Problem with Balmar Mc614 and LiFeP04

[ksanders]

I'd turn the question around... What requirement is there to be at > 80% SoC when the motor shuts down?

For those who cruise between slips and plug in regularly, or use solar and/or genset, or aren't going to fully discharge the battery when they stop this isn't a requirement. Burning more fuel to charge and keep the battery at a higher SoC range in this case seems unproductive to me. Use case matters when making these fine tuning adjustments.

I can appreciate your rationale, but IMO for many people there is a good argument for keeping the regulator values on the low side.

To folks that drive their boat all day and anchor off it could be duper important. to arrive at the anchorage for the night with a full set of batteries.

To others going to a dock, or planning to use their generator anyway, maybe not so much so.

 

[twistedtree]

Kevin, it looks like what's happening is that after being in float for about 6 hrs, it's kicking back into Absorb. I think this matches your settings and how the MC614 is supposed to work. Unfortunately there is no way to tell it to stay in float forever. But I think you can trick it.

I think you could get what you want by switching around how you have absorb set. Rather than setting it equal to bulk, set it equal to float. I seem to recall that with the Balmar, absorb can't be more than 0.1v less than bulk, and it may also have to be 0.1V more than float, so just be aware of those possible limits.

Anyway, if you set absorb equal to float, then as soon as bulk is complete (and you can adjust the duration), it will switch to absorb, but with a float voltage. Then it will switch to float, and after 6 hrs go back to absorb, but since float and absorb are the same voltage, it should hold where you want it indefinitely.

It otherwise seems like the MC614 voltage calibration is off since it's regulatign to 13.3V even though you have it set to 13.5V. I'd just bump it up 0.2V and see if it then holds where you want it.

A bunch of the problem is that LFP really just wants/needs two stage charging, not 3 stage. And certainly not all the stuff Balmar has that's intended to optimize lead batteries.

 

[twistedtree]

It may be a PITA to get too, but when operating the Balmar will display what it thinks the voltage is. I'll bet it's showing 0.2V higher than actual, hence regulating at 13.3V when you have it set to 13.5V.

A lot of older tech devices have crappy voltage sensing and crappy voltage regulation. With lead you could get away with it, but with LFP it's more important.

 

[Jeff F]

To folks that drive their boat all day and anchor off it could be duper important. to arrive at the anchorage for the night with a full set of batteries.

That's my use case. In my case I have a 7.5kw battery pack, and charge from alternator, solar and occasionally genset. My daily use is 2-5 kwh, and solar provides 1-4 kwh/day.

If I'm stopping for the night and am leaving the next day I have no need for a full battery. If I'm staying longer and the sun is shining I don't need a full battery. Maybe I have to start the genset if I'm anchored for days on end, and maybe that could be delayed if I started with a fuller pack, but that's for me the exception rather than the rule.

Obviously YMMV, but it's worth examining if you want to fiddle further.

 

[SteveK]

The default settings in post one Bulk 14.3v for 6 minutes appear to trigger Absorb 13.6v min 18 minutes & then float 13.4v before batteries are fully charged to 100% SOC . Then rebulk happens when bats drop to 12.7v which means hours of running the ALT doing nothing.
That is the Balmar default, what do the batteries want?
Maybe increasing time at bulk and absorb times before Float takes over is the answer.

 

[Jeff F]

I'm frankly very surprised by how high some batteries say to charge. 3.65 vpc is the outer limit for LFP cells, and many say 3.6 vpc. With so little incremental stored energy between charging to 3.5vpc and 3.6vpc, I just don't see much value is pushing it. Plus, you are so close to the edge that the slightest imbalance in cells should cause a disconnect. If you only charge to 3.5 vpc, there is a lot more leeway.

Thanks. I recall you making this point a long time ago and it's always made sense to me. I think the consumer products are mostly in a race for cheaper, lighter, smaller, and more powerful, so their desire to push it is understandable.

In the context of this discussion it may matter. It reminds me a bit of some EV charging systems where the standard charge cycle allows XX range but there is a 'high range' charge setting that extends the range by something like 10% by pushing into that incremental capacity. Bad for battery health, so use only when required.

My theory is that discussions about %SoC have to take this into consideration. Your 100% may correspond to 90% of the possible charge capacity of the bank under more aggressive thresholds. So while you can comfortably sit at 100% reported SoC someone using higher thresholds would be well advised to not keep their battery above 90% reported. Both batteries are storing the same amount of energy in this case.

ETA: I don't think we're disagreeing. It's mostly about where you mark the top end, and how far into it you want to venture. You have a conservative bulletproof arrangement.

 

[twistedtree]

Thanks. I recall you making this point a long time ago and it's always made sense to me. I think the consumer products are mostly in a race for cheaper, lighter, smaller, and more powerful, so their desire to push it is understandable.

In the context of this discussion it may matter. It reminds me a bit of some EV charging systems where the standard charge cycle allows XX range but there is a 'high range' charge setting that extends the range by something like 10% by pushing into that incremental capacity. Bad for battery health, so use only when required.

My theory is that discussions about %SoC have to take this into consideration. Your 100% may correspond to 90% of the possible charge capacity of the bank under more aggressive thresholds. So while you can comfortably sit at 100% reported SoC someone using higher thresholds would be well advised to not keep their battery above 90% reported. Both batteries are storing the same amount of energy in this case.

ETA: I don't think we're disagreeing. It's mostly about where you mark the top end, and how far into it you want to venture. You have a conservative bulletproof arrangement.

I agree that we agree

One challenge I have read about is that many drop-ins only balance when the voltage is high. I understand the reasons for this, but the side effect is that you have to charge up to their specified voltage range, even if it's higher than you would otherwise like. And I gather for many of them you need to hold the high voltage for longer than you normally would so the balancers have time to work.

One question for Barking Sands, since you seem to have a relationship with Epoch: Why not provide an Allow-to-Charge signal from the battery? All it needs to be is a wire terminal and one or two wires. These could be wired together for various battery parallel/serial battery configurations, and used to control chargers, even if just as a backup. It is such a simple solution to for alternator control and protection. It could even be used for alarming in some situations. It seems like it would be so simple to implement and phase into a battery line.

 

[Jeff F]

And I gather for many of them you need to hold the high voltage for longer than you normally would so the balancers have time to work.

...or they're trying to meet capacity ratings with less battery materials by routinely overcharging.

This doesn't apply to the OP with a 14.2v bulk, but interesting discussion.

 

[Jeff F]

It may be a PITA to get too, but when operating the Balmar will display what it thinks the voltage is. I'll bet it's showing 0.2V higher than actual, hence regulating at 13.3V when you have it set to 13.5V.

I've puzzled over this for a while. Here's my current unified theory.
- all battery mfgs seem to agree that 13.5 is an appropriate maximum float. If you charge to full and put a trickle discharge on the battery and stop at 13.5 the battery is at the top of its working range. That's 100% SoC in your case, but could be as low as 85% for super aggressive top charging.
- all voltages are quoted as resting voltages, and once below 13.5 a small V change represents a big SoC change. Further, working voltage doesn't correspond to resting voltage. If I turn on my microwave for a few minutes the battery may be 0.2v lower for a long time after. So you can't use spot voltage as a guide to much of anything.
So that 13.5 float target becomes very elusive in the case of variable and intermittent loads. You can't just apply charging current every time the voltage drops below that threshold.

I've watched the regulator's behaviour. It regularly cycles current up and down, even when bulk charging. I'm going to take a charitable view of this and suggest that it's doing this deliberately to support some non-trivial calculations in order to understand where the battery is on the charge/discharge curve based on voltage changes during the cycle, or some such data.

Your idea of setting absorb = float is interesting. That would maintain SoC in a tighter range.

In general I think we need to avoid bring LA concepts with us. The whole idea of float doesn't really fit with Lithium. You can describe it in terms of resting voltage, but in the real world it exists only as a concept.

 

[twistedtree]

I've puzzled over this for a while. Here's my current unified theory.
- all battery mfgs seem to agree that 13.5 is an appropriate maximum float. If you charge to full and put a trickle discharge on the battery and stop at 13.5 the battery is at the top of its working range. That's 100% SoC in your case, but could be as low as 85% for super aggressive top charging.
- all voltages are quoted as resting voltages, and once below 13.5 a small V change represents a big SoC change. Further, working voltage doesn't correspond to resting voltage. If I turn on my microwave for a few minutes the battery may be 0.2v lower for a long time after. So you can't use spot voltage as a guide to much of anything.
So that 13.5 float target becomes very elusive in the case of variable and intermittent loads. You can't just apply charging current every time the voltage drops below that threshold.

I've watched the regulator's behaviour. It regularly cycles current up and down, even when bulk charging. I'm going to take a charitable view of this and suggest that it's doing this deliberately to support some non-trivial calculations in order to understand where the battery is on the charge/discharge curve based on voltage changes during the cycle, or some such data.

Your idea of setting absorb = float is interesting. That would maintain SoC in a tighter range.

In general I think we need to avoid bring LA concepts with us. The whole idea of float doesn't really fit with Lithium. You can describe it in terms of resting voltage, but in the real world it exists only as a concept.

I agree that the LA charging terms largely don't apply to LFP. Yet we are all stuck trying to shoehorn LFP charging into chargers with a long legacy of LA thinking. plus a number of chargers with "advanced" LA thinking like Balmar and many others. The smarter the charger, the harder they are to outsmart. Eventually we will see chargers with programming that actually matches LFP.

I think what's needed is:

1) An initial charge voltage target for CC charging.

2) A hold time at that charge voltage target. This needs to have a wide possible range from a few minutes to hours to accommodate balancers and other battery specific stuff.

3) A maintenance voltage to hold SOC, cover DC loads, but not keep pumping A into the battery.

4) An easy way to adjust the maintenance voltage so you can maintain the battery are various SOCs depending on use.

All in all, it's a lot simpler than LA charging.

 

[Marco Flamingo]

If for some reason I want to be at full charge when stopping I'll manually kick off a full charge before stopping, but that's extremely rare. If I'm anchoring for one night I know that I have lots of power to get me through the night at > 30% SoC. That's a function of capacity vs use, but there is no harm in utilizing the bottom end of capacity with LI.

I'm curious about the "manually kick off a full charge" idea. My first generation Balmar regulator is now set for LA, but I will be changing that for LFP (by customizing my own parameters). In consideration of the issue of a BMS shutdown (whether real or not), I was looking at how to simply drop the alternator charge by switching off the exciter wire. My electrical draw is very small and I could go for hours (maybe days) with the alternator off.

Would temporarily turning off the alt "kick off a full charge" when re-energizing again and automatically be in the "bulk" phase? I assume the same could be done by shutting down the engine entirely, say when on the lunch hook at noon. I assume that would cause the regulator to run through its cycle again, beginning with bulk. If so, a simple switch could relieve amperage anxiety if a 70% SOC in the afternoon might not get one through an overnight anchorage. Simply turn off/on the switch and bulk charge the last couple hours before dropping anchor.

 

[DDW]

I've puzzled over this for a while. Here's my current unified theory.
- all battery mfgs seem to agree that 13.5 is an appropriate maximum float. If you charge to full and put a trickle discharge on the battery and stop at 13.5 the battery is at the top of its working range. That's 100% SoC in your case, but could be as low as 85% for super aggressive top charging.
- all voltages are quoted as resting voltages, and once below 13.5 a small V change represents a big SoC change. Further, working voltage doesn't correspond to resting voltage. If I turn on my microwave for a few minutes the battery may be 0.2v lower for a long time after. So you can't use spot voltage as a guide to much of anything.
So that 13.5 float target becomes very elusive in the case of variable and intermittent loads. You can't just apply charging current every time the voltage drops below that threshold.

I've watched the regulator's behaviour. It regularly cycles current up and down, even when bulk charging. I'm going to take a charitable view of this and suggest that it's doing this deliberately to support some non-trivial calculations in order to understand where the battery is on the charge/discharge curve based on voltage changes during the cycle, or some such data.

Your idea of setting absorb = float is interesting. That would maintain SoC in a tighter range.

In general I think we need to avoid bring LA concepts with us. The whole idea of float doesn't really fit with Lithium. You can describe it in terms of resting voltage, but in the real world it exists only as a concept.

This is the problem with trying to guess what is happening at the battery by looking at system bus voltage. It was a marginal idea with LA, and a bad idea with LFP. The Wakespeed and Zeus pretty much solve this problem. The best way to guess what is happening in the battery is to look at the voltage and current actually going into or out of the battery.

I couldn't make a Balmar regulator behave itself even with LA. It works OK on the sailboat where loads are somewhat steady and predictable, never worked well on the trawler. I fiddled with settings for hours. The Wakespeed does more or less what you tell it, and what you want from it.

 

[Bmarler]

This is the problem with trying to guess what is happening at the battery by looking at system bus voltage. It was a marginal idea with LA, and a bad idea with LFP. The Wakespeed and Zeus pretty much solve this problem. The best way to guess what is happening in the battery is to look at the voltage and current actually going into or out of the battery.

I couldn't make a Balmar regulator behave itself even with LA. It works OK on the sailboat where loads are somewhat steady and predictable, never worked well on the trawler. I fiddled with settings for hours. The Wakespeed does more or less what you tell it, and what you want from it.

i don't have any issue getting the balmar 618 to do exactly what i want. it doesn't just look at voltage, it looks at percent of field output (which is an indicator of current) as well to determine when to drop from absorb to float. comparing against the readings from the smart shunt at the house bank it's right on.
i'm using agm now, but don't expect any issues if and when i switch to lfp. setting advanced values using the app works great, and can be done from the pilothouse.

 

[Jeff F]

Would temporarily turning off the alt "kick off a full charge" when re-energizing again and automatically be in the "bulk" phase? I assume the same could be done by shutting down the engine entirely, say when on the lunch hook at noon.

Yes. That's what I do. I just turn the ignition off and restart. No stopping necessary. Kludgy, but not done often.

The one time my batteries are always at full charge is when I'm travelling the Canadian canals, where I'm shutting down at every lock.

 

[Jeff F]

This is the problem with trying to guess what is happening at the battery by looking at system bus voltage. It was a marginal idea with LA, and a bad idea with LFP. The Wakespeed and Zeus pretty much solve this problem. The best way to guess what is happening in the battery is to look at the voltage and current actually going into or out of the battery.

Agreed. Do these regulators use a shunt? That would go a long way in maintaining a target SoC.

 

[Jeff F]

The smarter the charger, the harder they are to outsmart.

I guess our approaches differ a bit on this. I see no need to outsmart. The deficiency we're discussing is the regulator's challenge in maintaining a specific SoC below full charge. If you're willing to give this up it's easy to live with.

 

[twistedtree]

I guess our approaches differ a bit on this. I see no need to outsmart. The deficiency we're discussing is the regulator's challenge in maintaining a specific SoC below full charge. If you're willing to give this up it's easy to live with.

I think it depends on how precisely you want to hold the SOC. If you want say 80% +/- a few %, then you probably have to use the SOC gauge and monitor/control Ah flow in/out of the battery.

Alternatively, and I what I have been doing, is to just shoot for an approximate SOC, and control everything with voltage. In the mid range of SOC for LFP, the voltage is really flat, so it alone doesn't give a very accurate SOC reading. But for storage I really don't care whether it's 60% or 70% or 80%, just that it's below about 80%. So setting the float voltage to 3.35 vpc might result in anything in the 50% to 80% range, but for storage that's just fine, and can be done with a simple regulator with only voltage sensing.

Or maybe we are talking about different things here? My point behind the previous comment is just that some LA chargers have a bunch of fancy stuff in them to optimize LA charging. Thing like adaptive absorption times, float time limits followed by forced rebulk, absorption times based on return amps, etc. Sometimes that can make it harder to get them to do what you want for LFP.

 

[Jeff F]

I should add that I've spent a couple of winters plugged in with my Multiplus set at a maximum 5a AC draw, so it cycles regularly between charging and inverting to support the house loads.

The behaviour is not much different than what I've seen with the Balmar regulation. If the battery gets low enough it'll kick off a full charge, but I can go weeks with SoC roaming around between 50-80%.

I don't see this as a Balmar issue.

 

[Jeff F]

Or maybe we are talking about different things here? My point behind the previous comment is just that some LA chargers have a bunch of fancy stuff in them to optimize LA charging. Thing like adaptive absorption times, float time limits followed by forced rebulk, absorption times based on return amps, etc. Sometimes that can make it harder to get them to do what you want for LFP.

I think we're on the same page. I'm fully in agreement that most the fancy LA stuff no longer applies. But I'm also assuming that the folks at Balmar had some insights that guided their default Lithium algorithm. I wouldn't presume to know more than they do.

Your approach is very conservative. By staying out of the danger zone you dodge the whole issue to a large degree.

I was thinking a bit more about a constant 3.5v power supply. The challenge with that approach in my mind is that if you apply a large load the constant V won't make up the difference. So with a large load the SoC will fall and is hard to recover without upping the charge voltage. So it'll work, but won't maintain SoC under all conditions.

 

[SteveK]

Will Prowse determined that 100% SOC charge was OK and also where balancing happens

 

[DDW]

i don't have any issue getting the balmar 618 to do exactly what i want. it doesn't just look at voltage, it looks at percent of field output (which is an indicator of current) as well to determine when to drop from absorb to float. comparing against the readings from the smart shunt at the house bank it's right on.
i'm using agm now, but don't expect any issues if and when i switch to lfp. setting advanced values using the app works great, and can be done from the pilothouse.

The percentage of field output is a (bad) surrogate for battery current. It is an estimate of an estimate, and says absolutely nothing about what is going into or out of the battery. It includes the 10 amps going into your refer, the 20 amps going into your engine, 150 amps going into your inverter, etc., all indistinguishable (to the Balmar) from current into the battery. In addition, the Balmar does not respect the setting you put into it in many instances. Don't know why, but measurements prove it doesn't.

The Wakespeed measures the current in the shunt going directly into the battery and only the battery. So when the battery is near fully charged and taking only 3 amps, but the other house loads are 80 amps, the Balmar guesses that the battery isn't charged as it is needing 80% field to maintain absorb voltage at 83 amps. The Wakespeed on the other hand, sees the 3 amps going into the battery, judges it to be charged, and (correctly) switches to float. Further, when the inverter turns on and draws 100 amps from the alternator, it will not (erroneously) switch back to absorb. There is no way to fix this without measuring the current into the battery.

 

[Jeff F]

Will Prowse determined that 100% SOC charge was OK and also where balancing happens

Sure. Then apply a load with a max 13.5v float value according to OEM specs and see what happens. You don't see 100% SoC again until the end of the next charge cycle. And triggering unnecessary charge cycles is not a good thing. That 100% is meant to be transient, not persistent.

Here's what I think a generic regulation looks like for LFP:
1) get the battery to 100% as part of a prescribed and well understood charge cycle
2) maintain constant 13.5v charge
3) monitor battery voltage, and kick off a charge cycle when the battery drops below some prescribed voltage thresholds.

Note that 3) will only happen if you persistently (even if intermittently) apply high draw on the battery.

That guarantees good service and good health.

It's only the arbitrary insistence on holding at a certain point in the cycle that's challenging.

In practical terms 13.5 may trigger a cycle at ~80% SoC as we saw in Kevin's case, or maybe ~60% with 13.4 and ~40% with 13.3. I think I'm at 13.2 and I'm not sure I've ever seen it trigger, but it's definitely providing power to the battery once it approaches the bottom end of its charge.

 

[SteveK]

Here is what I see remotely. The house bank (LFP) is 13.5v (13.4-13.6) for the past week. I cannot see SOC remotely.
When on the boat I do see SOC drop as there is plenty of other loads being used without a re bulk until 12.8v is reached for so many seconds (I cannot remember). It will then bulk charge, absorb and go to float 13.5v when 100% SOC is reached.
When off the boat, the only DC load is inverter standby and the fridge cycle. While the shore power is available the inverter charger only sends current when fridge comes on equal to demand, no draw off battery.
When I return I find it at 100% SOC.
What are others seeing?

 

[Bmarler]

The percentage of field output is a (bad) surrogate for battery current. It is an estimate of an estimate, and says absolutely nothing about what is going into or out of the battery. It includes the 10 amps going into your refer, the 20 amps going into your engine, 150 amps going into your inverter, etc., all indistinguishable (to the Balmar) from current into the battery. In addition, the Balmar does not respect the setting you put into it in many instances. Don't know why, but measurements prove it doesn't.

The Wakespeed measures the current in the shunt going directly into the battery and only the battery. So when the battery is near fully charged and taking only 3 amps, but the other house loads are 80 amps, the Balmar guesses that the battery isn't charged as it is needing 80% field to maintain absorb voltage at 83 amps. The Wakespeed on the other hand, sees the 3 amps going into the battery, judges it to be charged, and (correctly) switches to float. Further, when the inverter turns on and draws 100 amps from the alternator, it will not (erroneously) switch back to absorb. There is no way to fix this without measuring the current into the battery.

I don’t seem to have the issues you have. If I have a big draw to the inverter my alternator takes care of it. I don’t see any voltage fluctuations or go back into absorb. I have a big alternator and battery bank, so maybe that makes a difference too. Easier to keep the voltage flat.
I monitor the regulator, as well as the battery bank through the shunt, and it acts perfectly. Drops to float right when I’d expect.
The net effect of using the field vs a shunt are minor. If my regulator delays switching to float due to a high field load, the net effect to the bank is negligible as most of that power is consumed, not put into the bank. The voltage remains steady.
It will be interesting however, if I switch to lfp and modify that base algorithm, how differently it will react. I think I can tune it to do exactly what I want in that instance too.

 

[ksanders]

I don’t seem to have the issues you have. If I have a big draw to the inverter my alternator takes care of it. I don’t see any voltage fluctuations or go back into absorb. I have a big alternator and battery bank, so maybe that makes a difference too. Easier to keep the voltage flat.
I monitor the regulator, as well as the battery bank through the shunt, and it acts perfectly. Drops to float right when I’d expect.
The net effect of using the field vs a shunt are minor. If my regulator delays switching to float due to a high field load, the net effect to the bank is negligible as most of that power is consumed, not put into the bank. The voltage remains steady.
It will be interesting however, if I switch to lfp and modify that base algorithm, how differently it will react. I think I can tune it to do exactly what I want in that instance too.

Running the MC-614 regulator using FLA as the base profile I had the exact same results.

When I changed to LiFeP04 batteries and changed the profile the issues started.

Part of this is the nature of the LiFeP04 batteries being extremely flat in their voltage. A FLA battery has a higher internal resistance, so when you draw more current the voltage goes down. Same thing with SOC. As the FLA batteries drain, the voltage changes quite a bit. Not the same with LiFeP04 batteries.

That flat voltage makes it a challenge for any voltage regulator, as the only thing the regulator knows about the batteries is the voltage.

I will be switching to the Arco Zeus regulator in the next month or so when the work on the hard is complete. That regulator not only uses a shunt, it also through it's app provides visibility and the ability to configure the regulator.

 

[Bmarler]

Running the MC-614 regulator using FLA as the base profile I had the exact same results.

When I changed to LiFeP04 batteries and changed the profile the issues started.

Part of this is the nature of the LiFeP04 batteries being extremely flat in their voltage. A FLA battery has a higher internal resistance, so when you draw more current the voltage goes down. Same thing with SOC. As the FLA batteries drain, the voltage changes quite a bit. Not the same with LiFeP04 batteries.

That flat voltage makes it a challenge for any voltage regulator, as the only thing the regulator knows about the batteries is the voltage.

I will be switching to the Arco Zeus regulator in the next month or so when the work on the hard is complete. That regulator not only uses a shunt, it also through it's app provides visibility and the ability to configure the regulator.

Interesting. I get the allure of the lfp bank, the flat voltage is a huge plus. Regulation seems more like a resolution issue, much smaller swing than fla.
Maybe the balmar 618 I’m using has tighter voltage control than the 614. I had a 612 before, and replaced it with the 618 so I could use the phone app for setting and monitoring. Honestly, I never really scrutinized the performance of the 612. The 618 seems to hold very close to what I set it at. But, as you mentioned, yours worked well with fla. Possibly I will run into the same issue one day. I guess we’ll see.
I’ve heard good things about the zeus regulator, I’m sure you’ll be pleased.

 

[twistedtree]

I was thinking a bit more about a constant 3.5v power supply. The challenge with that approach in my mind is that if you apply a large load the constant V won't make up the difference. So with a large load the SoC will fall and is hard to recover without upping the charge voltage. So it'll work, but won't maintain SoC under all conditions.

The only experience I can relay on this is with a Wakespeed floating my batteries at 27.0V (3.375 vpc) while underway. Once the batteries are charged and floating, and sometimes before they are fully charged, I regularly run large AC loads on the boat via the inverter. This can be an electric oven, a clothes dryer, a watermaker, or even combinations of these. If the batteries aren't yet fully charged, the alternators keep putting out full power, the loads run fine, and the batteries continue charging with whatever alternator power is left over. Once the batteries are charged, the alternator regulates nicely to cover loads while maintaining 27.0V within 0.1V, and battery current hovers plus and minus, but nets out to zero. I'm very pleased with that, and think much of it is because the Wakespeed seems to do very good voltage regulation. I haven't used a Balmar with LFP, so can't compare. And the ONLY thing my wakespeed is regulating based on is voltage - well, alternator temp too, but I have yet to see it throttle the output based on temp. More specifically it has no idea what the battery current is because that input is not hooked up.

 

[twistedtree]

Sure. Then apply a load with a max 13.5v float value according to OEM specs and see what happens. You don't see 100% SoC again until the end of the next charge cycle. And triggering unnecessary charge cycles is not a good thing. That 100% is meant to be transient, not persistent.

That's interesting, and different from what I see and described in a previous post. What setup behaves this way? The chargers I have used seem to hold the SOC pretty well just by regulating float voltage. As discussed, it's a bit of a crap shoot just what SOC you will get for any given float voltage, but whatever it is, my stuff seems to hold it pretty well.

 

[SteveK]

The only experience I can relay on this is with a Wakespeed floating my batteries at 27.0V (3.375 vpc) while underway. Once the batteries are charged and floating, and sometimes before they are fully charged, I regularly run large AC loads on the boat via the inverter. This can be an electric oven, a clothes dryer, a watermaker, or even combinations of these. If the batteries aren't yet fully charged, the alternators keep putting out full power, the loads run fine, and the batteries continue charging with whatever alternator power is left over. Once the batteries are charged, the alternator regulates nicely to cover loads while maintaining 27.0V within 0.1V, and battery current hovers plus and minus, but nets out to zero. I'm very pleased with that, and think much of it is because the Wakespeed seems to do very good voltage regulation. I haven't used a Balmar with LFP, so can't compare. And the ONLY thing my wakespeed is regulating based on is voltage - well, alternator temp too, but I have yet to see it throttle the output based on temp. More specifically it has no idea what the battery current is because that input is not hooked up.

Does this mean it can be hooked up?

 

[twistedtree]

I will be switching to the Arco Zeus regulator in the next month or so when the work on the hard is complete. That regulator not only uses a shunt, it also through it's app provides visibility and the ability to configure the regulator.

I'll be real interested to hear how it works and how you like it.

The whole issue of a regulator being able to measure the current in/out of the battery I think is much less import for LFP than for lead. For lead, it's by far the best way to tell when a battery is fully charged. But because historically it has not been available to regulators, they have used other approaches like fixed absorb times, variable absorb times based on bulk times, using alternator amps rather than battery amps, etc. etc. But I would argue that with LFP you can do a very good job charging using voltage only. You can optimize a bit if you know return amps, but I think it's 5% or less. And since there is no compelling need to fully recharge an LFP battery like there is for an LA battery, who cares if you recharge to 97% or 98% or 100%? Actually, none of my LFP chargers either on the boat or at my house know anything about battery current, just battery voltage. Voltage vs SOC may be vague in the mid SOC ranges, but there is no mistaking when an LFP battery is full or empty based on voltage.

Now battery current it vital to tracking SOC, but that's a different function than charging, and the relationship between the two is really just re-synchronizing SOC Ah counting when the batteries reach full charge.

Bringing this back to the Zeus regulator, one reason I haven't tried one yet is because they require the battery shunt. Unless something has changed, it won't work without it. Putting a second battery shunt into my system would be a huge project, and all for something that I know isn't necessary for LFP. I hope someday they remove that restriction, in which case I'd be interested to try one.