should shut down.

THE BUGABOO !
The Charger MUST be a good Smart Charger designed for LFP Support. Chargers for Lead, AGM etc are not suitable. Being able to read Amps and have End Amps Monitoring (and preferably configurable). Why configurable ? at 100% Full 3.65V the accepted amp input will be low but not so at 90%, as the resistance won't be there "yet" so if you can program the end amps you can note the amps accepted at 90% and then program the Charger to cut off at that threshold. Most good Solar Controllers use can End Amps as well as many higher-end Inverter/Chargers.

Packs in Parallel will share Load & Charge, they also in a sense balance voltage across the packs in the bank as they are on the common DC Bus. The Common DC Bus for the Packs making up the Bank is quite critical to get right because if done improperly can lead to errors and potential imbalances. I'm sure you've seen it but I'll post this doc from Victron which is great at explaining the parallel pack setup (starting page 17) VICTRON Wiring-Unlimited-EN.pdf

Here is an Excellent write-up from Norkyn Marine as part of their larger series of docs: (There is much more on their site)

And of Course the "Marine How To": https://marinehowto.com/lifepo4-batteries-on-boats/

Hope it helps, Good Luck.
 
Yes. Four packs of 16 commodity 280AHr Eve cells makes up each battery. (There are two batteries one for each outboard motor). The discharge rate is 0.125C and the charge rate is even less (due to limited shore power) a max of 0.1C. So each pack can serve as the "last man standing" and not ever be stressed more than 0.5C

My concern is that I want two methods of stopping charging for each pack. As explained many times in this forum, I should use the BMS as the method of last resort. So I need another method for the charger to turn off on its own accord once a pack is full. It looks like I need a separate charger for each pack then. I guess I should work on that then.

One question. Is it best to taper the current as the pack begins to reach full, or should I just charge at the 0.1C rate until the cell reaches 3.60 volts and suddenly cut off?
 
My knowledge on AC to DC Chargers for LifePo is quite limited.

For example I use a Chargery BMS, if I was to use it in a Separate Port Configuration, which means one DC line for Charging & one DC line for output, each controlled by a Realy/Contactor. I would be able to use one of their chargers which connects directly to One BMS. In that instance, the single BMS would run the show, regardless of what the others state is. For me that would be unacceptable in your scenario. Now, if having Two packs in Parallel and using one BMS to control the charge that would not necessarily be bad thing, as long as you kept a watch, you may have to switch the charger over to the 2nd BMS.... Now that ties directly to the Chargery BMS system and is still limited.

I saw the marine folk have multiple bank charge controllers and limiters, plus some robust chargers. AC to DC Chargers. I think 48V/50A won't be inexpensive for anything good. Hopefully someone with AC to DC CHarging expertise will chime in.

Wish I could help more.
Steve
 
One question. Is it best to taper the current as the pack begins to reach full, or should I just charge at the 0.1C rate until the cell reaches 3.60 volts and suddenly cut off?

Bolded preferred. Although, since you're only charging at 0.1C, you will be very nearly fully charged at 0.1C and 3.60V.

A standard charger should behave that way. LFP and lead-acid batteries essentially have identical charge profiles in concept. Bulk, absorp, float

Charge LFP to 3.60V/cell at up to 0.5C (bulk), hold at 3.60V cell until current is at 0.05C (absorp). Full. Drop to float at 13.6V max.
 
Charger controlled current tapering is typically only driven as a function of cell temperature. Some of the higher density lithium chemistries may need a different current limit at high/low SOC.

Since you are talking less than 0.2C, there isn't any need to regulate charger current based on SOC. Temperature control might be called for, but only if the pack routinely gets hot.

As far as charging goes, just take them up to 3.6V (or lower if you prefer) and terminate. If you are going for maximum calendar life, terminating at 3.45-.3.5V may be more desirable. You give up a couple percent of capacity though.

In your case, I think you probably only need on/off signaling for the charger. That would be triggered by high cell voltage, or high/low cell temperature.

On the motor controller side, you may want current limiting for low SOC. Typically you taper the maximum drive current down to zero as the pack approaches your lower SOC limit. This avoids the BMS dropping out due to low cell voltage. Now, since your max current is fairly low this may not be required. Typically you would see this on EVs which cruise at 0.3C, but can go up to 1C or higher when acceleration is commanded.
 
Back
Top