question

damien-walker avatar image
damien-walker asked

Cable or busbar for correct midpoint voltage?

In this section of the Wiring Unlimited book, the midpoint connections are illustrated as cables. Should the midpoints be connected to a busbar like the positive and negative cables in order to get the same midpoint voltage across each series?


wiringBattery Balancer
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

3 Answers
JohnC avatar image
JohnC answered ยท

Good question @Damien Walker

In a perfect world you'd only need a smallish wire to help balance out minor midpoint differences. Certainly not a busbar.

In the real world a failed batt/cell anywhere could prompt high current flow between string midpoints, and you don't really want that. Better to let one batt/string fail 'quietly' than to have it cop the full stored energy of all strings together. Think 'meltdown'.

Many recommend fuses on the midpoint links, designed to blow before the wire melts, for pb's a hydrogen-explosion prevention thing. At least lowering the odds of that happening.

Then of course if you're monitoring remotely with a single midpoint reading (like from a BMV) you may never know a fuse is blown, and your bad string rolls on in the darkness..

Our imperfect world.. but I think a busbar is not the way to roll..


2 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

heinzie avatar image heinzie commented ยท

Curious.


Can you please explain why NOT to use busbars?


Thabks in advance!

0 Likes 0 ยท
JohnC avatar image JohnC โ™ฆ heinzie commented ยท

@Heinzie

Personal preference plays a part, but for me it would be the chance of that 'meltdown' scenario I theorized about before. I haven't ever seen that (hope I never do), but I've seen plenty hot batts from imbalance, and that's not nice either.

I use a secondary bank of 6x 6V old batts for a 12V buffer, and I'm not game to even connect the midpoints. Only lightly loaded and charged to only 13.5V for a limited time, and I periodically run a multimeter over the midpoints to check (I removed 2x (of 8) just a few weeks ago when they didn't test well).

If you have busbars across the midpoints and a good monitoring/alarm system then it might be ok.

My real preference is to have single strings only, a major reason I switched to 48V for my main bank, which carries my serious loads..

0 Likes 0 ยท
Ingo avatar image
Ingo answered ยท

This is the correct method. Equal length cable with bus bars at the main connections as per the Balancer manual.


1587660048694.png (117.0 KiB)
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

johnsmith avatar image
johnsmith answered ยท

The folowing is purely fictional and not intended to be used in any way for anyone in any cirumstance on or off earth.

Not a very helpful post re: lack of links, (as on ipad) from me here today as I don't have the literature handy at the moment, but will try and post it later.

The following is only relevant to SLA or Lead Acid monoblocks.

In summary the 4 strings max "thing" was taken from many years ago and is obsolete. Knock yourself out with as many strings as you can fit on the planet.

Even adding batteries a year down the line with extra strings has now shown to be acceptable as the strings will always do their own thing anyway. As long as you havn't thrashed them to death or totally discharged them to 0V.

The confusion comes from series/parallel in the automotive industry and experiences with lithium which don't apply here.


In summary LA are much happier in series strings, paralleled at each end. This is a whole day of reading which when I find it I will post.

Connecting strings with midpoints making them effectivley series/parallel configurations of 1 huge monoblock has not shown to be a great idea, this idea is used in the Automotive industry with lithium and is a COST EFFECTIVE lazy way of doing it. Most people sell a car after 3-4 years. Although the above wiring diagram is techically correct from a load carrying capacity of the cables, it's the fastest way to destroy and full bank of strings in a few days/weeks depending on cycle usage and SOC delpletion and give you some real pain when it comes to fault finding the bank. LA don't like balancing in parallel/series and it will always be a problem. Better to run your strings on their own and connect at each end with busbars.

However running midpoint connections for balancing with a battery balancer is a good idea and using say 4-6mm square cable with fuses between each parallel strings is all good. Maybe 6-10 amp fuses, I suggest using 10x38 fuse carriers on the wall next to the strings and using gg fuses with 20kA AIC. This will allow the strings to be balanced and stop a bad monoblock from destroying your whole bank in a week. If you work out the Adiabatic equation the 4-6mm square cable will allow more than that parallel string can throw at it in short circuit current in 200 milli seconds and quench what is needed in a real major fault current situation, off the top of my head that cable will also be good for @300Amps for a few seconds before its bonfire night. Bearing in mind your whole string can only supply the full Short circuit potential for 200 milliseconds and if doesn't explode will discharge in the next 2 seconds and be toast, hence why we have 20kA AIC+ fuses that wonโ€™t explode and fill you with ceramic making you a human teabag.

In summary aim for 2.5kA of potential carnage per string of 4x12V 100Ah batteries. (A very quick way is Cvalue * capacity) so say 20x130A=2600A. If you have AGM then double it and get some steaks for the bonfire. 4 strings will set you back between 10kA-20Ka of potential disaster, LA and AGM respectively based on 100ah batteries.

Now with lithium this gets interesting as Tesla owners that have crashed and then driven into a river are the only people who can testify they have tried finding the AIC value of a sizeable lithium bank for AIC rating calculations. I would estimate you could multiply the above figures from LA by 4 with lithium if things go very bad.

Fusing also becomes a nightmare when series/parallel configurations are setup using the same size cable as the string cable, as resistance starts to play a big part and really make it even more complex in deciding which way the currents and angry pixies will flow.

Itโ€™s a lot easier to fuse each string using AIC calcs above.

In summary use a decent set of battery balancers, e.g Victron balancers and some decent PV cable left overs, a few decent 600VDC 32A fuse cariers and some 20kA 10*38 fuses not those automotive blade fuses which are only good for welding when they fail. Connect up your midpoints with 4-6mm cable and your fuses between midpoints and also fuse between balancers and source and you are good to go. Worst that will happen is you will pop a fuse SAFELY without and damage to the carrier and you will atleast save your batteries and know there is a problem.

Just remember that when dealing with strings of batteries especially in 48VDC configurations, mega fuses and midi fuses are as much use as a choclate tea-pot as they are simply too short for the arc generated. Off topic Mega and Midi types should only be used in my opinion if protected on the incoming by a decent NH, HRC, T class fuse for the reasons explained above. When fuses that are underrated for AIC get exposed to high short circuit DC current they explode violently and in most cases Iโ€™ve seen blow rather large holes in whatever they are attached to. With fuses and DC it is a size thing, longer the arc, longer the fuse. I see a lot of people paying attention to cable size but fuse AIC and curves are far more important than any of that. This is of course if you are talking about off grid large banks and not 4 batteries in your boat or RV.


I think this post is relevant for the question asked even if the owner has lithium banks and not LA.

Most large banks will probably be LA in the DIY world. However if there are people with multiple strings of lithium, what I have written is still relevant. It does concern me that lithium is still relatively new to DIY and the consumer setting up his own gig with multiple strings of lithium is potentially real carnage if AIC is not understood correctly. These lithium batteryโ€™s can produce some serious output for seriously prolonged periods compared to SLA and it really does bother me what the potential of short circuit currents in a runaway situation can do in the DC world.

The Victron Video of the Romanian factory install in the container is a really good example of how seriously the designers have taken AIC. Look at the NH fuses used on their copper busbars in the incoming from the lithiums. They are for each 2*24V pack of lithiums and there are 10 of them. Then look at the incoming fuses pre-distribution for their main. 4 huge NH fuses in parallel. Gives you a realisation of what those lithiums can supply and NOT what those Quattros can draw.


This post ended completely off topic to the original question but I think is relevant to the original question from a safety point of view.

Stay safe

1 comment
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

arildo avatar image arildo commented ยท
Thanks for a great post. I have been looking for someone to confirm my own thoughts on this. My only comment, which I think you will agree, is that a string of 4 LA batteries giving 48V does not increase the AIC as the internal resitance in the string is equally larger. I worry more about those large 12V banks of AGM batteries where the AIC can get pretty high. Anyway, a much better design it is to get a blown fuse in the midpoint if any of the strings are bad.
0 Likes 0 ยท

Related Resources

Additional resources still need to be added for this topic