Cheap LiFePO4 BMS?

[Richard10002]

5 minutes ago, Tony1 said:

 

I still have the option to buy a victron battery protect for about £100, 

 

 

You can get the Smart 220A version for a bit over £100, so there is probably no rush to buy one?

 

https://www.batterymegastore.co.uk/product/victron-energy-smart-battery-protect-12-24v-220a-bpr122022000/

[Tony1]

13 minutes ago, Richard10002 said:

 

You can get the Smart 220A version for a bit over £100, so there is probably no rush to buy one?

 

https://www.batterymegastore.co.uk/product/victron-energy-smart-battery-protect-12-24v-220a-bpr122022000/

 

 

Thanks Richard, and it seems you may have the biblical power of prophesy - it looks like I might need a battery protect unit after all, but not for a low voltage disconnect.  

 

My plan was to use a BMV712  'in reverse' (e.g. the low SoC value as a high SoC, and not inverting the relay). 

But it doesnt seem to fully work when I want to control both SoC and voltage, for reasons that only Nick will understand.

But now that I have so much solar charge coming in, I do want a high-SoC disconnect to make sure the batteries stay below 85-90% on a day-to-day basis, so I think I'll use the BMV for that as planned.

That means I still need a high voltage disconnect for the chargers, so I might get a victron BP to manage that.

 Am I right in thinking you can use for that? And can you set its cut off voltage values in an app, or does it only receive the connect/disconnect signals from another unit? 

If it needs a trigger signal sending to it in order for it to act, then hopefully I can use one of the two existing BMVs....

 

 

Edited September 8, 2021 by Tony1

[nicknorman]

7 hours ago, Tony1 said:

 

Well here's an interesting and somewhat frustrating thing- the reason I wasnt able to trigger high-voltage events last night by fiddling with the relay values seems to be because when use 'in reverse', as I'm doing, the BMV doesnt always seem to respond as expected when you are monitoring both low SoC and low voltage, and using both 'in reverse'. 

 

You'll know why this is I'm sure, but if one value (say SoC) is at a value that keeps the relay closed, and the other parameter (say voltage) reaches a value that should cause the realy the relay to open, it seems like it does not happen- I suspect if one of the parameters is in a certain state (whether closed or open relay), this state affects the other parameter. 

 

This obviously does not happen when you use the parameters correctly (e.g. use low voltage settings to monitor and respond to low voltage events), but it seems if you have two parameters and one of them hits a value that should set the relay to a different state, the relay wont change over to the new state. 

It also doesnt happen if only use one parameter, e.g. only low SoC, or only low voltage. 

 

I'd be very grateful if you know of a way around this issue, but if not I still have the option to buy a victron battery protect for about £100, and I believe that although it is intended to be used to disconnect loads in low voltage events, it can also be used to disconnect chargers in high voltage events. It does also give an extra level of protection I guess. 

Yes you’re right. Nand vs Nor and all that. I don’t think there is a way round it. But as I said, I think that provided you synchronise the BMV timeously, a high SoC event will always be a precursor to a high voltage event. In other words, I don’t think it’s anything to worry about because even if it had worked as we’d hoped, the weak link remains the fact that your “protection” relies on the various charge sources responding correctly to a shutdown command. Provided the programmed target voltage for the various charge sources is reasonable, the probability of a failure to regulate correctly and yet still responding correctly to a shutdown command, seems extremely low.

[Tony1]

42 minutes ago, nicknorman said:

Yes you’re right. Nand vs Nor and all that. I don’t think there is a way round it. But as I said, I think that provided you synchronise the BMV timeously, a high SoC event will always be a precursor to a high voltage event. In other words, I don’t think it’s anything to worry about because even if it had worked as we’d hoped, the weak link remains the fact that your “protection” relies on the various charge sources responding correctly to a shutdown command. Provided the programmed target voltage for the various charge sources is reasonable, the probability of a failure to regulate correctly and yet still responding correctly to a shutdown command, seems extremely low.

 

Cheers Nick, not great news but I thought that would be the case.

So I have a few options.

First thought is that since over-voltage will kill a battery very quickly, whereas a prolonged 100% SoC will only shorten its life, the priority has to go to over-voltage protection. 

So my next step will be to configure my new BMV so that it only does over-voltage protection.

 

I could use the original BMV for this job, as I am doing now in fact, but I still like the idea of disconnecting only the loads when there is a low voltage event, and only the chargers when there is a high voltage event- so I am still thinking of two separate disconnect systems.

 

So do I now have to  invert the relay on the new BMV (as I did with the first one)?

And can I check- does this mean I have to move any of the three wires in the back of the new BMV? And will it change any settings I have to make to control the MPPTs?

 

I think for the time being I will have to give up on my idea of an 85% SoC disconnect, as a day to day safeguard to help prolong battery life.

I'm not yet ready to spend £160 on a third BMV, just for the sake of protection in those times when I am away from the boat.. 

I may go that way in future, depending on how much time I spend away from the boat over the next year or so. 

 

I take your point that my high voltage disconnect relies on four charger units all responding correctly to a relay change at the BMV, and put like that it doesnt sound all that safe. 

But I would add that the MPPTs seem very well behaved, and automatically went into float (13.2v) when I was trying to push the SoC up to 100% earlier to synchronise. 

And the B2Bs have very conservative settings, in fact they're actually too conservative.

The one B2B that is currently wired up didnt even get to start charging in my test earlier, because the 50 amps of solar had already pushed the battery voltage up to 14.1v, and the B2B is set to charge at 13.8v. Or maybe the control wire from the B2B stopped it. As a test, I changed the charge profile to AGM, which uses a higher voltage, and the B2B did them start charging. Certainly more testing is needed. 

 

The SoC cant be relied on either. I twiddled with my SoC value last night and forgot to set it back to a vaguely correct number, so it was reading falsely low when I tested this morning, and was saying the batteries were at 85% SoC when in fact they had gone right up the knee phase, and were heading for 14.5v. 

I need to do another test where I know I have reached 100% SoC, but I think I came close today. 

 

 

Edited September 8, 2021 by Tony1

[MtB]

19 minutes ago, Tony1 said:

I could use the original BMV for this job, as I am doing now in fact, but I still like the idea of disconnecting only the loads when there is a low voltage event, and only the chargers when there is a high voltage event- so I am still thinking of two separate disconnect systems.

 

Forgive me if I am just demonstrating my lack of understanding of this thread, but I watched a Will Prowse video the other day (or maybe read a comment in the comments section) that mentioned that "twin channel" BMSs were now becoming available, which do exactly this. They disconnect just the load or just the charge supply depending on what happened, rather than disconnecting the battery from everything.

 

Annoyingly I don't remember any specific BMS models being mentioned though. 

 

 

[Richard10002]

8 hours ago, Tony1 said:

 

 

Thanks Richard, and it seems you may have the biblical power of prophesy -

 

 

 

8 hours ago, Tony1 said:

That means I still need a high voltage disconnect for the chargers, so I might get a victron BP to manage that.

 Am I right in thinking you can use for that? And can you set its cut off voltage values in an app, or does it only receive the connect/disconnect signals from another unit? 

If it needs a trigger signal sending to it in order for it to act, then hopefully I can use one of the two existing BMVs....

 

It doesn't need a trigger, it acts in its' own right, based upon voltage. However, in the settings section, it refers only to under voltage settings and to the shut down voltage and restart voltage.

 

It looks like it is designed to protect against low voltage only, but I'm no expert. I havent tried setting them so that the shut down voltage is higher than the restart voltage, and am not confident enough to try it.

 

Having said that, I am sure that somebody, somewhere, will have  

[nicknorman]

4 minutes ago, MtB said:

 

Forgive me if I am just demonstrating my lack of understanding of this thread, but I watched a Will Prowse video the other day (or maybe read a comment in the comments section) that mentioned that "twin channel" BMSs were now becoming available, which do exactly this. They disconnect just the load or just the charge supply depending on what happened, rather than disconnecting the battery from everything.

 

Annoyingly I don't remember any specific BMS models being mentioned though. 

 

Yes you’re right. The problem is that with BMSs with internal switchery, often maximum current is quite limited and less than say a 2kw inverter will pull. Those that use external switchery seem not to have considered the use of bistable relays that require a voltage pulse to operate. I think it is problematic to find a “twin channel” BMS that has a high enough current rating or is capable of operating a zero static power consumption switching device such as the Tyco bistable relay. This of course being the original topic of this thread!

[Tony1]

16 minutes ago, MtB said:

 

Forgive me if I am just demonstrating my lack of understanding of this thread, but I watched a Will Prowse video the other day (or maybe read a comment in the comments section) that mentioned that "twin channel" BMSs were now becoming available, which do exactly this. They disconnect just the load or just the charge supply depending on what happened, rather than disconnecting the battery from everything.

 

Annoyingly I don't remember any specific BMS models being mentioned though. 

 

 

 

Thanks MtB, I'll have a gander at Will's recent videos and see if I can find it. 

My problem is that I've kind of gone down a certain route (well, stumbled, perhaps) in terms of designing lithium battery protection, instead of taking a step back on day 1, and picking the best option based on a big-picture comparison.

 

So if the irrepressible Will has a great solution but it costs £800, I'd have to think really carefully.

I dont want a dog's breakfast of electronics and fix-ups, but I also have to consider the budget, and to factor in my current situation (even if I got here by sub-optimal decisions). 

 

The frustrating thing is that none of the safeguards needed are rocket science (to an electronics engineer anyway), but still we are suspiciously eyeing the various potential solutions to find the reason why this great gadget is no good. E.g the victron battery protect- at first glance it has the ability to disconnect a charger if the voltage goes too high. Its under £100. How could I go wrong?

 

Well, even my inept and cursory perusal of the manual revealed that the overvoltage protection only happens at 16v, by which time lithium batteries will be pining for the fjords. So in practical terms, its no good.

If lithium is ever to be adopted by the boating masses, there has to be a bit more 'plug and play' about it I think. 

 

 

 

Edited September 8, 2021 by Tony1

[nicknorman]

1 hour ago, Tony1 said:

 

Cheers Nick, not great news but I thought that would be the case.

So I have a few options.

First thought is that since over-voltage will kill a battery very quickly, whereas a prolonged 100% SoC will only shorten its life, the priority has to go to over-voltage protection. 

So my next step will be to configure my new BMV so that it only does over-voltage protection.

 

I could use the original BMV for this job, as I am doing now in fact, but I still like the idea of disconnecting only the loads when there is a low voltage event, and only the chargers when there is a high voltage event- so I am still thinking of two separate disconnect systems.

 

So do I now have to  invert the relay on the new BMV (as I did with the first one)?

And can I check- does this mean I have to move any of the three wires in the back of the new BMV? And will it change any settings I have to make to control the MPPTs?

 

I think for the time being I will have to give up on my idea of an 85% SoC disconnect, as a day to day safeguard to help prolong battery life.

I'm not yet ready to spend £160 on a third BMV, just for the sake of protection in those times when I am away from the boat.. 

I may go that way in future, depending on how much time I spend away from the boat over the next year or so. 

 

I take your point that my high voltage disconnect relies on four charger units all responding correctly to a relay change at the BMV, and put like that it doesnt sound all that safe. 

But I would add that the MPPTs seem very well behaved, and automatically went into float (13.2v) when I was trying to push the SoC up to 100% earlier to synchronise. 

And the B2Bs have very conservative settings, in fact they're actually too conservative.

The one B2B that is currently wired up didnt even get to start charging in my test earlier, because the 50 amps of solar had already pushed the battery voltage up to 14.1v, and the B2B is set to charge at 13.8v. Or maybe the control wire from the B2B stopped it. As a test, I changed the charge profile to AGM, which uses a higher voltage, and the B2B did them start charging. Certainly more testing is needed. 

 

The SoC cant be relied on either. I twiddled with my SoC value last night and forgot to set it back to a vaguely correct number, so it was reading falsely low when I tested this morning, and was saying the batteries were at 85% SoC when in fact they had gone right up the knee phase, and were heading for 14.5v. 

I need to do another test where I know I have reached 100% SoC, but I think I came close today. 

Obviously it’s your choice, but IMO I think you are making the wrong choice. The tradeoff of keeping the SoC away from a prolonged 100% on a day to day basis, vs catering for a failure of one of the charging devices such that the voltage becomes excessive and yet it still responds to a shutdown command. Surely it is a better use of resources to take care of the day to day issues rather than dump them in favour of the improbable event of a voltage regulation failure?

 

If you keep the high SoC protection on BMV#2 you still have high voltage disconnect function on BMV#1. Yes this will result in the boat’s services no longer being supplied by the batteries, but bear in mind this isn’t going to happen in the dark (unless the boat is cruising in the dark) and once the source of over voltage is dealt with the batteries will be reconnected automatically. This highly improbable scenario can be catered for with blankets over the solar panels (or disconnecting the panels) and or stopping the engine. And it is pretty much guaranteed to work, it doesn’t rely on the rogue device responding to a shutdown command.

 

If despite all that you still want to use BMV#2 only for high voltage protection then you EITHER need to invert the relay  OR swap the wires on NO and NC.

Edited September 8, 2021 by nicknorman

[Richard10002]

21 minutes ago, Tony1 said:

 

My problem is that I've kind of gone down a certain route (well, stumbled, perhaps) in terms of designing lithium battery protection, instead of taking a step back on day 1, and picking the best option based on a big-picture comparison.

 

 

I am conscious that hindsight is a wonderful thing, and all that, but I found this guy, and this video, helpful and interesting.

 

He also does a good one on fitting a Victron Orion DC-DC B2B, to protect the alternator.

 

 

[Tony1]

24 minutes ago, nicknorman said:

Obviously it’s your choice, but IMO I think you are making the wrong choice. The tradeoff of keeping the SoC away from a prolonged 100% on a day to day basis, vs catering for a failure of one of the charging devices such that the voltage becomes excessive and yet it still responds to a shutdown command. Surely it is a better use of resources to take care of the day to day issues rather than dump them in favour of the improbable event of a voltage regulation failure?

 

If you keep the high SoC protection on BMV#2 you still have high voltage disconnect function on BMV#1. Yes this will result in the boat’s services no longer being supplied by the batteries, but bear in mind this isn’t going to happen in the dark (unless the boat is cruising in the dark) and once the source of over voltage is dealt with the batteries will be reconnected automatically. This highly improbable scenario can be catered for with blankets over the solar panels (or disconnecting the panels) and or stopping the engine. And it is pretty much guaranteed to work, it doesn’t rely on the rogue device responding to a shutdown command.

 

If despite all that you still want to use BMV#2 only for high voltage protection then you EITHER need to invert the relay  OR swap the wires on NO and NC.

 

I totally see where you're coming from on this question, and I'm persuaded to rethink my approach. 

The truth is that when I'm aboard, I will always have half an eye on the SoC, thats just what boaters do I think. 

And more so in sunny weather, because of the fast and prolonged charging potential of 1.4kw of panels. This morning it was a tad hazy at times, but when the sun was high I was getting about 50 amps of charge. And then when the engine gets involved, the charging voltages can climb very quickly.

 

But can the boat look after itself if for some reason I'm not around? Thats the thing that probably niggles at me more. if I'm honest.

I feel ok, apart from the constant desire to torture small mammals- but I'm in my late 50s, and unexpected things can happen...

What if something happens to me tomorrow, and I'm taken away and dont get back to the boat for a month?

In that scenario, if I dont put measures in place, the batteries will be kept at around 100% for the whole time, if there's even a little bit of sun.

What are the consequences of that for the batteries?

Who knows? Maybe it will take two years off of their useful life, Certainly it wont be good for them. 

 

So, driven by the objective of prolonging their life, and with the scenario of a novice cruising my boat, I'm coming back around to using my second BMV to manage the SoC within healthy limits. 

Could my batteries ever get to say 15volts, and the SoC still be below 85%? It doesnt seem feasible, unless I mess about with the SoC value as I did last night. 

That's one other thing- the voltage is measured live, and is a true real-time safety indicator.

The SoC is useful, but in my own case it was at least 10% wrong this morning, and if I hadnt been watching so closely, my manual editing of the SoC could have seen the batteries get to dangerously high voltage.

 

I'll be doing more tests once I have both B2Bs plus the BMV wires all connected, so I'll see how I feel after that....

[nicknorman]

8 minutes ago, Tony1 said:

 

I totally see where you're coming from on this question, and I'm persuaded to rethink my approach. 

The truth is that when I'm aboard, I will always have half an eye on the SoC, thats just what boaters do I think. 

And more so in sunny weather, because of the fast and prolonged charging potential of 1.4kw of panels. This morning it was a tad hazy at times, but when the sun was high I was getting about 50 amps of charge. And then when the engine gets involved, the charging voltages can climb very quickly.

 

But can the boat look after itself if for some reason I'm not around? Thats the thing that probably niggles at me more. if I'm honest.

I feel ok, apart from the constant desire to torture small mammals- but I'm in my late 50s, and unexpected things can happen...

What if something happens to me tomorrow, and I'm taken away and dont get back to the boat for a month?

In that scenario, if I dont put measures in place, the batteries will be kept at around 100% for the whole time, if there's even a little bit of sun.

What are the consequences of that for the batteries?

Who knows? Maybe it will take two years off of their useful life, Certainly it wont be good for them. 

 

So, driven by the objective of prolonging their life, and with the scenario of a novice cruising my boat, I'm coming back around to using my second BMV to manage the SoC within healthy limits. 

Could my batteries ever get to say 15volts, and the SoC still be below 85%? It doesnt seem feasible, unless I mess about with the SoC value as I did last night. 

That's one other thing- the voltage is measured live, and is a true real-time safety indicator.

The SoC is useful, but in my own case it was at least 10% wrong this morning, and if I hadnt been watching so closely, my manual editing of the SoC could have seen the batteries get to dangerously high voltage.

 

I'll be doing more tests once I have both B2Bs plus the BMV wires all connected, so I'll see how I feel after that....

If the MPPTs and B2Bs all have sensible settings for charge voltages, I don’t see how the battery voltage could become excessive unless there was a fault. And as I mentioned, if there is such a fault it seems quite likely that a shutdown command wouldn’t work.

[MtB]

21 minutes ago, Tony1 said:

The truth is that when I'm aboard, I will always have half an eye on the SoC, thats just what boaters do I think.

 

This is one of the LA battery mindsets that lithium users must wean themselves from. SoC does not matter (other than that one does not want to arrive a the boat at 10pm to find the batts virtually flat). A well designed BMS ensures it is neither too high nor too low. 

 

 

26 minutes ago, Tony1 said:

What if something happens to me tomorrow, and I'm taken away and dont get back to the boat for a month?

In that scenario, if I dont put measures in place, the batteries will be kept at around 100% for the whole time, if there's even a little bit of sun.

What are the consequences of that for the batteries?

 

As a leisure user, this is what bothers me too. Low SoC and the conventional BMS will disconnect, but will hold high SoC from the solar for long periods when the boat is resting in the marina unused. The cells degrade quickly like the way dog-years pass, apparently, when kept at 100% SoC or close. So any BMS that charges them up fully then holds them there indefinitely is defective, IMO.

 

What we leisure users really need is a SoC controller with an adjustable hysteresis that can be set to hover between say 40% and 60% when the boat is sleeping, but can be widened to say 20% to 90% when the boat is occupied and in use. Or even just constantly topped up to 90% during regular use.

 

But one way or another lithiums like to be exercised rather than constantly fully re-charged like LA batts have to be, and this takes a bit of adjusting to by the boater, psychologically. 

[Richard10002]

1 hour ago, Tony1 said:

 

This morning it was a tad hazy at times, but when the sun was high I was getting about 50 amps of charge. And then when the engine gets involved, the charging voltages can climb very quickly.

 

But can the boat look after itself if for some reason I'm not around? Thats the thing that probably niggles at me more. if I'm honest.

 

Could my batteries ever get to say 15volts,

 

Peterboat here tells us that he fitted his Valence batteries, set the solar to charge at no more than 13.9V, and forgot about them, (I'm exaggerating a little, but you will get the picture). I think this means a max SOC a bit below 90%.

 

Assuming there is a facility on your solar controller to set max voltage to something around say, 13.6V to 13.9V, then solar can never increase the voltage above that and, as far as I understand things, all is good. In user settings, I think mine, (MT50), is set to 13.9V "boost" and 13.4V "Float". I think they do a max of 3 hours boost each day, then revert to float.

 

You could also set similar values in your mains charger and your B2B charger/s, so they never get close to 100% SOC

 

Except that, every now and then, you need to charge up to 14.6V, 100% SOC, in order to recalibrate the SOC reading on your monitor, (BMV712).

[Tony1]

1 hour ago, MtB said:

 

This is one of the LA battery mindsets that lithium users must wean themselves from. SoC does not matter (other than that one does not want to arrive a the boat at 10pm to find the batts virtually flat). A well designed BMS ensures it is neither too high nor too low. 

 

 

I can expand my explanation a bit- the liveaboard LA users will always keep half an eye on the voltage (particularly in the afternoon) , to get an idea of the SoC, and to know whether they need to run their engine for a while, to make sure they don't get too low overnight. 

 

I dont think most lithium users worry much about the battery SoC  getting too low- their anxiety- especially in summer and if they have solar- is make sure it doesnt stay too high, for too long. I had a friend along for a leisure cruise on a sunny day not long ago, and around lunchtime she asked me why I was checking my batteries on the phone so often. I explained that the batteries were starting to approach 85% SoC, and I was picking the right time to switch off the B2B.   

Its just a thing that I do at the moment, certainly in the summer months. I'll get away with leaving them up around 100% plenty of times, of course, but the more I can avoid that., the longer they'll last.

So with that knowledge in mind, keeping an eye on the SoC is not a mindset that I want to wean myself from, at least until there is some SoC control built into my system.

 

I would qualify your statement by saying that SoC does not matter, as long as its between 20% and 80 or 90% (depends who you talk to). 

If its getting close to those limits then it does matter, and in my case, with only 400Ah, I'm finding that it approaches the upper limit most days, and at the moment with all the solar it needs a bit of manual management.

That will probably change in the winter of course, when I wont have to worry about about the solar contribution.

But I do agree with the point that a well designed BMS should take SoC into account, as it does voltage levels. 

My challenge at the moment is to finish the construction of what I hope will be an adequate BMS, but the kit involved is not dirt cheap, so there is a pressure to research and choose wisely, and without a lot of advice from Nick and others, I'd still be floundering- there just isnt enough clear and detailed info in the public sphere for electrics novices like myself of modest intelligence to get hold of, understand, and use.

 

 

 

Edited September 9, 2021 by Tony1

[MtB]

Thank you for the discussion. 

 

8 hours ago, Tony1 said:

I would qualify your statement by saying that SoC does not matter, as long as its between 20% and 80 or 90% (depends who you talk to). 

If its getting close to those limits then it does matter, and in my case, with only 400Ah, I'm finding that it approaches the upper limit most days, and at the moment with all the solar it needs a bit of manual management.

 

This, I don't understand.

 

What I meant by "SoC does not matter" in the context of a well designed lithium installation compared to a LA installation, is the BMS doss all the worrying and the managing of SoC for you, rather than you having to do it as with LA. In particular I'm not getting why you need to intervene because you have lots of solar. Does your BMS not disconnect the solar charging when SoC get to 90% (or whatever you decided on as your max SoC)?

 

This leads on to another thing I notice about lithium threads. They tend to start off with someone asking about a specific point, then become less and less focused as ther discussion becomes ever more complex, then eventually turn into a discussion of some other aspect of lithiums completely. This makes finding the knowledge and information quite difficult as thread titles rarely match the discussions within. This thread is a good example. I don't have a solution though. I started a thread long ago about how to determine the SoC of lithiums and my question never really came to a conclusion beyond 'its very complicated', IIRC. I lost the will to live, eventually. Well not literally, but I'm sure you know what I mean. 

 

So I'm going to do the same again now, and ask how you are determining the SoC of your cells  

 

How are you determining the SoC of your cells? Are you using cell voltage as a proxy? Or AH counting? Or summink else? 

 

Oh and what's all this about B2Bs? This thread is the first I've heard of them being used for lithium management. I first took them as an inventive method of limiting the alternator charge current to stop it getting hot, but there seems to be more to them than that... 

 

Thanks! 

 

 

[Tony1]

8 minutes ago, MtB said:

Thank you for the discussion. 

 

 

This, I don't understand.

 

What I meant by "SoC does not matter" in the context of a well designed lithium installation compared to a LA installation, is the BMS doss all the worrying and the managing of SoC for you, rather than you having to do it as with LA. In particular I'm not getting why you need to intervene because you have lots of solar. Does your BMS not disconnect the solar charging when SoC get to 90% (or whatever you decided on as your max SoC)?

 

This leads on to another thing I notice about lithium threads. They tend to start off with someone asking about a specific point, then become less and less focused as ther discussion becomes ever more complex, then eventually turn into a discussion of some other aspect of lithiums completely. This makes finding the knowledge and information quite difficult as thread titles rarely match the discussions within. This thread is a good example. I don't have a solution though. I started a thread long ago about how to determine the SoC of lithiums and my question never really came to a conclusion beyond 'its very complicated', IIRC. I lost the will to live, eventually. Well not literally, but I'm sure you know what I mean. 

 

So I'm going to do the same again now, and ask how you are determining the SoC of your cells  

 

How are you determining the SoC of your cells? Are you using cell voltage as a proxy? Or AH counting? Or summink else? 

 

Oh and what's all this about B2Bs? This thread is the first I've heard of them being used for lithium management. I first took them as an inventive method of limiting the alternator charge current to stop it getting hot, but there seems to be more to them than that... 

 

Thanks! 

 

 

 

I've mentioned some of this in various earlier posts, but in fairness, and as you said, the info is scattered throughout what is now a very long thread.

My problem is that I dont have a BMS fully set up yet- that is what I'm in the process of doing. The batteries dont have an internal BMS either (not that they are very use, by the sound of things- but thats another discussion).

 

I got hold of some excellent quality Valence lithiums last November at a good price, and in early Jan they were installed to replace my tired LAs.

What I immediately found was that my A2B charger was unable to control the charging voltage or current, the way it did with LAs.

Left unregulated, the lithiums were sucking 120 amps from the alternator, which went from around 50 degrees to 150 degrees very quickly, and at the end the charge cycle the voltage started to shoot up towards 15v in a matter of a minute or two, and only my careful monitoring and prompt intervention saved the batteries from being ruined. 

 

An easy to understand and affordable solution for this was a B2B system. So I got a Sterling BB1260, which limits the current to 50-55 amps, and also controls the voltage very well. 

 

Phase 3 was when my thoughts turned to protecting the batteries from low voltages and low SoC, and for that I got a BMV712 with a motorised BEP switch to do the disconnecting. 

 

Phase 4 was when I realised I needed to improve my charging capacity to avoid running the engine for 2-3 hours each day in the winter, and that involved replacing my smaller 50 amp alternator with a 90 amp unit. To safely draw the extra current now available, I got a second B2B (so one for each alternator), which has yet to be connected up.

 

Phase 5 was installing another 750 watts of solar recently, which is great but it introduces the reality (during summer) of constant charge being fed into the batteries via two MPPTs. The MPPTs seem to manage the voltage pretty well, but my next thing to look at was phase 6, which was improved management of the higher SoC. 

 

 To do this I got a second BMV712, and only use the SoC monitoring function on it, and in the next few weeks I'm going to set it up so that it disconnects the two B2Bs and the two MPPTs via their remote switching functions, when the SoC goes above a nominated value (again, thats another discussion, but at the moment I'm assuming 85% on day to day basis). 

 

The next step (I hope) will be to rewire the cabling such that the original BMV (which handles low voltage and low SoC) will only disconnect the loads, and the second BMV (monitoring high SoC) will only disconnect the chargers. At the moment the BEP switch on BMV 1 disconnects everything, which I dont think is ideal (again, thats another discussion). 

 

So thats where I'm at now, via a long (and expensive) evolution, having picked up knowledge and awareness of issues as I've gone along (and usually bumped into them). 

 

As things stand, my cobbled together BMS setup will not have a high-voltage disconnect, but it will have a high-SoC disconnect, set at 85% so that it is almost impossible for a high-voltage event to happen without a high-SoC event having already happened. 

 

I may come back and revisit this question, and at the moment it seems like another dedicated BMV712 plus a BEP switch would the the solution, but thats another £400, so I'm not rushing into that yet.

 

So thats where we are, and thats (sort of) why we got there! 

 

 

[Tony1]

10 hours ago, Richard10002 said:

 

Peterboat here tells us that he fitted his Valence batteries, set the solar to charge at no more than 13.9V, and forgot about them, (I'm exaggerating a little, but you will get the picture). I think this means a max SOC a bit below 90%.

 

Assuming there is a facility on your solar controller to set max voltage to something around say, 13.6V to 13.9V, then solar can never increase the voltage above that and, as far as I understand things, all is good. In user settings, I think mine, (MT50), is set to 13.9V "boost" and 13.4V "Float". I think they do a max of 3 hours boost each day, then revert to float.

 

You could also set similar values in your mains charger and your B2B charger/s, so they never get close to 100% SOC

 

Except that, every now and then, you need to charge up to 14.6V, 100% SOC, in order to recalibrate the SOC reading on your monitor, (BMV712).

 

Thanks Richard, and tbh I've been taking a similar approach myself, i.e. not really worrying too much about high SoC. 

Something I did pick up on (again from the very helpful Nick) is that the voltage during charging is a general indicator of SoC, but it also depends on the charging current- and its the charging current that has changed for me in recent days, with 1.4kw of panels now running.

 

E.g. a month ago, on a cloudy day and with only my engine and 650 watts of panels charging, I was seeing maybe 50 amps of charge happening (the B2B is underperforming, and thats something I need to resolve as well).  

At this relatively low charging current the voltage in the batteries was staying low, right up to 90-95% SoC. But yesterday I was getting about 50 amps from the solar alone, plus about 45 from the engine, and when the new B2B is hooked up (and they are both running properly, there will be summer days when I see 170+ amps charging.  

When there is more charging current, it seems to be push the voltage higher than it was a month ago with say 50 amps charging), even when the battery is at a similar SoC.  

 

I'm not disagreeing with you on this point, but its worth mentioning that in my case, I'll need to do a  bit of testing to see how the voltage is when I'm pushing in a lot of amps. 

 

One thing I found yesterday was that my previous charging voltage setting for the B2B might  not be suitable when the MPPTs are pushing in 50 amps and the SoC is getting up to about 70%. The voltage used to stay below 14v the whole time, right up to about 100%, but now the MMPTs are pushing the voltage up a bit, so I need to reduce them to say 13.8v, or else increase the B2B charging voltage to around 14v, otherwise it wont kick into action. 

 

I only have the most tenuous grasp of how the various factors play into it to be honest, so I'm going to be testing and tweaking to find safe and appropriate settings for both MPPTs and B2Bs that will allow both to get involved in charging, and wont stop either one charging too early, but also will prevent them raising the voltage too high. 

I suspect Peter's recommended settings of 13.8 is where I'll end up, for both B2Bs and MPPTs. 

 

 

 

 

[MtB]

12 minutes ago, Tony1 said:

 

One thing I found yesterday was that my previous charging voltage setting for the B2B might  not be suitable when the MPPTs are pushing in 50 amps and the SoC is getting up to about 70%. The voltage used to stay below 14v the whole time, right up to about 100%,

 

 

 

May I ask... how are you measuring SoC please?

 

Thanks.

 

 

Edited September 9, 2021 by MtB
Edit to pick out the right quote

[Tony1]

29 minutes ago, MtB said:

 

 

May I ask... how are you measuring SoC please?

 

Thanks.

 

 

 

Of course- and my apologies, you did ask earlier. 

The BMV and its shunt monitor the SoC- from Nick I came to understand that you tell it the total capacity, get the battery to 100%, and it then measures current movements (in and out) and deduces the SoC from those, so obviously over time it becomes less accurate and needs synchronising. 

 

As usual, there is a twist in the tale. 

I'm using the second BMV specifically to manage high SoC events, but it doesnt have high SoC as a normal parameter, so I'm using the low SoC parameter- I'm using it in reverse, so to speak.

So I set the 'threshold' value to 85%, and I did not invert the relay. So when the SoC rises above 85%, the relay opens and the chargers are triggered to switch off by a zero voltage sent to them down the control cable.

The other setting of the pair (which is to reconnect the chargers) is currently about 70% SoC. This is to allow some charge to be used and the SoC to reduce a bit, before it reconnects the chargers, and the SoC then gets pushed back up by the solar. This cycle might take a week to happen if the boat was left unattended). 

 

Using the BMV 'in reverse' with the above values means (I think) that the BMV is treating the lithiums as though they have a charge floor of 70%.

And so when it displays a value for for SoC, I think its taking the 70% as its zero point on the scale.

So when the other BMV is showing a (true) SoC of say 80%, the second BMV shows a lower value. 

Thats just a theory of course, and Nick will know one way or the other why the second BMV is showing a lower SoC (I think it was reading about 70%) within half an hour of them both being synchronised to 100%.

Certainly I saw the two BMVs giving  different SoC values yesterday anyway (after I'd synchronised both units at 100%), and I put it down to the fact that the BMV checking for high SoC is working in the way I described above, but tbh I dont really have a clue. 

 

I am a bit wary of relying too much on SoC at the moment, especially since I saw this random low value on the second BMV yesterday. 

The danger is that if the BMV thinks the SoC is lower than it really is, it will allow charging to continue even after 100% is reached - which I think I saw happening yesterday.

The voltage was rising fast, well above 14.0v, but the SoC said it was only 80% (in that case it was because I'd manually adjusted it the day before, as part of some testing).

 

 

Edited September 9, 2021 by Tony1

[nicknorman]

I don’t think the BMV discrepancy is anything to do with the SoC relay settings. I suggest you check that the battery capacity, Peukert and charge efficiency settings are all the same as per BMV #1.

 

And check the wiring. Everything that was on the load side of the first BMV shunt must be moved to the load side of the second BMV shunt. So there should be nothing on the battery side of the second BMV except the load side of the first BMV, and nothing on the load side of the first BMV except the battery side of the second BMV.

Edited September 9, 2021 by nicknorman

[nicknorman]

2 hours ago, MtB said:

 

May I ask... how are you measuring SoC please?

 

Thanks.

Ah counting using BMV 712. One problem with that is of course that without the frequent charges to 100% as per lead acid, they do drift and yet the SoC display is very alluring. After I fitted our batteries and left them for months over lockdown on shore power but with a fixed float voltage of 13.2v, the SoC indication was higher than expected for the voltage. And when I fully charged, the indicated SoC hit 100% long before actual 100% (as in perhaps 20% before, ie 120Ah).

 

So yes voltage is a good guide provided you can measure it accurately and take a reading with minimal load. Although the battery voltage doesn’t change much with load (much less than for LA) by the same token the voltage vs SoC profile is much flatter.

 

In my case, setting a precise float voltage either on the Combi or the alternator controller and then looking at SoC once any current has subsided, gives me a pretty good clue about the accuracy of the SoC display. And let’s face it, even if it is 10% out it doesn’t much matter.

Edited September 9, 2021 by nicknorman

[Tony1]

1 hour ago, nicknorman said:

I don’t think the BMV discrepancy is anything to do with the SoC relay settings. I suggest you check that the battery capacity, Peukert and charge efficiency settings are all the same as per BMV #1.

 

And check the wiring. Everything that was on the load side of the first BMV shunt must be moved to the load side of the second BMV shunt. So there should be nothing on the battery side of the second BMV except the load side of the first BMV, and nothing on the load side of the first BMV except the battery side of the second BMV.

 

Thanks Nick- to be fair I haven't seen a discrepancy again, apart from +/- a percent or so on the SoC, and a small amount on the voltage, but as you explained that is to be expected with two units next to each other in series. 

The battery Ah and other related values all seem to the same in the settings for the two BMVs, so I can only put it down to something silly that I did in the half hour after I completed the synchronisation. 

 

I must say it is very gratifying to watch the B2B stop its charging when the SoC gets up to 85%, and I'm very optimistic that when the VE cables arrive, the MPPTs will also be switched off at the designated SoC. 

I love it when a plan comes together.  Especially because in my case its a rare event.

 

Of course when boaters or civilians visit, I shall throw open the electrical cupboard and regale them- with barely a blush in my cheek- with tales of how I wired it all up.

It is a high risk strategy though- the first person that actually knows anything and asks a technical query will uncover my deceit in a moment... 

 

 

Edited September 9, 2021 by Tony1

[MtB]

On 09/09/2021 at 13:40, nicknorman said:

So yes voltage is a good guide provided you can measure it accurately and take a reading with minimal load. Although the battery voltage doesn’t change much with load (much less than for LA) by the same token the voltage vs SoC profile is much flatter.

 

 

Thanks. This is what I thought. 

 

In my thread from about three years ago about determining SoC, Dr Bob was of a different opinion IIRC. He held that cell voltages swing around significantly for a given mid range SoC. (Never minding the hockey sticks at each end of SoC.) Also for a given SoC, rested, no load cell voltage will be different according to whether the last thing to happen to the cell was a charge or a discharge. There used to be a graph on the nordkyndesign.com site illustrating this but I can't find it now that site has been revamped. Maybe it has been discredited. I'd be interested if you notice this effect or agree this effect exists, as a discerning real life user.

 

 

 

Edited September 10, 2021 by MtB
Clarify a point.

[nicknorman]

6 minutes ago, MtB said:

 

Thanks. This is what I thought. 

 

In my thread from about three years ago about determining SoC, Dr Bob was of a different opinion IIRC. He held that cell voltages swing around significantly for a given mid range SoC. (Never minding the hockey sticks at each end of SoC.) Also for a given SoC, rested, no load cell voltage will be different according to whether the last thing to happen to the cell was a charge or a discharge. There used to be a graph on the nordkyndesign.com site illustrating this but I can't find it now that site has been revamped. Maybe it has been discredited. I'd be interested if you notice this effect or agree this effect exists, as a discerning real life user.

 

There is certainly some memory effect from what happened several minutes/ an hour ago, but it's quite small. I suspect this might be more problematic for people with solar. But for us, when the engine isn't running we are in discharge mode and so the voltage does give some reasonable guide. If you are alternating between charge and discharge it might be harder.