BMS Settings for 12V LiFePO4 Using NOCO GEN5x1 Charger

[Dr_Zu]

Hey everyone. I'm an 30-year veteran medical device design engineer, but my training is mechanical / biomedical . . . so I'm relatively unequipped to understand the nuances of electrical engineering and I'm a total noob to LiFePO4 batteries. This is my first post here, but I'll make it a really long one to compensate for all of the posting that I haven't done in the past! Seriously though, I really appreciate those of you who are active on the forum. I've learned a TON from all of you! I appreciate, even more, those of you who have the stamina to read through all of this post. I'll be as brief as possible, but that's not my strong suit!


I just finished building a 12V LiFePO4 4-cell pack (see pics below) based on the knowledge that I gained here, from Will's videos, and from many late nights drinking good beer while reading and watching videos from all over the web. My BMS is from Overkill Solar and my application for the cell pack is powering the accessories on my pontoon boat (the boat is always used on a small ~45 acre lake so my application is far from "mission critical" . . . unless you consider losing tuneage mission critical, that is). All cells (ExpertPower 100Ah cells) were top balanced (using a constant voltage / currant power supply) and capacity checked (using one of the inexpensive load fixtures Will has reviewed) before being built into a pack. All cells checked out at over the 100Ah spec using the discharge fixture at 0.2C. I'm using a NOCO GEN5x1 Charger to charge the cell pack (set to Li-Ion setting, which NOCO claims is LiFePO4 compatible).



So I first had the BMS protection for cell overvoltage set to 3650 mV, which seemed reasonable based on what I had read online. What I quickly noticed is that the BMS would alarm for cell overvoltage before the charger had completed charging the cell pack (according to its proprietary algorithm). I did some searching and realized that the charger was designed to charge to 14.6 V (= 4 x 3650 mV) before shutting down. Even with a slight (and inevitable) cell imbalance of a few mV, the BMS would detect a cell overvoltage and fault. Hmmm.


More research (and beer) and I decided (wrongly?) that it would be OK to tweak the cell overvoltage setting to 3700 mV (see below for current BMS settings). After that, all was copesetic! The GEN5x1 charger fully charged the cell pack and stopped charging without creating any faults. All seemed right in the world . . . except that nagging feeling that it wasn't!



I reached out to NOCO to get their recommendations on how I should set up my BMS, which was a dumb question on my part! That's way too much liability for them to take on; helping me build my custom cell pack with components they neither make nor specify!?!? I should know better based on my own product development experience! I blame the beer on that one! However, good on them for giving me enough information (without disclosing any proprietary algorithms) to understand the basics of how the NOCO GEN5x1 charger works in Li-Ion mode:

• Constant current charge at ~5 amps until the cell pack hits 13.2 V
• Reduce constant current charge to ~2.5 amps (2 amps in my observations) until the cell pack hits 14.6 V
• Stop charging and monitor cell pack voltage.
• If cell pack voltage drops below 12.6 V, restart charging algorithm.
• Every 24 hr., restart charging algorithm (topping up cell pack)

Sooooo, there are a few questions still keeping me up (and drinking beer) at night:

• Is a BMS protection for cell overvoltage setting of 3700 mV too high (assuming that I want to optimize cell life)?
• If "yes", how do I keep the BMS from triggering a fault if I reduce it since I can't change the charger settings?
• Is the charger's daily "top off" mode OK for LiFePO4 cells. My gut is telling me "no" based on my research, but I can't point to anything specific. In any case, it's inherent to the GEN5x1 charging algorithm so there's nothing I can do to change it (except get a different charger).
• Is there a different/better charger that I should be using? My understanding is that LiFePO4 cells like to live in the 20% to 80 % SOC range, which GEN5x1 won't accommodate (since it always charges to 100%). Also, this isn't exactly a "marine" application (fresh water), but I need a charger that can handle an occasional indirect splash.
• Adjacent question, but what is the "Trig. Delay" setting in the BMS? I had assumed seconds, but it doesn't say and I haven't found any specific documentation on it.

Underlying motive for this post is a second build that I'm working on to power my 48V Torqeedo electric boat motor. The little lake that I live on is designated "electric-motor-only" to protect the community drinking water (there are well-heads in the area). I've been using AGM cells for the past 5 years, but they're on their last legs and I want to build a LiFePO4 cell pack to replace them with (the 16 cells in question are being top-balanced as I type - below). This 12 V accessory battery is my test run!


Full off-grid solar power once I retire and have more time (and less money) to play! I really appreciate any feedback you all have. Thanks in advance!!!

 

[sunshine_eggo]

Eff me in the aay.

Noco is so overrated. What a stupid algorithm. It essentially guarantees that your BMS MUST terminate the charge - a great way to get a BMS fail.

answer to all your questions:

Shitcan the Noco and use your benchtop PS as a charger. Charge to 14.4V. Done when current has dropped to 5A.

 

[time2roll]

Nothing particularly terrible or immediate will come of it for staying at 3.700 protection. I believe most premade batteries recommend up to 14.600 volts charging and must have protection set 3.700 to 3.800 or as you found there will be BMS protection. These batteries have a warranty.

The only real solution would be to replace the NOCO with a charger with a lower peak voltage. Possibly use the GEL or AGM setting or other program that does not equalize but comes in a bit lower.

I use a Meanwell NPB series. Has a couple standard programs to go 14.2 - 14.4 zone.

 

[Dr_Zu]

Eff me in the aay.

Noco is so overrated. What a stupid algorithm. It essentially guarantees that your BMS MUST terminate the charge - a great way to get a BMS fail.

answer to all your questions:

Shitcan the Noco and use your benchtop PS as a charger. Charge to 14.4V. Done when current has dropped to 5A.

I was afraid someone would say that, but it's what my gut was telling me. Old addage - "Never ask a question that you don't want the answer to"!

Unfortunately, a benchtop power supply won't work well for my application. It would almost certainly get wet at some point in my boat, though I suppose I could put it in some type of box (as I did the cell pack / BMS). More importantly though, I also need a charger that I can plug in and walk away from. A nice to have would be a charger that would double as a power source when I'm playing the radio while I'm at the dock, which the GEN5x1 seems to do well (no apparent voltage spikes when disconnecting the cell pack during charging under load).

As noted earlier, I'm not an electrical engineer so this may be a dumb question: Could I put a voltage divider one of the charger legs so the max voltage that the cell pack sees is, say, 14.4 V instead of 14.6 V and then let the GEN5x1 take care of the rest? It would still do the daily top-off, but it would do it at a lower SOC, which seems harmless???

Thanks!

 

[Dr_Zu]

Nothing particularly terrible or immediate will come of it for staying at 3.700 protection. I believe most premade batteries recommend up to 14.600 volts charging and must have protection set 3.700 to 3.800 or as you found there will be BMS protection. These batteries have a warranty.

The only real solution would be to replace the NOCO with a charger with a lower peak voltage. Possibly use the GEL or AGM setting or other program that does not equalize but comes in a bit lower.

I use a Meanwell NPB series. Has a couple standard programs to go 14.2 - 14.4 zone.

The Mean Well NPB-750-48 looks like it would be a solid choice for my 48V cell pack, though I don't think it would like the wet / humid environment on my boat. I was originally thinking of getting a NOCO GX4820 to charge it, but expect that it has a similar algorithm to the GEN5x1 (I have an email out to NOCO to confirm) . . . which may be sub-optimal? To your point, premade packs which get subjected to NOCO and other LiFePO4 compatible chargers must have cell level protection set to > 3.65 V or they'd have BMS fault triggers on every charge.

Thanks!

 

[mikefitz]

Consider the Victron range of chargers, all can be user programmed for a suitable charge profile. The 12v Blue Smart IP 67 would seem ideal for a boat 12v battery charger.

premade packs which get subjected to NOCO and other LiFePO4 compatible chargers must have cell level protection set to > 3.65 V or they'd have BMS fault triggers on every charge.

No, the majority seem to have cell protection set at 3.65 volts, and based on issues reported on this forum, many ready built batteries suffer from BMS shutdown due to cell overvolts.

Mike

 

[sunshine_eggo]

I was afraid someone would say that, but it's what my gut was telling me. Old addage - "Never ask a question that you don't want the answer to"!

Unfortunately, a benchtop power supply won't work well for my application. It would almost certainly get wet at some point in my boat, though I suppose I could put it in some type of box (as I did the cell pack / BMS). More importantly though, I also need a charger that I can plug in and walk away from. A nice to have would be a charger that would double as a power source when I'm playing the radio while I'm at the dock, which the GEN5x1 seems to do well (no apparent voltage spikes when disconnecting the cell pack during charging under load).

As noted earlier, I'm not an electrical engineer so this may be a dumb question: Could I put a voltage divider one of the charger legs so the max voltage that the cell pack sees is, say, 14.4 V instead of 14.6 V and then let the GEN5x1 take care of the rest? It would still do the daily top-off, but it would do it at a lower SOC, which seems harmless???

Thanks!

Would probably work but handling the current might be an issue.

A single suitably rated diode or two would lower the voltage as well.

 

[Dr_Zu]

Consider the Victron range of chargers, all can be user programmed for a suitable charge profile. The 12v Blue Smart IP 67 would seem ideal for a boat 12v battery charger.



No, the majority seem to have cell protection set at 3.65 volts, and based on issues reported on this forum, many ready built batteries suffer from BMS shutdown due to cell overvolts.

Mike

The IP67 is nice! Wish they had a 48 volt version for my boat motor battery build!

Thanks!

 

[Dr_Zu]

Would probably work but handling the current might be an issue.

A single suitably rated diode or two would lower the voltage as well.

Diodes are a great idea! I'm going to give that one a try and see how it works out. Thanks!!!

 

[Substrate]

Hi Dr.. The first thing that jumped out at me was that your trigger voltage of 3.4v for the start of balancing is too low. 3.45, perhaps 3.5v would be more appropriate.

Using 3.4v as the trigger is right at the bottom of the upper charge-knee - which sounds good but isn't. It has been discussed in older threads, but depending on the brand-specific LFP mix, 3.4v (as the trigger) can make the cell go into absorb prematurely, rather than charge to 100 percent. This is just a very long way of saying that if the trigger is set to 3.4v, one may never balance the system - too much work for the usual dinky bleeders to handle (or more of less be negligable for balance)

It depends on the cell(s) manufacture. Some people might get away with 3.4v, and some won't since this is sitting right on the bottom edge of the upper charge-knee. At the very least, nudge the trigger to 3.45v.

And, at 3.4v, the usually small bleeder resistor/circuit is more or less swamped with negligable balancing capability. You bank may be getting more unbalanced all the time at 3.4v balance trigger. But here's the fun part - when you up your trigger v to 3.45, 3.475 etc, now that the cells are "given the chance" (by going above 3.4v), you may encounter more bms trips. Just means it will take much cycling to give the bleeders a chance.

Re: the Noco. Sunshine eggo is right, but Noco did actually design it with consumer LFP in mind. Unlike most CC/CV chargers, when nearing the end of charge, it purposely ramps down the current (sometimes in a 3 or more steps) in the attempt to not have unbalanced batteries prematurely overshoot - with the forced lower current ramp down, it gives a long time for the dinky bleeder-balancers in the consumer box a better chance to do their job. So yeah, they did try to design around the common unbalanced consumer boxes. It might take a few cycles for the unknown consumer-supplied bms bleeders to balance - and whala, the bms is no longer tripping, and the Noco acts normally like it should. They just have no control over what the consumer-supplied bms and initial cell balance is.

P.S. Nice tidy build with wire loom - I like it!