Cheap LiFePO4 BMS?

[eid]

2 minutes ago, Dr Bob said:

It is information that is stored at the

https://vrm.victronenergy.com

site.

The normal way of doing this is to buy a Victron venus GX or the more expensive display (the CCGX?) and connect your BMV to it. I was going to buy the Venus GX but it was £250 (IIRC). Instead I had a spare Rasp Pi around so downloaded the Victron system to put on the Pi  and connected that up to the BMV. The Pi is on all the time, connected to the BMV and to the router so is sending data every minute to the Victron VRM site which can be accessed from anywhere. We are off the boat for a few days this week and I can log in and see the power status of my batteries from 200 miles away.

If you are more interested have a look at

 

Go to post 59 which starts to detail how I did it. I am not a whizz at programming or Pi's but still managed it. In the end I had to buy a better Pi. The Pi 2 didnt work so I ended up buying a Pi 3 for £40.

 

Great! Thanks Bob.

 

On another note, I've put my battery purchase on hold while I do more research. This conversation has made me realise how much more I need to consider.

Edited November 4, 2019 by eid

[peterboat]

8 hours ago, Dr Bob said:

Oops.

You are using the Valance batteries. Wot I wrote above was for bare cell batteries. I do not know how these work on charging to 100%. Peter can you help? Can you monitor cell voltage on charging and do the isolate themselves on full charge?

Personally I would try and source bare cells i.e. Thunderskys, as you will have more control. Will th Valance batteries cope with 200-300A loads if you are using them with your start batteries? Peter doesn't have LAs in his system.

As they reach full charge they start active balance, up to that point they balance passively and as the video shows do it very well, they can hand out large amps without issue they were designed for cars/vans/minisubs, so it shouldnt be an issue for them. They do not isolate themselves when full the mater BMS does that also it balances batteries when charging, I have had these batteries in series/parallel and have had no issues with the batteries staying balanced, just connect the connectors thats what they are there for so that the batteries talk to each other, or so a very youg geek from the states proved in a language I couldnt really understand dude

[Dr Bob]

3 minutes ago, eid said:

Not cheap at £320 but it seems to deal with all of the problems raised in an alternator based lithium install. With this, some used cells, and the necessary isolators/BMS I'd still consider this a cheap system when compared to Victrons offering.

 

Before I got my Li's, I looked at the Balmar regulator but was put off by the number of wires needed to be required ....and not a clue how to interface it to my 90A Beta 43 bog standard alternator. It would be interesting to hear if you can control that alternator with the balmar controller as it does seem to do what is needed. If you are looking at spending that sort of money then maybe the sterling AtoB may work. Ring Sterling up and ask them what the voltages and current is on their 'lowest' setting.

On my system the voltages and currents I quote are unique to me. I have 660Ahr of LAs and 480Ahrs of Li's. On charge the Li voltage is circa 0.2V lower than the LA bank so when I talk about 13.7V termination point, that is 13.9V on the LA batteries. The Li bank is about 5 metres away from the LA's hence there is probably some voltage drop (50mm cables).

 

[Dr Bob]

11 hours ago, eid said:

I bought 2 at ~ 600 cycles as they were cheaper and I didn't think the difference was worth it. I have read from Peter and seen on you-tube people saying that these batteries have an internal BMS of some sort which at the very least balances the cells. That will do for me.

 

Time will tell if I made the right choice ?

Just a few comments on the Valence batteries.

If you choose these then you will be on the leading edge if charging via an alternator. The main experience on here, Tom, MP, Peter and myself have not done this. Tom, MP and myself all charge from an alternator but have bare cells so we can monitor cell voltage closely during charge. I think it is important as in a previous post to map out the charging voltage and current and take it up to 100% looking first for the voltage knee and second for the drop in current whilst monitoring individual cell currents. It is vital to monitor cell voltage in case there is an in-balance.

Now, I dont know the Valence batteries but looking at the video, I am not sure how easy it is to monitor voltage on all the cells at the same time. These batteries have hundreds of individual cells and not sure if it easy to monitor (Peter..what do you think...I take the point that they self balance but can you check the balance easily when getting up to 100%). On my system, I have a cell monitor display (always available...computer not needed) that displays the 4 cells. It is very quick to see if an in-balance is present. It looks to me that Thundersky bare cells are much easier to understand.

You need to be able to go up to 100% (as measured by voltage and tail current …..as MP keeps saying) to scope out your voltage/current relationship for different charging currents and understand your voltage vs SoC relationship. Without getting to 100% SoC you cant really create an accurate voltage vs SoC graph. I take mine up to 100% every 3 months to synchronise the BMV and to avoid the memory effect (which may or may not be real!).

Can you safely take the Valence batteries up to 100% and ensure the cell balance is ok as you can with the Winston Thunderskys?

[peterboat]

9 minutes ago, Dr Bob said:

Just a few comments on the Valence batteries.

If you choose these then you will be on the leading edge if charging via an alternator. The main experience on here, Tom, MP, Peter and myself have not done this. Tom, MP and myself all charge from an alternator but have bare cells so we can monitor cell voltage closely during charge. I think it is important as in a previous post to map out the charging voltage and current and take it up to 100% looking first for the voltage knee and second for the drop in current whilst monitoring individual cell currents. It is vital to monitor cell voltage in case there is an in-balance.

Now, I dont know the Valence batteries but looking at the video, I am not sure how easy it is to monitor voltage on all the cells at the same time. These batteries have hundreds of individual cells and not sure if it easy to monitor (Peter..what do you think...I take the point that they self balance but can you check the balance easily when getting up to 100%). On my system, I have a cell monitor display (always available...computer not needed) that displays the 4 cells. It is very quick to see if an in-balance is present. It looks to me that Thundersky bare cells are much easier to understand.

You need to be able to go up to 100% (as measured by voltage and tail current …..as MP keeps saying) to scope out your voltage/current relationship for different charging currents and understand your voltage vs SoC relationship. Without getting to 100% SoC you cant really create an accurate voltage vs SoC graph. I take mine up to 100% every 3 months to synchronise the BMV and to avoid the memory effect (which may or may not be real!).

Can you safely take the Valence batteries up to 100% and ensure the cell balance is ok as you can with the Winston Thunderskys?

You could Bob but as I have discovered over the years why bother? The cells passively balance because they are high quality, in real life car makers don't do this unless a long journey is being undertaken, so they must have a good reason, we have learned that if it shows 13.8 they are full and 12 volts is empty.to answer your question I have taken them to 100% and watched active balance do it's stuff but every time we do that it shortens the battery life, so I don't do it anymore 

[cuthound]

What am I missing?

 

If LifePo's have a life of 5,000 plus cycles, why are so many used ones with just a few hundred cycles on them so readily available?

[peterboat]

8 minutes ago, cuthound said:

What am I missing?

 

If LifePo's have a life of 5,000 plus cycles, why are so many used ones with just a few hundred cycles on them so readily available?

Car crash

[Richard10002]

11 hours ago, Dr Bob said:

Oops.

You are using the Valance batteries. Wot I wrote above was for bare cell batteries. I do not know how these work on charging to 100%. Peter can you help? Can you monitor cell voltage on charging and do the isolate themselves on full charge?

Personally I would try and source bare cells i.e. Thunderskys, as you will have more control. 

That's why I have given these Valances a miss, rightly or wrongly. Peter seems to do OK with them, but I prefer the idea of single cells. Assuming I trust the supplier in Prague, I can do a new 300Ah system with Sinopolys or Winstons for £1500. Alternatively, there is a UK based seller where the same system is £1900.

 

I think you got lucky with your timing and the availability of your Winstons but, if a used set appear before I take the plunge, all could be good.

[Richard10002]

2 hours ago, peterboat said:

As they reach full charge they start active balance, up to that point they balance passively and as the video shows do it very well, they can hand out large amps without issue they were designed for cars/vans/minisubs, so it shouldnt be an issue for them. They do not isolate themselves when full the mater BMS does that also it balances batteries when charging,

Hi Peter,

 

The stuff I have seen suggests that the internal BMS stuff balances the cells at the end of an absorption period, which suggests that you have to charge them to the highest voltage every now and then in order to make this happen. Seems to make sense.

 

However, you then mention a "master BMS" which I havent heard you talk about before. Is this the Valence BMS that Lithiumwerks say these batteries need, or a different BMS?

 

BTW: Are your Valences the green ones, or the black ones that we are seeing for sale?

Edited November 4, 2019 by Richard10002

[nicknorman]

Re the multiple cells in a valence battery, don’t forget that cells in parallel must by definition all have the same terminal voltage. So the apparent number of cells to monitor is identical to a thunder sky, ie 4 for a 12v system.

[Richard10002]

1 hour ago, cuthound said:

What am I missing?

 

If LifePo's have a life of 5,000 plus cycles, why are so many used ones with just a few hundred cycles on them so readily available?

They don't seem to be quite so readily available when I want to buy some!  (although I havent contacted the guy in Colchester yet).

 

They seem to be more readily available in the US....

 

I could guess that, as well as vehicle crashes, there might be some applications where, once one cell in a bank, or a few cells, have got down to 80% capacity, or so, they have to be replaced. There is lots of talk about medical applications for the Valences.

[Dr Bob]

39 minutes ago, nicknorman said:

Re the multiple cells in a valence battery, don’t forget that cells in parallel must by definition all have the same terminal voltage. So the apparent number of cells to monitor is identical to a thunder sky, ie 4 for a 12v system.

I dont think it works like that.

For my Thunderskys I have 12 * 160Ahr cells. They are wired in parallel in threes to give  3.3V 480Ahr cells which are then in series to give the overall 13.3V. You then measure the 4 in-series cells.

With the Valence, you have a 13.3V battery - just like a 12V LA - say 160Ahr which has 4 cells in series and those 4 cells are made up of many many small cells in parallel. To get a 480Ahr bank you would buy 3 of these 13.3V batteries so you will have to monitor 4 cells per battery so 12 Cells overall. It looks like each  13.3V battery will keep its own 4 cells in balance and the daisy chain of wires enables the 3 batteries to talk to each other and balance. You wouldnt have a cell voltage display like mine then as I am only looking at 4 cells.

 

Image of my cell monitor below.

 

 

1 minute ago, Richard10002 said:

They don't seem to be quite so readily available when I want to buy some!  (although I havent contacted the guy in Colchester yet).

 

They seem to be more readily available in the US....

 

I could guess that, as well as vehicle crashes, there might be some applications where, once one cell in a bank, or a few cells, have got down to 80% capacity, or so, they have to be replaced. There is lots of talk about medical applications for the Valences.

You need to give Jeremy a call. I would be surprised if he didnt have any.

I linked to another thread a bit earlier today - in response to Eid's question on my Rasp Pi - and in that thread there is a post from someone called Jan who put a link to someone in the Netherlands who supplies 2nd hand LiFePO4s to the marine market from cars/vans. Maybe worth a follow up.

[Dr Bob]

2 hours ago, cuthound said:

What am I missing?

 

If LifePo's have a life of 5,000 plus cycles, why are so many used ones with just a few hundred cycles on them so readily available?

As Peter said, car crashes. I dont think anyone would take the 2nd hand batteries out of a write off and put them in a new car.....so they end up with guys who build EVs from parts. These 2nd hand ones are ideal for us in boats. Jeremy, the guy I got mine from, builds electric cars so has a supply of the battery cells. For EV use they must be capable of delivering say a 200A load from a 200Ahr battery for X mins etc etc. Some of the 2nd hand ones are not quite up to this so cant go back in a car. MP and Tom's were like this. Not quite up to car use but ideal for boats where we never need that sort of draw. I had to pay a little more for mine as mine did pass the current test but I doubt if you would see any difference.

As time goes by, there will be more and more of these 2nd hand ones appearing. At the moment they seem be in demand so prices are high (£1000 for 480Ahrs) but I am sure this will drop as EV's become more popular.

[nicknorman]

40 minutes ago, Dr Bob said:

I dont think it works like that.

For my Thunderskys I have 12 * 160Ahr cells. They are wired in parallel in threes to give  3.3V 480Ahr cells which are then in series to give the overall 13.3V. You then measure the 4 in-series cells.

With the Valence, you have a 13.3V battery - just like a 12V LA - say 160Ahr which has 4 cells in series and those 4 cells are made up of many many small cells in parallel. To get a 480Ahr bank you would buy 3 of these 13.3V batteries so you will have to monitor 4 cells per battery so 12 Cells overall. It looks like each  13.3V battery will keep its own 4 cells in balance and the daisy chain of wires enables the 3 batteries to talk to each other and balance. You wouldnt have a cell voltage display like mine then as I am only looking at 4 cells.

 

 

Yes you are right of course, if you have multiple batteries. I was thinking of having just the one. Unless you can somehow take the lid off the Valence cells and connect them in parallel at cell level. Probably too difficult!

[MtB]

Is not there not a problem with paralleling multiple small cells? 

 

As the cells each probably degrade/lose capacity/fail at slightly different rates, would you not end up persistently passing too many coulombs through the faster degrading ones and accelerating the degradation? And vice versa on discharge? Or do the better cells take up the strain and give the poorer ones an easier time?

 

 

 

 

 

[eid]

6 hours ago, Dr Bob said:

Just a few comments on the Valence batteries.

If you choose these then you will be on the leading edge if charging via an alternator. The main experience on here, Tom, MP, Peter and myself have not done this. Tom, MP and myself all charge from an alternator but have bare cells so we can monitor cell voltage closely during charge. I think it is important as in a previous post to map out the charging voltage and current and take it up to 100% looking first for the voltage knee and second for the drop in current whilst monitoring individual cell currents. It is vital to monitor cell voltage in case there is an in-balance.

Now, I dont know the Valence batteries but looking at the video, I am not sure how easy it is to monitor voltage on all the cells at the same time. These batteries have hundreds of individual cells and not sure if it easy to monitor (Peter..what do you think...I take the point that they self balance but can you check the balance easily when getting up to 100%). On my system, I have a cell monitor display (always available...computer not needed) that displays the 4 cells. It is very quick to see if an in-balance is present. It looks to me that Thundersky bare cells are much easier to understand.

You need to be able to go up to 100% (as measured by voltage and tail current …..as MP keeps saying) to scope out your voltage/current relationship for different charging currents and understand your voltage vs SoC relationship. Without getting to 100% SoC you cant really create an accurate voltage vs SoC graph. I take mine up to 100% every 3 months to synchronise the BMV and to avoid the memory effect (which may or may not be real!).

Can you safely take the Valence batteries up to 100% and ensure the cell balance is ok as you can with the Winston Thunderskys?

 

I think this project is on hold until I understand how to regulate my alternators voltage and current output. The last thing I want is a burnt out alternator. The Balmar regulator suggested by PeterF seems perfect and I'll be reading up on it soon. Once (if) that is understood then I'll decide whether to carry on reading and look into cells, BMS etc, or buy some Valence batteries.

 

Something I've noticed over the past few months is that Lifepo4 batteries, new and used, are coming down in price.

 

Edited November 4, 2019 by eid

[eid]

Regarding Balmar regulators:

 

Quote

The Balmar ARS-5 provides precise voltage control for Balmar
high-output 12-volt alternators and other externally regulated
P-type (positive field) alternators,

 

My alternator has a regulator fitted to it which I would think is internal. Peter mentioned earlier that I may have to remove this (I guess this turns mine into an "externally regulated alternator").

 

However, how do I know if it is "P-type"?

Edited November 4, 2019 by eid

[Dr Bob]

18 minutes ago, eid said:

 

I think this project is on hold until I understand how to regulate my alternators voltage and current output. The last thing I want is a burnt out alternator. The Balmar regulator suggested by PeterF seems perfect and I'll be reading up on it soon. Once (if) that is understood then I'll decide whether to carry on reading and look into cells, BMS etc, or buy some Valence batteries.

 

Something I've noticed over the past few months is that Lifepo4 batteries, new and used, are coming down in price.

 

Let me know how you get on researching the Balmar regulator and how easy it is to fit to your alternator. What alternator have you got and what sort of volatges/currents do you get out of it with your LAs?

 

One thing I have found very useful was to have the Rasp Pi installed and working to monitor the power prior to the Li's as it gives you a good way to look at the data. If you are looking at doing it, you have plenty of time while your mull over the Li issues.

 

When I put my Li's on, my initial view in March when I got the batteries was to only charge the Li's via solar and see how the alternator would work the the Li's - being prepared to just disconnect them manually after an hour if I couldnt control the termination voltage etc. This was as we moved towards summer and our normal boating ....ie 2-3 hours a day of cruising. I was thinking as we got to winter when moored up most of the time, I would be running the engine for an hour or two a day to heat the water and therefore I could manage any 'overcharging'. I wasnt really aware at that time of the potential to destroy the alternator with high current/temperature. For us, 'winter mode' is very different to 'summer mode' and once I have put a bilge blower on to blow cold air from the bottom of the engine bay onto the alternator then it should be ok to run the AtoB on a higher setting and get 60A-70A out for an hour without getting over 95°C. I'll report back once I have fitted it next week.

[WotEver]

11 minutes ago, eid said:

However, how do I know if it is "P-type"?

The spec might tell you, but probably not. I’d suggest that there are more N-type Alts around than P-type as standard. All it means is that the regulator is in the +ve feed to the field (‘P’) as opposed to in the -ve connection (‘N’). 
 

Some alternators are simple to convert from N to P, some aren’t. Obviously, some come as standard as P-type. Best to speak to your friendly alternator specialist. 

[Dr Bob]

2 minutes ago, WotEver said:

The spec might tell you, but probably not. I’d suggest that there are more N-type Alts around than P-type as standard. All it means is that the regulator is in the +ve feed to the field (‘P’) as opposed to in the -ve connection (‘N’). 
 

Some alternators are simple to convert from N to P, some aren’t. Obviously, some come as standard as P-type. Best to speak to your friendly alternator specialist. 

Tony, while you are here, can you comment on voltage drop on battery cables. Obviously high charge current for hours will be bad for alternators but I wonder if my situation is eased by having long battery cables.

I have my 660Ahrs of LAs under the cruiser deck with the alternator wired to the domestic bank (via the AtoB). The wiring then goes forward maybe 2 metres to the inverter for the +ve and the -ve shunt for the BMV. I have then put my Li's in parallel with that circuit with the batteries under the bed and 5 metres of wiring back to the inverter and shunt. There are therefore 7 metres of cable between the two banks. All cable is 50mm.

When charging at 30-50A I see 0.2V higher voltage on the LAs. On similar high discharge the LA's are 0.2V lower. This seems to be in line with the loss if I pull the data into a voltage loss calculator

I guess the first thing here is that if I want to terminate charge on the Li's at 13.9V, then that will be 14.1V on the LAs... so we are going in the right direction. Would the current always be the same if the wiring were a different legnth? I guess the aim here is to reduce the current.

What would happen if you reduced the cable to 25mm? Obviously the voltage drop will be reduced but I guess also the current would be reduced so the alternator not working as hard(???). Someone earlier did suggest a thinner wire!

[BEngo]

Strikes me ths best way to charge LI standing alone would  be  to find a modern ECU linked alternator and control its output via its LlN interface.  The alternator is then "Standard" so is easily replaced if the magic smoke escapes.

That of  course requires an understanding of LIN and something to emulate the ECU and decide what voltage and/or current to demand from the alternator. Another raspberry job?

Any one know owt about LIN?

N

[WotEver]

3 minutes ago, Dr Bob said:

Would the current always be the same if the wiring were a different legnth?

Yup. It’s just the voltage drop that would change. 
 

4 minutes ago, Dr Bob said:

What would happen if you reduced the cable to 25mm? Obviously the voltage drop will be reduced...

No, it would be increased. But is that what you meant to write?

 

5 minutes ago, Dr Bob said:

Someone earlier did suggest a thinner wire!

That would reduce the voltage at the LIs because you’d use some of the energy to heat the cables instead. 

Just now, WotEver said:

Yup. It’s just the voltage drop that would change. 

I need to clarify that.
 

That’s assuming that the alternator (or controller if present) is in bulk. If the alternator is already at full voltage and steadily reducing current (absorption if you like) then the increased cable resistance will lower the current still further. 

[eid]

16 minutes ago, WotEver said:

Yup. It’s just the voltage drop that would change. 
 

No, it would be increased. But is that what you meant to write?

 

That would reduce the voltage at the LIs because you’d use some of the energy to heat the cables instead. 

I need to clarify that.
 

That’s assuming that the alternator (or controller if present) is in bulk. If the alternator is already at full voltage and steadily reducing current (absorption if you like) then the increased cable resistance will lower the current still further. 

 

So if I use thin enough cables I can reduce the charge voltage, heat the boat and provide light all at the same time?

 

Why didn't anyone mention this before?

[Detling]

10 minutes ago, eid said:

 

So if I use thin enough cables I can reduce the charge voltage, heat the boat and provide light all at the same time?

 

Why didn't anyone mention this before?

overdo it and the fuse effect will provide a rapid disconnect for you.

[PeterF]

1 hour ago, eid said:

Regarding Balmar regulators:

 

 

My alternator has a regulator fitted to it which I would think is internal. Peter mentioned earlier that I may have to remove this (I guess this turns mine into an "externally regulated alternator").

 

However, how do I know if it is "P-type"?

I just noticed they only do P type, some external regulators can be bought in both flavours, my alternator is P type. I have the 614 regulator, faff to program but very flexible. You could argue that it may not need the BMV to dhut it down as it has a proper float setting, so use that to limit the charge and then use the BMV as overcharge protection isolating ignition to the Balmar.