Jump to content

Cheap LiFePO4 BMS?


jetzi

Featured Posts

9 hours ago, Richard10002 said:

Hope this all makes sense... perhaps Dr. Bob could take a look and see how things look :)

 

 

Richard,

All good so far. I wouldnt be too worried. You are doing the right thing and making progress. The key here is to learn  about your battery voltages and see in practice what us guys with Li's have been prattling on about.

One caveat here is that I know little about Valance drop in 12V batteries. @peterboat is the expert and hopefully he will comment. I can help with the generics.

Firstly at 13.9V, 45A and charging each battery to a 6A tail current, you were almost full. The at rest voltages dont seem too bad. Yes when at the end of charge you will have 13.9V but on resting that will drop back to 13.35 - 13.4V ish. Very normal. Not a clue why "3" was 13.6V - need to watch that one.

I have read that you shouldnt keep your batteries at 100% so have always discharged them down from 100%. I always do this. You have had yours at 100% for nearly a week. Dont leave 'em up there for too long! (I am not sure what too long is!!!)

 

Ok, so far so good. You have charged all your batteries to full and they are at rest with reasonable voltages (except the 13.6V one). Now on to the charge stuff on "9".

I have never charged at 14.4V (and I dont think Peter has). I would be nervous at that and maybe go lower at 35A....but your regime worked. You were right to stop with cell 4 at 3.63V. It is clear the cells arent balanced as you have a 217mV delta.

 

The 6.152V maybe a concern but is it real? My BMV (when using my LAs last year) had a high voltage reading of 19.8V - I never saw it get there and no way was that 'normal' so guess it must have been a spike the system saw. Is your valance cell similar. That reading may be real or not. For now I would see if the battery behaves ok. For me it seems to be but I dont know valances!

 

You then talk about balancing. The problem with your second pic, is that the charge is now off, the voltage has dropped from 3.4/3.5V on charge to 3.35V per cell and are now away from the knee. You would expect the cell delta to be less than 20mV so likely not much balancing has occurred. Again I dont know valances but I guess they will only balance on charging. The test will be to take the battery up to 13.9V again and see if the delta is lower. This is why I like my bare cells as I can get at the individual cells and either charge or drain them at a reasonable rate. I guess the balance current in yours will be low ie <1A so could take quite a few hours at a delta of 200mV. Over to Peter, but I would be tempted to put "9" on charge with a low current and a low voltage -13.8V (????) so the balancing happens and see if the delta reduces. Maybe the balancing happens all the time but I dont see how that can happen if the deltas are so low with the batteries full but at rest.

 

So I dont think you have any real problems at the moment and need to continue to check out the voltages. Repeat what you did on "9" on the other 3 to see the cell balance. I would then be tempted to wire them all together in parallel and charge to full but I would keep the voltage down to say 13.9V and stop when any cell gets to 3.6V and then leave them connected for 24hrs then discharge a bit. Then repeat and check the cell balances.

 

eta...oops changed a 13.6 to 3.6V

 

Edited by Dr Bob
Link to comment
Share on other sites

17 minutes ago, peterboat said:

Richard the battery tells you its state of charge in the second picture it shows 45% ish remember these batteries arnt LifePo4s they have an addition of Magnesium so be careful of following the wrong advice.

The high cell voltage on that one cell is a bit of a worry, why cant you read the other batteries? if you bought them with known cycles you can put those on charge and watch them in charge

Peter, can you rely on the SoC reading on the batteries? Do yours ever give the right SoC? I notice that the SoC on the first pic is also 40% when the battery is clearly nearly full.

I have a BG8S cell meter that tells me I am 40% full when I am fully charged. It has a LiFePO4 setting but you just have to ignore it.

 

I notice now I look more at the two pics, the first one says balancing active and the second says inactive. Ref my previous post it looks like you only get balancing on charging.

Link to comment
Share on other sites

The other thing I noticed is that the transition to the knee is very marked, with concomitant rapid drop off in current (at my high charge current anyway). The point being that a very small imbalance in terms of AH / SoC, translates into a very large voltage differential when the first battery gets to the knee. So what I am saying is that it’s probably not worth getting too excited about an imbalance only visible right at the top. Provided one stops charging when the top cell gets to 3.6v or whatever, one will be pretty close to fully charged and even with a significant voltage delta the reduced SoC on the other cells probably only amounts to a % or two.

Link to comment
Share on other sites

1 minute ago, nicknorman said:

The other thing I noticed is that the transition to the knee is very marked, with concomitant rapid drop off in current (at my high charge current anyway). The point being that a very small imbalance in terms of AH / SoC, translates into a very large voltage differential when the first battery gets to the knee. So what I am saying is that it’s probably not worth getting too excited about an imbalance only visible right at the top. Provided one stops charging when the top cell gets to 3.6v or whatever, one will be pretty close to fully charged and even with a significant voltage delta the reduced SoC on the other cells probably only amounts to a % or two.

Yes exactly.

You are charging at higher voltage and higher current than I am and therefore things will happen faster. I assume the higher voltage will make the transition faster????

It is great you are now seeing in practice the voltage profiles.....it really helps cement what is happening.

You are spot on. If you keep down at 80% SoC then balancing is not an issue unless on of the cells is way out.

Link to comment
Share on other sites

1 minute ago, Dr Bob said:

Yes exactly.

You are charging at higher voltage and higher current than I am and therefore things will happen faster. I assume the higher voltage will make the transition faster????

It is great you are now seeing in practice the voltage profiles.....it really helps cement what is happening.

You are spot on. If you keep down at 80% SoC then balancing is not an issue unless on of the cells is way out.

I don’t intend to keep down at 80% - that seems a big waste of 20% of capacity! It seems to me that if the cells are reasonably well balanced one should be able to get to 95% or more before the first cells ramps up the knee. The important thing is to monitor individual cell voltages and stop charging when the first starts to climb the knee.

Link to comment
Share on other sites

1 minute ago, nicknorman said:

I don’t intend to keep down at 80% - that seems a big waste of 20% of capacity! It seems to me that if the cells are reasonably well balanced one should be able to get to 95% or more before the first cells ramps up the knee. The important thing is to monitor individual cell voltages and stop charging when the first starts to climb the knee.

The whole 80% vs 95% argument has not really been well discussed on any of the forums. Obviously you get better life time if you avoid 100% and maybe 95% but are there any other benefits of keeping lower than 90%? I get the impression (ie no specific data to quote) that cells go out of balance more if you are pushing to the top of the charge curve so by keeping down at 80% the cells stay in balance longer. This is not going to be a problem if you have a good cell balancing system. Are there any issues over the 'memory' effect?

I know MP spent a lot of time up near 100% this summer. The rest of us seem to be operating much lower.

One benefit of having an upper limit of 80% is that you are that much further away from destruction caused by overcharging so it is far less likely to happen! On my system I have a number of levels of security ie 2 separate audible alarms and 2 separate relay disconnects (but both to just one switch) but the 80% 'normal' termination keeps me further from trouble.

I notice Tesla have 80% as the 'default' charging maximum.....but is that just to maximise battery life which is important in an EV?

  • Greenie 1
Link to comment
Share on other sites

I have only ever had my batteries at 100% once! That was when I first charged them, it was using solar and watching the computer at the same time, I could see cell balancing happening but my amps going in were very low .25 amps and less.

My current system is exactly the same on solar, on absorb its down to fractions of an amp before it goes into float

My bulk is 13.9 absorb 13.8 and float is 13.6 this normally achieves an 80% charge state on the battery according to valence software which I do believe.

I do not believe a set of cells has hit over 6 volts as the battery would not be working now [remember this battery has 400 cells in it 100 per section].

I would connect them in parallel and charge them as a bank now, I did this on my 36 volt batteries from the beginning and they had voltages ranging from 2.7 to 30 volts! at the end after charging in groups of 10 they were all balanced and now after a few months are staying that way [these batteries are running an internal BMS so are gradually discharging]

If all the batteries are connected together by the communication leads, you can read them very easily, changing from battery to battery in seconds.

I do think the BMS balances all the time not just in charging I have watched this numerous times and you can see the cell voltages wandering up and down all the time

Link to comment
Share on other sites

I wouldn’t have though that taking the battery up to a high SoC would increase imbalance, but it certainly shows it up more.

 

I think it is true that high SoC is not good for longevity. But then when the batteries are likely to outlive their owners, does it matter? And as you’ve mentioned, it is perhaps keeping at 100% for long periods that is problematic, rather than charging to 100% and then immediately starting to discharge.

 

I notice that with the latest version of iOS for my iPhone, they’ve introduced a thing called “optimised battery charging” which aims to charge only to 80% until shortly before you need it, by learning your charging routine. So in my case I plug the phone in when I go to bed around midnight, and get up around 8am. So the phone will charge to 80% between midnight and 1am, hold at 80% until 7am, then charge the last 20% in time for me to get up. Well something like that, I haven’t woken up in the middle of the night to check it! And of course those are lithium polymer batteries.

 

So for my system I propose a selector knob, 50%, 80%, 100% or whatever. Normally left on 80% but if full capacity needed, switch can be moved to 100% an hour or so before end of cruising day. Or set it to 50% if arriving at the marina to leave the boat for a few weeks.

  • Greenie 2
Link to comment
Share on other sites

15 minutes ago, Dr Bob said:

The whole 80% vs 95% argument has not really been well discussed on any of the forums. Obviously you get better life time if you avoid 100% and maybe 95% but are there any other benefits of keeping lower than 90%? I get the impression (ie no specific data to quote) that cells go out of balance more if you are pushing to the top of the charge curve so by keeping down at 80% the cells stay in balance longer. This is not going to be a problem if you have a good cell balancing system. Are there any issues over the 'memory' effect?

I know MP spent a lot of time up near 100% this summer. The rest of us seem to be operating much lower.

One benefit of having an upper limit of 80% is that you are that much further away from destruction caused by overcharging so it is far less likely to happen! On my system I have a number of levels of security ie 2 separate audible alarms and 2 separate relay disconnects (but both to just one switch) but the 80% 'normal' termination keeps me further from trouble.

I notice Tesla have 80% as the 'default' charging maximum.....but is that just to maximise battery life which is important in an EV?

Bob Tesla are a very clever company and the 80% can be overridden but its not advisable if you want to have a long relationship with the battery bank. Lets be honest Bob my bank very rarely gets below 13.1 volts, even now its at 13.3 and that is the norm so why would you want to go beyond the 13.9 volts that I go to in Bulk?

Link to comment
Share on other sites

1 hour ago, peterboat said:

Bob Tesla are a very clever company and the 80% can be overridden but its not advisable if you want to have a long relationship with the battery bank. Lets be honest Bob my bank very rarely gets below 13.1 volts, even now its at 13.3 and that is the norm so why would you want to go beyond the 13.9 volts that I go to in Bulk?

Yes, I agree with you here.

Nick seems to be taking the approach that taking the system to 95% ( ie just up to the knee) is optimum for his needs - with one of the benefits that you have more capacity. Sure it will reduce battery life time but is that significant?

My viewpoint (but only relevant for me and you) is that you loose very little only going to 80% and it prolongs battery life. Now I dont know if that is going to be 1 year, 2 years, 5 years or 15 years...but it would be signifcant if holding below 80% doubles life. It may well do. I dont find capacity an issue as in the summer we can go 4-5 days without a charge with solar and in the winter when out and about we run the engine for a hour minimum to heat the water each day so we can park up for a few days without extra charging. The big plus of 80% ish max is that I know I am many hours from the 'full' with solar blasting power in so very unlikely to get up to the knee if something goes wrong. Its just that extra layer of safety.

Also, one point that is not mentioned much on here, is the safety issue of Li's. Now we all say that it is very difficult to destroy an LiFePO4 battery (safer than all the other Li's) but the risk of damage to the surroundings is significantly higher on a 100% full battery compared to a lower charged battery in the event of a thermal runaway. The company I am involved with in Scotland are doing a project with a big Eu Aerospace company (EU grant funded) on packaging standards for air transportation of lithium batteries. We have been testing loads of cells to destruction at different states of charge. I must try and get a couple of videos to post but once you get a thermal runaway going on the worst type of Li, the effects are spectacular when 100% SoC compared to 30% SoC. I will say that it is very difficult to actually get a thermal runaway going - ie a blow lamp is not good enough (!) so we can all sleep safe in our beds - especially when our batteries are under the bed.:) They've not tried any LiFePO4s yet but I will get them to look at them. Bottom line is charging to 80% is safer than 100% but I havent got the data to compare the risks.

In any risk study you need to consider the likelyhood of the risk occuring and the effect of that risk. Obviously the effect of 100% charge will be greater than 80% charge (ie  a bigger explosion) but the likelyhood is also bigger as the voltages will be higher and thermal runaways (very unlikely on LiFePO4s) are likely to be caused by pin point temperature rises deep in the batteries ie cells shorting out and I assume this is more likely to happen as cell voltages reach 3.6V etc but I may be wrong.

I will try and get chapter and verse of the boys in the lab and report back.

I dont think all this talk of safety should sway anyone from getting LiFePO4s 'cause LAs are just as bad (lots of hydrogen around!) but you are one level of safety better if you are away from 100% with more time to react. I only ever take mine to 95% when I am there, paying close attention and ready to turn off the charge if I see a 'bad' change in voltage.

 

 

eta. Just read Nicks response on having multiple target levels for charge. I like it.

Edited by Dr Bob
  • Greenie 1
Link to comment
Share on other sites

17 minutes ago, Dr Bob said:

Yes, I agree with you here.

Nick seems to be taking the approach that taking the system to 95% ( ie just up to the knee) is optimum for his needs - with one of the benefits that you have more capacity. Sure it will reduce battery life time but is that significant?

My viewpoint (but only relevant for me and you) is that you loose very little only going to 80% and it prolongs battery life. Now I dont know if that is going to be 1 year, 2 years, 5 years or 15 years...but it would be signifcant if holding below 80% doubles life. It may well do. I dont find capacity an issue as in the summer we can go 4-5 days without a charge with solar and in the winter when out and about we run the engine for a hour minimum to heat the water each day so we can park up for a few days without extra charging. The big plus of 80% ish max is that I know I am many hours from the 'full' with solar blasting power in so very unlikely to get up to the knee if something goes wrong. Its just that extra layer of safety.

Also, one point that is not mentioned much on here, is the safety issue of Li's. Now we all say that it is very difficult to destroy an LiFePO4 battery (safer than all the other Li's) but the risk of damage to the surroundings is significantly higher on a 100% full battery compared to a lower charged battery in the event of a thermal runaway. The company I am involved with in Scotland are doing a project with a big Eu Aerospace company (EU grant funded) on packaging standards for air transportation of lithium batteries. We have been testing loads of cells to destruction at different states of charge. I must try and get a couple of videos to post but once you get a thermal runaway going on the worst type of Li, the effects are spectacular when 100% SoC compared to 30% SoC. I will say that it is very difficult to actually get a thermal runaway going - ie a blow lamp is not good enough (!) so we can all sleep safe in our beds - especially when our batteries are under the bed.:) They've not tried any LiFePO4s yet but I will get them to look at them. Bottom line is charging to 80% is safer than 100% but I havent got the data to compare the risks.

In any risk study you need to consider the likelyhood of the risk occuring and the effect of that risk. Obviously the effect of 100% charge will be greater than 80% charge (ie  a bigger explosion) but the likelyhood is also bigger as the voltages will be higher and thermal runaways (very unlikely on LiFePO4s) are likely to be caused by pin point temperature rises deep in the batteries ie cells shorting out and I assume this is more likely to happen as cell voltages reach 3.6V etc but I may be wrong.

I will try and get chapter and verse of the boys in the lab and report back.

I dont think all this talk of safety should sway anyone from getting LiFePO4s 'cause LAs are just as bad (lots of hydrogen around!) but you are one level of safety better if you are away from 100% with more time to react. I only ever take mine to 95% when I am there, paying close attention and ready to turn off the charge if I see a 'bad' change in voltage.

 

 

eta. Just read Nicks response on having multiple target levels for charge. I like it.

I think it is obvious that a battery charged to 100% has the potential to be more damaging that one at 80% - it is simply a matter of stored energy. But whether the battery at 100% has a higher propensity to suddenly decide to dump all its energy in a few seconds, compared to one at 80%, is something I don't know. As you say, LiFePO4 has a pretty good reputation for not self-destructing, and anyway I agree that it would be a bad idea to keep the battery at 100% for prolonged periods.

Link to comment
Share on other sites

Incidentally I have gone away from including current measurement in my system since it occurs to me that if I am to control the Combi's charging via CANbus (Masterbus) I will also have access to the data from the other device on the Masterbus, ie the Mastershunt which is an AH counting SoC gauge, measures battery current, SoC, voltage, battery temperature etc and sends it all down CANbus. I just hope it won't prove too difficult to decode the CANbus messages! I have now selected a PIC micro with built in CANBUS controller for the project.

Link to comment
Share on other sites

2 hours ago, nicknorman said:

As you say, LiFePO4 has a pretty good reputation for not self-destructing...

There’s that Chinese video on YouTube where they shot them, knifed them, burned them... and the batteries just sat there like a brick. 

2 hours ago, Dr Bob said:

They've not tried any LiFePO4s yet but I will get them to look at them.

I’ll be surprised if they get them to do anything untoward. 

Edited by WotEver
Typo
Link to comment
Share on other sites

On 02/12/2019 at 09:33, Dr Bob said:

Peter, can you rely on the SoC reading on the batteries? Do yours ever give the right SoC? I notice that the SoC on the first pic is also 40% when the battery is clearly nearly full.

I have a BG8S cell meter that tells me I am 40% full when I am fully charged. It has a LiFePO4 setting but you just have to ignore it.

 

I notice now I look more at the two pics, the first one says balancing active and the second says inactive. Ref my previous post it looks like you only get balancing on charging.

After posting on Saturday night I read a later Valence XP instruction manual and, amongst other stuff, it said that the batteries should be charged at 14.6V absorption, and that the external BMS, (which we don’t have), would stop the charging when one cell reached 3.8V. It talks about active balancing and sensing and correcting SOC over time. It also states that there is no problem with leaving the batteries full and on a continuous float charge.... it almost seems to expect this to be the useage.

 

So I bit the bullet, set the charger at 14.6V. The volts quickly rose to 14.6V, and the Amps dropped to a couple. When the highest cell reached just over 3.8V I stopped the charger and the voltages began to fall. I think the spread was about 250mV.

 

When the highest cell had dropped to about 3.65V, I noticed that the SOC had changed from its previous 42% or so, to 100%. It therefore looks like 3.8V on one cell is a trigger point for things to happen internally.

 

The active balancing continued after the charger had been turned off, and didn’t stop until the voltages of all the cells had fallen to about 3.45V and the spread had reduced to 30mV.

 

I am guessing that, if I left them on a float charge, with no amps flowing, balancing would continue over a long period and, maybe, the cells would end up balanced.

 

I’m going to do the same thing with the other 3 batteries, (charge at 14.6V until one cell reaches 13.8V, then stop), and see what happens. I should then have 4 full batteries, all showing 100%, ready for connection to the domestics and, hopefully, over a period, they should all self balance.

 

As Peter has said, I’ll be watching like a hawk for a while then, hopefully, they will become boringly reliable :) 

  • Greenie 1
Link to comment
Share on other sites

On 02/12/2019 at 09:55, Dr Bob said:

 

I know MP spent a lot of time up near 100% this summer. The rest of us seem to be operating much lower.

I've added a user selectable mode to the controller, for use when we're getting a wriggle on and moving every day, which terminates charge at 80%. Better to do daily cycling between 60% and 80% than between 80 and 100%.

 

For balancing, my aim it to keep the bank sufficiently well top balanced that the bank voltage/current charge termination always happens well before any cell reaches the 3.65v overvoltage threshold. This has not been difficult to do, in practice.

 

MP.

 

Link to comment
Share on other sites

19 minutes ago, Richard10002 said:

Here is the link to the later software and manual on the Valence XP Lithium batteries:

 

http://dicksbluebirdbus.x10host.com/valencefiles.html

 

Section 8 is the charging section, starting at page 31. 

I notice that these are not just LiFePO4 batteries, but also have magnesium in them. Does this affect the characteristics (voltages etc) compared to LiFePO4 without magnesium?

Link to comment
Share on other sites

23 hours ago, nicknorman said:

Incidentally I have gone away from including current measurement in my system since it occurs to me that if I am to control the Combi's charging via CANbus (Masterbus) I will also have access to the data from the other device on the Masterbus, ie the Mastershunt which is an AH counting SoC gauge, measures battery current, SoC, voltage, battery temperature etc and sends it all down CANbus. I just hope it won't prove too difficult to decode the CANbus messages! I have now selected a PIC micro with built in CANBUS controller for the project.

That sounds like a sensible approach, though I have had fun doing all the current sensing and AH counting myself and playing with Kalman filters to get better Soc numbers. Doing accurate current sensing over four decades in two directions is definitely not a trivial electronic design task.

 

MP

 

Link to comment
Share on other sites

15 minutes ago, MoominPapa said:

That sounds like a sensible approach, though I have had fun doing all the current sensing and AH counting myself and playing with Kalman filters to get better Soc numbers. Doing accurate current sensing over four decades in two directions is definitely not a trivial electronic design task.

 

MP

 

My original plan wasn’t to do an AH counting gauge - as you say, if you are integrating current over a long period it needs to be spot on with a fairly fast sample rate to avoid accumulating large errors. I just intended to get an approximation of the charging current so that I would know when to reduce the alternator voltage from “absorption” to “float” - an accuracy of 5A or so would have been sufficient.

 

But obviously I already have this data - if I can read it. I am just a bit worried that with lots of binary data on the CANbus, how I’m going first of all to work out which label is what parameter, and secondly decoding a parameter’s binary value into engineering units!
 

The Mastershunt has a Li-ion battery type option so hopefully it should cope with near unity charge efficiency factor.

 

Ive ordered a CANbus to USB sniffer from eBay, hopefully that will allow me to get an idea of what I’m letting myself in for, before committing.

Link to comment
Share on other sites

52 minutes ago, Richard10002 said:

 

The active balancing continued after the charger had been turned off, and didn’t stop until the voltages of all the cells had fallen to about 3.45V and the spread had reduced to 30mV.

 

I am guessing that, if I left them on a float charge, with no amps flowing, balancing would continue over a long period and, maybe, the cells would end up balanced.

 

 

It looks like the balancing is only happening when you are into the voltage knee ie above 3.45V. The 30mV delta is typical of what you will see at 3.45V ish - the test is to take them back up to 3.5V+ and see whether the delta goes back up to 200+mV.

 

Be careful about 'float'. If you have a cell up to 3.8V then that is full. Maybe the internal BMS will disconnect at that point but the general wisdom is you stop charging when full and not float at anything over 13.5V total bank voltage. Be ultra careful when you are charging at that high voltage and going so high in cell voltage. You are a whisker away from wrecking your bank.

 

  • Greenie 1
Link to comment
Share on other sites

16 minutes ago, nicknorman said:

I notice that these are not just LiFePO4 batteries, but also have magnesium in them. Does this affect the characteristics (voltages etc) compared to LiFePO4 without magnesium?

Yes Nick it does, they were originally for mini subs so safety was paramount, our batteries were for Smith's electric Van's, my 36 volt ones are out of buses. They all seem to  have live  a Bms in them

Link to comment
Share on other sites

2 hours ago, Richard10002 said:

After posting on Saturday night I read a later Valence XP instruction manual and, amongst other stuff, it said that the batteries should be charged at 14.6V absorption, and that the external BMS, (which we don’t have), would stop the charging when one cell reached 3.8V. It talks about active balancing and sensing and correcting SOC over time. It also states that there is no problem with leaving the batteries full and on a continuous float charge.... it almost seems to expect this to be the useage.

 

So I bit the bullet, set the charger at 14.6V. The volts quickly rose to 14.6V, and the Amps dropped to a couple. When the highest cell reached just over 3.8V I stopped the charger and the voltages began to fall. I think the spread was about 250mV.

 

When the highest cell had dropped to about 3.65V, I noticed that the SOC had changed from its previous 42% or so, to 100%. It therefore looks like 3.8V on one cell is a trigger point for things to happen internally.

 

The active balancing continued after the charger had been turned off, and didn’t stop until the voltages of all the cells had fallen to about 3.45V and the spread had reduced to 30mV.

 

I am guessing that, if I left them on a float charge, with no amps flowing, balancing would continue over a long period and, maybe, the cells would end up balanced.

 

I’m going to do the same thing with the other 3 batteries, (charge at 14.6V until one cell reaches 13.8V, then stop), and see what happens. I should then have 4 full batteries, all showing 100%, ready for connection to the domestics and, hopefully, over a period, they should all self balance.

 

As Peter has said, I’ll be watching like a hawk for a while then, hopefully, they will become boringly reliable :) 

I have a feeling that the stste of charge is done by amp counting so self discharge fools it. Your charge to full like I did when I first got my batteries must reset it

Link to comment
Share on other sites

Apologies about this long post, but just catching up and wanted to reply to a few points. Didn't realise when I started this post it would end up so long though!

 

 

On 16/11/2019 at 20:02, nicknorman said:

 

But what about when the battery temperature is below 0 and the engine is started? One wants to keep the battery connected for domestic loads. One also wants to supply boat domestic loads from the alternator. So how should the system handle this? Should it set the alternator regulation voltage to a low value so that, whilst the batteries don’t receive any charge, domestic loads can be supplied by the alternator? (if so what voltage?)   Or should it just isolate the battery and connect the alternator directly to the boat’s services without a battery in circuit (likely to get lots of ripple I suspect)? Or should it isolate the alternator from the batteries and boat services, whilst leaving the battery to supply boat services (requires a second motorised switch)? Or what?

My Arduino based alternator controller handles thiis by operating in float if battery temp is 0c or below. Battery remains connected to boat loads, and alternator. In float mode, alternator supplies loads, but aims to keep amps in/out battery as close to 0A as possible. It does this by taking info from existing shunt. 

 

 

On 16/11/2019 at 21:16, Dr Bob said:

Tom, MP and I all use an LA as a dump so there is always a battery in circuit to power the appliances with the engine running. Tom uses 2 switches, one on the charge side and one on the discharge side. The more complex you make it (ie not having an LA) and the more you try to cover all eventualities the more foolproof (or idiot proof) it has to be. 

 

If you are not going to have an LA in circuit then it does become more complex and issues like you are talking about have to be solved.

 

Just to clarify, I don't have a LA in parallel. Alternator controller and other charge sources are set so overcharge should not occur. High voltage disconnect is purely a last ditch safety system and has so far never activated. 

 

Wouldn't say it's overly complicated though. I haven't provided any protection to charge sources if disconnected due to HV event, but consider a fault would have already occurred by that point. Only difference really is my use of alternator controller to solve alternator charging problems 

 

 

On 21/11/2019 at 21:39, Richard10002 said:

I've ordered one of these - a bit like the gizmo that Will Prowse uses, and which looks quite handy for not much money:

 

https://www.amazon.co.uk/gp/product/B07KW2GMS1/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1

 

 

Having read your later posts as well, I think you should just connect then up in parallel to your existing LA bank, set your max charge voltage to 13.9 (or 13.8 if you want ti be conservative) and just start using them and benefiting from your investment 

 

If you used an automatic switch such as motorised BEP switch already discussed, or even a simple relay, you could use the above device to set an over voltage protection of 13.9, and under voltage protection of 12.2. Then just use and forget about them. With the low charge currents you talk about you'd be extremely unlikely to overcharge them just charging to 13.9v and terminating at that voltage (which the over voltage protection would do if your charge sources didn't).

 

You can over think these things sometimes (I know I did when installing mine!).

 

 

On 22/11/2019 at 15:34, peterboat said:

It does seem that high amps are the enemy for us LI users, not at the beginning but definitely beyond 75% it could easily be a disaster waiting to happen. 

Surely the ability to charge at high rates is one of the big advantages of Li over LA? Seems a shame not to be able to benefit fron thiis. 

 

 

On 22/11/2019 at 15:40, cuthound said:

 

Hopefully this has been addressed on newer Li battery BMS designs, but how would a novice buyer know?

The simple BMS @Dr Bob and I use (BMM8v2) activates low voltage disconnect at (from memory) 2.5v, and completely shuts itself down at 2.0v preventing any parasitic drain on completely flat battery. 

 

 

On 23/11/2019 at 16:24, Dr Bob said:

 

I tried the module Tom suggested to balance and didnt find it much use. Maximum current being transferred was 200mA so to transfer 10 A was going to take 50 hours.

 

That's why we decided just to leave it permanently connected and let it do its own thing. If well balanced to start with, it really doesn't take very much to keep them balanced. Ours had drifted a bit after 3 months, but we often charged to 100%, and it's only in the top knee as you get above 95% that it really stays to show. 

 

We often charge our 320Ah bank at around 80-100A which is also higher than others here. 

 

 

On 23/11/2019 at 17:05, nicknorman said:

Probably not the right place to post this, sorry, but just took the lid of the Iskra 175A alternator to see how easy it would be to connect an external regulator. Interesting to see the big diodes and the teensy (relatively) field diodes. The brushes and regulator are in one module and look to be brazed/silver soldered onto the regulator chip. I think it will be safest to buy a new module that I can butcher and attach flying leads to, keeping this one intact for “backup”! (ie cockup!). They are about £15 - £20 on eBay. Also, interesting to note that it has fans front and rear and apparently little or no flow right through. Field resistance is about 3 ohms so the max field current limit of 12A on my chosen smart regulator should be plenty.

 

73F72C8B-D19F-421C-B7B8-0906926B123B.jpeg.3e97483ae5ab0d762284988e31ee1642.jpeg

Good to see it shouldn't be too difficult to modify. I thought most alternators could probably be modified without too much difficulty, our A127 certainly was (although we've already had to replace it due to burnt out rotor!).

 

 

On 23/11/2019 at 18:29, nicknorman said:

 

The trouble with that might be that at high charge currents, differences in cell resistance might make comparing cell voltages rather inaccurate. I’ll have to experiment. I plan to experiment with some very small LiFePO4 cells so it will be easy to create charge currents in proportion to those for the real system.

But surely comparing cell voltages at end of charge is all you need? As long as all cells reach the knee as closely as possible at your charging currents what more is needed? We probably charge at higher currents than others here, and haven't noticed any difference in which cell reaches the knee first whether charging at 90A via alternator, or 10-15A via solar (in summer!).

 

 

On 23/11/2019 at 19:00, nicknorman said:

The trouble with MP’s method is that he only does it on charge, and with a fast charge that ain’t very long. I think it would be preferable to get a sense of how many AH to take out of each cell (except the lowest one, obvs!) whilst charging, using your method, then implement during the remainder of the charge, and afterwards if necessary.

 

I think the balancing is overplayed. If balancing every charge, it really should only be a very small amount so even with fast charge, time shouldn't be a problem. 

 

 

On 23/11/2019 at 19:49, nicknorman said:

Yes I did wonder about the alternator warning light and buzzer. Of course the supervising microcontroller could always operate it, the regulator chip has various possible fault flags in the LIN response frames. It would probably be useful for the microcontroller to know the ignition was on (ie time to wake up!) so a connection from switched ignition /warning light to the controller board.

 

Our alternator controller had option of output to warning light, but I preferred to keep it as visual indication of excitation and controller actually working as it should. Have therefore kept ours wired to D+ (and one side of the brushes), with the reg connected to the other brush. 

 

 

On 02/12/2019 at 09:44, nicknorman said:

I don’t intend to keep down at 80% - that seems a big waste of 20% of capacity! It seems to me that if the cells are reasonably well balanced one should be able to get to 95% or more before the first cells ramps up the knee. The important thing is to monitor individual cell voltages and stop charging when the first starts to climb the knee.

Completely agree. On our 320Ah 20% just seemed to much capacity to waste. Having said that, we rarely seem to get above 70% at the moment, but that's through choice, at least we have the option to use the full amount. I'm also of the opinion charging to 100% doesn't do much harm, it's storing them at that long term that doesn't do them any good. 

 

 

On 02/12/2019 at 09:55, Dr Bob said:

The whole 80% vs 95% argument has not really been well discussed on any of the forums. Obviously you get better life time if you avoid 100% and maybe 95% but are there any other benefits of keeping lower than 90%? I get the impression (ie no specific data to quote) that cells go out of balance more if you are pushing to the top of the charge curve so by keeping down at 80% the cells stay in balance longer. This is not going to be a problem if you have a good cell balancing system. Are there any issues over the 'memory' effect?

I know MP spent a lot of time up near 100% this summer. The rest of us seem to be operating much lower.

 

Ours were up around 100% quite often in summer. As in my reply to Nick above, from reading around its time keept at 100% that does the long term damage vs charging to 100% as such. Agree you will get better life by only charging to 80%, but by how much? Even regular charging to 100% they should last many years. 

 

I tend to agree that cells seem to drift slightly quicker if pushed to 100% though, but that's not based on any actual evidence. 

 

 

On 02/12/2019 at 11:51, nicknorman said:

Incidentally I have gone away from including current measurement in my system since it occurs to me that if I am to control the Combi's charging via CANbus (Masterbus) I will also have access to the data from the other device on the Masterbus, ie the Mastershunt which is an AH counting SoC gauge, measures battery current, SoC, voltage, battery temperature etc and sends it all down CANbus. I just hope it won't prove too difficult to decode the CANbus messages! I have now selected a PIC micro with built in CANBUS controller for the project.

Our alternator controller has CANbus interface for voltage, current, and SOC, but not having any CANbus devices no idea what information it would use or how it decides what data is relevant!

 

Tom

  • Greenie 2
Link to comment
Share on other sites

22 hours ago, nicknorman said:

I notice that these are not just LiFePO4 batteries, but also have magnesium in them. Does this affect the characteristics (voltages etc) compared to LiFePO4 without magnesium?

I did a bit of a search on tinternet but no real info to be had. This paper was the nearest I could get.

https://cpb-us-e1.wpmucdn.com/blog.umd.edu/dist/7/477/files/2018/02/45-2ljspy7.pdf

A bit complex but it seems to be saying that the Magnesium doping is giving better diffusion of Lithium ions between the cathode and anode hence (I assume) better charge and discharge rates than standard LiFePO4.  It doesnt look like the voltage changes. The technology has been around for 10 years.

In a similar manner, my Winston Thunderskys contain Yttrium so are really LiFEYPO4. The ytttrium is in there to change the low temperature performance - so the makers claim.

I looks like both these variants and the basic LiFePO4s all have the same voltage ie 3.1-3.3V ish compared to the other Li-ion batteries (ie the Cobalt oxides, manganese oxides, nickel manganese cobalt oxides etc) which have a higher specific energy (ie Wh/Kg) and hence voltages of 3.5V+. If there was any significant difference from the doping in voltage, the makers would be shouting it from the rooftops as the lower voltage of LiFePO4s is seen by many as a negative trade off. Peter seems to be quoting similar voltages to the rest of us.

 

This is bad, you've now got me looking at lithium battery chemistry!

 

 

Link to comment
Share on other sites

7 minutes ago, Dr Bob said:

I did a bit of a search on tinternet but no real info to be had. This paper was the nearest I could get.

https://cpb-us-e1.wpmucdn.com/blog.umd.edu/dist/7/477/files/2018/02/45-2ljspy7.pdf

A bit complex but it seems to be saying that the Magnesium doping is giving better diffusion of Lithium ions between the cathode and anode hence (I assume) better charge and discharge rates than standard LiFePO4.  It doesnt look like the voltage changes. The technology has been around for 10 years.

In a similar manner, my Winston Thunderskys contain Yttrium so are really LiFEYPO4. The ytttrium is in there to change the low temperature performance - so the makers claim.

I looks like both these variants and the basic LiFePO4s all have the same voltage ie 3.1-3.3V ish compared to the other Li-ion batteries (ie the Cobalt oxides, manganese oxides, nickel manganese cobalt oxides etc) which have a higher specific energy (ie Wh/Kg) and hence voltages of 3.5V+. If there was any significant difference from the doping in voltage, the makers would be shouting it from the rooftops as the lower voltage of LiFePO4s is seen by many as a negative trade off. Peter seems to be quoting similar voltages to the rest of us.

 

This is bad, you've now got me looking at lithium battery chemistry!

 

 

Just a slightly higher voltage 14.6 rather than 14.4 Richards and my batteries are from 2006 but dont know if earlier ones are around? Not that it matters mine have only been that high once it wont happen again barring a disaster!!

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.