Cheap LiFePO4 BMS?

[nicknorman]

I think you have to try pretty hard to get lithium batteries to heat themselves. In an engine bay they will get hot from the hot engine etc, so insulating a bit will help. Since the problem seems to be condensation I would suggest a thin layer of insulation on the things onto which water condenses. So wrap the exposed area of the bus bars in some sort of foam tape. Attach thin foam sheet to the underside of the lid. The aim being to ensure that any exposed metallic surfaces are covered with even just a mm or so of foam. This will stop the condensation.

 

I am a bit dubious about fixing the problem with a fan. Condensation occurs when water vapour in warmer air hits colder metal. A fan will happily transport new air containing water into the vicinity. Yes it will also help to equalise the temperature in time, but short term it could make things worse.

 

As to worrying about battery connection layout, I never really thought it was an issue with lead acid and I think it is much less of an issue with lithium.

[MoominPapa]

1 hour ago, rusty69 said:

What kind on cable length run have you got, and where did you hide the cables? To get mine under the bed would probably require a 15 ft run. The wardrobe, maybe only 7-8 ft.

It's only a few feet. The existing high current cables all terminate in a cupboard in the engine room on the bulkhead between the engineroom and  the back cabin, so the cables just have go through that bulkhead to reach the batteries on the other side. Long cables have to be fat, and long, fat, cables are expensive.

 

MP. 

[MtB]

36 minutes ago, MoominPapa said:

Long cables have to be fat, and long, fat, cables are expensive.

 

I'm not entirely convinced on this point. It depends on the planned usage. If the loads you plan to apply are small, and the current you intend to charge at is low, then the long cables don't need to be as fat. I had about 20m of 25mm2 cable on my last boat. 24V hybrid lithium bank and it worked fine. My 30A 24V alternator charged it happily at 35A without overheating, and the biggest load I ever applied was the 12V fridge, via a buck thing. 

 

Edit to add:

On reflection the biggest load was probably the water pump and shower waste drain pump running together, along with the extractor fan and lights. 

 

On even more reflection the biggest load was actually the Whispergen pre-heat, which was 20A at 24V for about 15 minutes during the start-up sequence. 

 

All the above still worked fine though! 

 

 

 

Edited January 14, 2023 by MtB

[Ronaldo47]

2 hours ago, MtB said:

 

1) You appear to have both cable connections at one end. I think it might be better for the cells long term if you moved either the +ve or the -ve wire to the opposite end of the bank. 

 

If  both connections are  made to  the cell at one end, that end cell is directly connected to the cables, whereas the cell at the opposite end  has the resistance of two strap conductors in series with itself and the cables, and pro rata for the intermediate cells.  This added resistance would mean that the cell directly connected to the cables is working harder than the other cells, with the most remote cell having the easiest life.

 

Connecting the positive cable to one end cell and the negative to the cell at the other end, means that all cells will have the resistance of one strap in series  between them and the cables, so all should, in principle, be worked equally hard.

 

Whether there is any difference in practice, depends on the resistance of the straps. If the strap resistance really is negligible compared with the cable resistance, it's not going to matter how you connect the cables. 

[rusty69]

1 hour ago, nicknorman said:

I think you have to try pretty hard to get lithium batteries to heat themselves. In an engine bay they will get hot from the hot engine etc, so insulating a bit will help. Since the problem seems to be condensation I would suggest a thin layer of insulation on the things onto which water condenses. So wrap the exposed area of the bus bars in some sort of foam tape. Attach thin foam sheet to the underside of the lid. The aim being to ensure that any exposed metallic surfaces are covered with even just a mm or so of foam. This will stop the condensation.

 

I am a bit dubious about fixing the problem with a fan. Condensation occurs when water vapour in warmer air hits colder metal. A fan will happily transport new air containing water into the vicinity. Yes it will also help to equalise the temperature in time, but short term it could make things worse.

 

As to worrying about battery connection layout, I never really thought it was an issue with lead acid and I think it is much less of an issue with lithium.

I might just insulate the whole of the copper bus bar with shrink tubing.😂

 

Actually, I wonder if it is worthwhile doing the entire thing and then cutting a bit off where the battery terminals are.

[MtB]

3 minutes ago, Ronaldo47 said:

If  both connections are  made to  the cell at one end, that end cell is directly connected to the cables, whereas the cell at the opposite end  has the resistance of two strap conductors in series with itself and the cables, and pro rata for the intermediate cells.  This added resistance would mean that the cell directly connected to the cables is working harder than the other cells, with the most remote cell having the easiest life.

 

Connecting the positive cable to one end cell and the negative to the cell at the other end, means that all cells will have the resistance of one strap in series  between them and the cables, so all should, in principle, be worked equally hard.

 

Whether there is any difference in practice, depends on the resistance of the straps. If the strap resistance really is negligible compared with the cable resistance, it's not going to matter how you connect the cables. 

 

 

That's exactly what I was trying and failing to express. Thank you.

 

I'm not so sure the resistance of the straps themselves is the problem, so much as the resistance of all the joints.

 

Fine if all your joints are perfect with absolutely zero resistance (relative to the already microscopic cell resistance), but not so fine if the odd joint is a bit dodgy and a milliOhm of resistance has developed in the year since fitting it. 

[rusty69]

Nope. Lost me. Not difficult though, to be fair.

[MtB]

Just now, rusty69 said:

Nope. Lost me. Not difficult though, to be fair.

 

 

It's one of those points people love to squabble endlessly about, which over there in the real world don't matter a toss. 

 

I guess i only mentioned it to head off the inevitable post from someone who doesn't really understand, pointing out the theoretical error and suggesting your batteries will explode or the sky will fall in if you don't correct it IMMEDIATELY. It happens in lots of threads when people post pictures of their battery banks. 

 

 

[rusty69]

Just now, MtB said:

 

 

It's one of those points people love to squabble endlessly about, which over there in the real world don't matter a toss. 

 

I guess i only mentioned it to head off the inevitable post from someone who doesn't really understand, pointing out the theoretical error and suggesting your batteries will explode or the sky will fall in if you don't correct it IMMEDIATELY. It happens in lots of threads when people post pictures of their battery banks. 

 

 

 

I would agree with you (both), but I haven't got a clue what you are going on about.😂

[MtB]

Just now, rusty69 said:

 

I would agree with you (both), but I haven't got a clue what you are going on about.😂

 

Nor have I really. 

 

 

[rusty69]

9 minutes ago, MtB said:

 

 

It's one of those points people love to squabble endlessly about, which over there in the real world don't matter a toss. 

 

I guess i only mentioned it to head off the inevitable post from someone who doesn't really understand, pointing out the theoretical error and suggesting your batteries will explode or the sky will fall in if you don't correct it IMMEDIATELY. It happens in lots of threads when people post pictures of their battery banks. 

 

 

It's a bunch of seven cells connected in parallel. I'm not sure how else it could be done.

 

ETA. I have seen different approaches using cable on the smartgauge website, but not sure how easy that would be with bus bars.

 

http://www.smartgauge.co.uk/batt_con.html

Edited January 14, 2023 by rusty69

[MtB]

SEVEN?!!!!!

 

 

 

11 minutes ago, rusty69 said:

ETA. I have seen different approaches using cable on the smartgauge website, but not sure how easy that would be with bus bars.

 

http://www.smartgauge.co.uk/batt_con.html

 

Yes that's exactly it. You've used Method 1 when Method 2 is theoretically superior. ISTR suggesting to Gibbo on here back in the mists of time it doesn't matter a toss and getting shouted down by just about everyone in the thread at the time! 

 

 

[rusty69]

11 minutes ago, MtB said:

SEVEN?!!!!!

 

 

 

 

Yes that's exactly it. You've used Method 1 when Method 2 is theoretically superior. ISTR suggesting to Gibbo on here back in the mists of time it doesn't matter a toss and getting shouted down by just about everyone in the thread at the time! 

 

 

Well, it often pays to go against the crowd.😂

[MtB]

2 minutes ago, rusty69 said:

Well, it often pays to go against the crowd.😂

 

Ok everyone, we've cleared it up by PM, and its me being a pillock. AGAIN! 

 

 

[rusty69]

3 minutes ago, MtB said:

 

Ok everyone, we've cleared it up by PM, and its me being a pillock. AGAIN! 

 

 

 

 

Hey, we can't be right all of the time. Well, I can, obviously, but not everyone can. 😂

[nicknorman]

Never been able to detect any difference in current shared by several batteries or cells in parallel with "suboptimal" wiring topology using a clamp meter, and that is with a couple of hundred amps of drain. Maybe there is 1 or 2% difference but so what?

[cuthound]

12 minutes ago, nicknorman said:

Never been able to detect any difference in current shared by several batteries or cells in parallel with "suboptimal" wiring topology using a clamp meter, and that is with a couple of hundred amps of drain. Maybe there is 1 or 2% difference but so what?

 

I agree. I have seen 400 volt UPS batteries made up of up to.6 parallel strings (200 cells in each string) with the load connections at one end, and no measurable difference between cell performance.

[George and Dragon]

On 08/01/2023 at 12:13, Ronaldo47 said:

That reminds me of a newspaper report of car accident I read some years ago where the driver, who had been wearing a seat belt, died of his injuries. The report went on to say that, if he hadn't been wearing a seat belt, his injuries would have been worse!

 

Of course, but he wouldn't have been any deader.

 

 

[BEngo]

4 hours ago, rusty69 said:

 

 

Hey, we can't be right all of the time. Well, I can, obviously, but not everyone can. 😂

You have not met my Mother-in-law.

 

N

[Tony Brooks]

On 14/01/2023 at 15:11, MtB said:

 

Yes that's exactly it. You've used Method 1 when Method 2 is theoretically superior. ISTR suggesting to Gibbo on here back in the mists of time it doesn't matter a toss and getting shouted down by just about everyone in the thread at the time! 

 

 

 

I think it is even less important in this case because unlike typical lead acid battery interlinks that are  usually a set of short individual links daisy chained so there are two connections at each battery post this one has solid copper bus bars that run the full length of the bank so unless there is a high enough load to cause volt drop along the bus bar all cells see the same voltage. The same would apply if lead acids were connected in this manner.

[PeterF]

Enlarging the photo it does not look like solid bus bars along the whole length but a series of short interlinks between each adjacent battery, so a double interlink connection at most battery terminals in the daisy chain.

[magnetman]

Well spotted. 

 

Also the A2 (304) stainless bolts might have some issues with being in direct contact with the copper. Not sure if galvanic action happens here but it seems plausible..

 

Threaded connections to the main battery terminals seems to me to be a sub optimal solution. not sure what the inserts themselves are made of but it could be 3 different metals interacting. 

 

Trouble? 

 

[Tony Brooks]

1 hour ago, PeterF said:

Enlarging the photo it does not look like solid bus bars along the whole length but a series of short interlinks between each adjacent battery, so a double interlink connection at most battery terminals in the daisy chain.

Having enlarged it I agree, so I was talking ollox for that instalation but might be something to bear in mind if starting anew.

Edited January 15, 2023 by Tony Brooks

[IanD]

Given that LFP batteries have a very flat SoC vs. voltage curve compared to LA, any small unequal voltage drops in the inter-battery connections will mean bigger SoC imbalance between them.

 

Making connections at diagonally opposite corners is easy to do and avoids this potential problem... 😉

[nicknorman]

44 minutes ago, IanD said:

Given that LFP batteries have a very flat SoC vs. voltage curve compared to LA, any small unequal voltage drops in the inter-battery connections will mean bigger SoC imbalance between them.

 

Making connections at diagonally opposite corners is easy to do and avoids this potential problem... 😉

However it is not exactly the SoC vs voltage that is relevant, it is the internal resistance. Which yes of course is a lot lower for Li. But anyway the effect is still minuscule and virtually unmeasurable. What exactly is the problem that is trying to be addressed? It in my experience there is no significant SoC imbalance and even if there was, so what? Such a SoC imbalance would only occur in mid SoCs and as upper or lower knees were approached, it would all come out in the wash. There isn’t a problem to address.

Edited January 15, 2023 by nicknorman